UvA-DARE (Digital Academic Repository) Evidence-based ... › ws › files › 2811566 › 177075_PhD... · Evidence-based knowledge creation on orthodontic mini-implants: ‘Why

UvA-DARE is a service provided by the library of the University of Amsterdam (http://dare.uva.nl)

UvA-DARE (Digital Academic Repository)

Evidence-based knowledge creation on orthodontic mini-implants: ‘Why we know so little’

Meursinge Reynders, R.A.

Link to publication

Citation for published version (APA):Meursinge Reynders, R. A. (2016). Evidence-based knowledge creation on orthodontic mini-implants: ‘Why weknow so little’.

General rightsIt is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s),other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).

Disclaimer/Complaints regulationsIf you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, statingyour reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Askthe Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam,The Netherlands. You will be contacted as soon as possible.

Download date: 15 Jul 2020

https://dare.uva.nl/personal/pure/en/publications/evidencebased-knowledge-creation-on-orthodontic-miniimplants-why-we-know-so-little(9c3ff667-8923-4c39-b24e-74d3efcdf91d).html

Protocols of the 5 systematic reviews included in the thesis:


R.A. Meursinge Reynders

ISBN: 979-12-200-1137-2 Lay-out, editing and cover design: Reint Meursinge Reynders, Milan, Italy Printed by: Allnet multiservice center, Sede Università Statale, Via Festa del Perdono 6, 20122 Milan, Italy Copyright © 2016 Reint Meursinge Reynders All rights reserved. No part of this publication may be reported or transmitted in any form or by any means without permission of the author

Protocols of the 5 systematic reviews included in the thesis:


Protocollen van 5 systematic reviews van het academische proefschrift

Evidence-based knowledge creation on orthodontic mini-implants: ‘Why we know so little’ ter verkrijging van de graad van doctor

aan de Universiteit van Amsterdam op gezag van de Rector Magnificus

prof. dr. ir. K.I.J. Maex ten overstaan van een door het College voor Promoties ingestelde commissie,

in het openbaar te verdedigen in de Agnietenkapel op woensdag 26 oktober 2016, te 10:00 uur door Reinhart Albert Meursinge Reynders

geboren te Haarlem

This appendix contains the non-edited protocols for the following 5 systematic reviews included in this thesis: Reynders R, Ronchi L, Bipat S. Mini-implants in orthodontics: a systematic review of the literature. Am J Orthod Dentofacial Orthop. 2009;135: 564.e1-19. Meursinge Reynders R, Ronchi L, Ladu L, Van Etten-Jamaludin F, Bipat S. Insertion torque and orthodontic mini-implants: a systematic review of the artificial bone literature. Proc Inst Mech Eng H. 2013 Nov;227(11):1181-202. Meursinge Reynders RA, Ronchi L, Ladu L, van Etten-Jamaludin F, Bipat S. Insertion torque and success of orthodontic mini-implants: a systematic review. Am J Orthod Dentofacial Orthop. 2012;142(5):596-614.e5. Meursinge Reynders RA, Ladu L, Ronchi L, Di Girolamo N, De Lange J, Roberts N, Plüddemann A. Insertion torque recordings for the diagnosis of contact between orthodontic mini-implants and dental roots: a systematic review. Syst Rev. 2016 Mar; 31;5:50. Meursinge Reynders RA, Ronchi L, Ladu L, Di Girolamo N, De Lange J, Roberts N, Mickan S. Barriers and facilitators to the implementation of orthodontic mini-implants in clinical practice: a systematic review. Status: Submitted

http://www.ncbi.nlm.nih.gov/pubmed/27036120

Author’s contributions All protocols of the 5 systematic reviews were presented to the PhD committee in this separate appendix, which permitted the verification of the a priori developed research methods. Reint Meursinge Reynders conceived and designed all studies and protocols and wrote all manuscripts included in the thesis and in this appendix. He also structured both the thesis and the appendix, wrote all chapters, and is the guarantor and the first author of all protocols and full systematic reviews included in this dissertation.

Funding and conflict of interest All expenses for developing and conducting the research studies in this dissertation and its appendix were paid evenly by each contributing author. No commercial association, personal, political, academic or other possible factors can pose a conflict of interest with any part of the critical appraising or reviewing procedures.

Data sharing statement Extra data can be accessed by e-mailing [email protected]

Reprints and permission The publishing company Biomed Central granted permission to reproduce in this appendix the protocols that were published in the journal ‘Systematic Reviews’. The pertinent URL on this information is listed under here. BioMed Central [online] Available from: www.biomedcentral.com/about/policies/reprints-and-permissions (accessed June 26th 2016).

mailto:[email protected]

http://www.biomedcentral.com/about/policies/reprints-and-permissions

Contents Title page …………………………………………………………………………………………………………………..…………………… 1 Protocols for the 5 systematic reviews included in the thesis ………………………………….….……………….. 4 Author’s contributions ………………………………………………………..…………………………..…………..…….………….. 5 Funding and conflict of interest …………………………………………………………..……………………..….…..……..….. 5 Data sharing statement …………………………………………………………………………………..……..……..…….………… 5 Reprints and permission ………………………………………………………………………………………………….……………… 5 Protocol 1. ……………………………………………………………..…………………………………………..…………………………… 7 Mini-implants in orthodontics: protocol for a systematic review Protocol 2. ……………………………………………………………..………………………………………….……..…………………… 18 Insertion torque and orthodontic mini-implants: protocol for a systematic review of the artificial bone literature Protocol 3. ……………………………………………………………..………………………………………….…………….………….… 34 Insertion torque and success of orthodontic mini-implants: protocol for a systematic review Protocol 4. ……………………………………………………………..………………………………………………………..….………… 69 Insertion torque recordings for the diagnosis of contact between orthodontic mini-implants and dental roots: protocol for a systematic review Protocol 5. ……………………………………………………………..…………………………………………………..……………….. 110 Barriers and facilitators to the implementation of orthodontic mini-implants in clinical practice: a protocol for a systematic review and meta-analysis

7

Protocol 1

Mini-implants in orthodontics: protocol for a systematic review

This is the non-edited protocol for the manuscript: Reynders R, Ronchi L, Bipat S. Mini-implants in orthodontics: a systematic review of the literature. Am J Orthod Dentofacial Orthop. 2009;135: 564.e1-19.

8

Introduction Osseointegrated implants are considered reliable sources of anchorage for orthodontists1-6. However, the large size of such implants limits their usage. To overcome this problem, mini- implants have been developed7-13. Their advantage, in addition to size, include: minimal anatomic limitations, minor surgery, increased patient comfort, immediate loading, and lower costs11-15. Because these devices are used for specific time periods, mostly rely on mechanical retention, and not always osseointegrate, other terms like: miniscrews, mini-screw implants, microscrews, and temporary anchorage devices have been used16,17. At this point there is no general agreement on the nomenclature18,19. This protocol will use the term “mini-implant” in the title, because it is currently the most frequently used in the orthodontic literature. A large number of mini-implants are now available, and orthodontists are trying to incorporate them in various clinical situations. However, with the introduction of new techniques, questions normally arise. Clinicians desire information on actual success rates and possible adverse effects involved with the use of mini implants for orthodontic anchorage. They also would like to identify those variables that could influence success. Although numerous articles on these topics are available in the orthodontic literature20-26, confusion arises from the differences in their findings. Further, the currently available reviews on mini-implants are either not systematic or asked different clinical questions16,17,27-34. Given this circumstance, a systematic review of the literature is deemed appropriate. The Cochrane handbook for systematic reviews of interventions, the CONSORT guidelines, and the QUOROM statement will be used as the framework for this paper35,36.

Objectives The objectives of this review will be: 1) to record the actual success and possible negative effects of mini-implant insertion: 2) to analyze which variables influence success.

Methods

Criteria for considering studies for this review Types of studies Two categories of selection criteria will be established. General measures will be applied to find studies on mini-implants and specific selection criteria to improve the quality of the articles to include. General selection criteria for studies on mini-implants

• Studies that analyse success of mini-implants for orthodontic anchorage will be included. • Only clinical studies on humans with a minimum sample size of 10 miniscrews will be

considered. Technique articles, case reports, opinion papers, reviews, laboratory, animal and in vitro studies will be excluded.

• Implants with a diameter larger than 2.5 mm will not be included. This limit was chosen because larger implants would not classify for specific orthodontic indications; for example interradicular positioning.

9

• Articles on mini-plates will be excluded, because of their different biomechanical characteristics.

Specific selection criteria for studies on mini-implants

• Only studies that present a definition of success will be included. • Studies that do not define the duration of the application of force will be excluded. • Studies that measure implant success at time periods shorter than 120 days of force

applications will be excluded37. This arbitrary time frame will be chosen, because most orthodontic objectives cannot be completed in less than 3 months.

• Studies that measure success either at a predetermined treatment time or at the completion of orthodontic anchorage objectives will be included.

Types of participants Patients of either gender without age restrictions and with a need of absolute anchorage for orthodontic purposes will be included. Types of intervention Mini-implants will be placed in one or both jaws and will be used as absolute anchorage during orthodontic treatment. It will be evaluated if these screws could be used successfully for this objective. Types of outcome measures Outcomes will be divided in primary and secondary measures. Primary outcomes will be success or failure of mini-implants as anchorage devices during orthodontic tooth movement. Secondary outcomes will define possible complications of such treatment. Primary outcomes Immobility, mobility, displacement, and failure will be used as parameters to classify primary outcomes. These measures will be examined from the start of the application of orthodontic forces during 120 days or more37.

• Success without mobility (Score 0): Implants that presented no clinical detectable mobility and could fulfil all necessary orthodontic anchorage objectives.

• Success with mobility (Score 1): Implants that had become mobile, but could still fulfil all necessary orthodontic anchorage objectives.

• Success with displacement (Score 2): Implants that had become displaced, but could still fulfil all necessary orthodontic anchorage objectives.

• Failure (Score 3): Implants that were lost or had become unusable. This latter group includes: -- Implants that had become inoperative because of excessive tissue proliferation that could not be reversed by simple excision.

-- Implants that had caused irreversible biologic damage. -- Implants that could not be used because of the risk of causing irreversible biologic damage. -- Implants that fractured at insertion, during orthodontic treatment or at the removal of the screw.

• Not Specified Success (Score NSS): The type of success of implants was not specified and included scores 0, 1 and 2.

Secondary outcomes Secondary outcomes will be divided into 3 categories: biologic damage, inflammation, and pain and discomfort measures.

10

Biologic damage Biologic damage will be analyzed from the day of implant placement till implant removal. Biologic damage that occurred or was detected after the removal of the implant will be classified under a separate heading.

• No biologic damage (Score 0): No Biologic damage had occurred and no correcting dental procedures were considered necessary. • Reversible biologic damage (Score 1): Biologic damage that is completely reversible with simple dental procedures. This group includes: -- Removal of hyperplastic tissue

-- Fractured mini implant that could be removed without causing irreversible damage. • Irreversible biologic damage (Score 2): Biologic damage that is not completely reversible

with simple dental procedures. This group includes: -- Tooth, nerve, sinus, and blood vessel damage.

-- Fractured mini-implants that could not be removed. -- The need of orthognathic surgery caused by uncontrolled biomechanics with mini-

implants. • Not specified biologic damage (Score NSBD): The presence of biologic damage was described, but the type of damage was not identified. • Post implant biologic damage (Score PIBD): Biologic damage that was caused by treatment with mini implants, but occurred or was detected after the removal of the screw.

Inflammation Inflammation will be measured either within the first month of implant placement or beyond this time limit.

• No inflammation: (Score 0): No signs of inflammation were present during the entire period of treatment with mini-implants. • Temporary inflammation (Score 1): Inflammation was confined to the first month. • Continuing inflammation (Score 2): Inflammation that lasted longer than the first month. • Not specified inflammation (NSI): The duration of inflammation was not specified.

Pain and discomfort Pain and discomfort will be measured within the first two weeks of implantation or beyond38.

• No pain or discomfort (Score 0). No pain or discomfort were present during the entire treatment period with mini-implants. • Moderate pain or discomfort (Score 1). Moderate pain or discomfort within the first

two weeks. • Severe pain or discomfort (Score 2). Severe pain or discomfort within the first two

weeks. • Continuing pain or discomfort (Score 3). Continuing pain lasting longer than two weeks. • Not specified pain (Score NSP): The presence of pain and discomfort was described, but their quality or duration were not specified.

Selection criteria for the analysis of variables influencing success of mini-implants The second part of this review will address variables that might influence success rates of mini-implants. These variables will be classified under the following six headings: implant, patient, location,

11

surgery, orthodontics, and implant maintenance related factors. A correlation between mini-implant success and these parameters will be tested according to the following criteria:

• A proposed association with success will only be rejected when the article presents direct proof that one or more influencing variables were not controlled. Lack of information about the control of those factors will not be considered sufficient to reject a correlation. • Only factors that had been tested for their statistical significance will be included in the analysis of variables.

Search strategy for identification of studies The following electronic data bases will be consulted through March 31 2008: Google Scholar Beta, PubMed, Medline, Embase, Science Direct, all seven Evidence Based Medicine Reviews (EBMR), Web of Science, Ovid, and Bandolier. Librarians who specialized in computerized searches of the health sciences at the American Dental Association will assist with the examination of these databases. The main subject heading “orthodontics” will be combined with the following keywords: implant, screw, mini-implant, mini-screw, micro-implant, screw implant, and temporary anchorage device. The search strategy for PubMed (Medline) will be: (orthodont*) AND (implant* OR mini-implant* OR mini-screw* OR miniscrew* OR micro implant* OR microimplant* OR screw* OR screw implant* OR temporary anchorage device*). For each individual search engine the appropriate characters will be used to truncate or explore search terms. To avoid inappropriate exclusion, the nouns, adjectival, singular and plural forms of all keywords will be inserted. Literature in the English, French, German, and Italian languages will be considered. To analyze if the keywords had covered all articles on mini implants, a manual screening of the following journals will be undertaken: The American Journal of Orthodontics & Dentofacial Orthopedics, The Angle Orthodontist, The European Journal of Orthodontics, The Journal of Orthodontics , The Journal of Clinical Orthodontics, Seminars in Orthodontics, and The International Journal of Adult Orthodontics and Orthognathic surgery. In addition, references found in each identified paper will be manually screened for articles that possibly are missed by the electronic search engines. Finally, all manual and electronic searches will be solicited for review articles35. References found in each review article will also be screened for pertinent papers. This analysis will provide a list of studies on mini implants with their respective success rates.

Study selection All abstracts will be read and the full text of all relevant articles will be collected and reviewed. Ambiguous articles will also be read to avoid inappropriate exclusion. All procedures will be performed independently by two review authors (Reint Reynders and Laura Ronchi). Differences between authors will be resolved by rereading and discussion till consensus will be reached18,35. The selection procedures will be illustrated in a flow diagram. Excluded articles will be listed in a table together with an explanation for their exclusion.

Analysis of methodological quality Studies will also be assessed for eligibility and methodological quality without considering the outcome. For each included study, a value will be given based on the quality of the following four criteria: definition of success, design of the study, description of methodology, and the control of

12

variables. A clear description of each criteria will account for 1 point. Studies will then be classified as clear (3 or 4 points), partially clear (2 points), or unclear (0 or 1 point). Incorporating assessments of study validity will not be used as a threshold for inclusion of studies, but exclusively as a possible explanation for differences in results between studies35.

Data synthesis A decision to perform a meta-analysis will be made in the case of sufficient similarities between studies in the types of participants, interventions and outcomes. Outcomes that will be rated as partially clear or unclear will not qualify for data synthesis35.

13

Data collection forms35 Data collection form 1. Study identification and key characteristics*

Entry Description and character of the information Authors, title, journal of article Source of article Language of the article Type of study Definition of success Success rates Variables that influence success rates Adverse effects Purpose Comments Rating

*Indicate page number and column in article Data collection form 2. Patient characteristics and selection procedures*

Entry Description and character of the information Patient selection criteria Consecutively treated Randomization Informed consent Number of patients Gender and age of patients Number of sites Physical and dental health status Periodontal status Malocclusion Crowding TMD symptoms Additional description of patient characteristics and selection procedures that could be important

Points of interest *Indicate page number and column in article Data collection form 3. Implant characteristics*

Entry Description and character of the information Implant type and company Implant numbers Implant length and diameter Self-drilling or self-tapping design Form of implant (cylindrical or tapered) Design head of implant Design neck of implant Platform of implant Surface of implant Additional description of implant characteristics that could be important

Points of interest *Indicate page number and column in article

14

Data collection form 4. Diagnostic tools* Entry Description and character of the information Pre-surgical radiographs Surgical guide Post-surgical radiographs Points of interest

*Indicate page number and column in article Data collection form 5. Location characteristics*

Entry Description and character of the information Location Location criteria (why chosen) e.g. based on the relationship between the application of the force and the location of the implant or based on the direction of the tooth movement

Edentulous area Description of bone thickness Attached or free mucosa Additional description of location characteristics that could be important

Points of interest *Indicate page number and column in article Data collection form 6. Surgical characteristics*

Entry Description and character of the information Type of insertion technique (flap or flapless) Exposed or closed implants Tissue punch used in mucosa or gingiva Small pilot hole within cortical bone Pilot hole for entire screw length Diameter of pilot hole Diameter pilot bur Self-drilling or self-tapping technique Timing of insertion Angle of insertion Angle of insertion and relationship with the direction of the orthodontic force

Chlorhexidine rinse prior to surgery Drilling speed Saline irrigation/or other cooling methods Mono-cortical or bi-cortical insertion Insertion of screw manually or with handpiece Who conducted the intervention? Orthodontist or surgeon Always same surgeon? Number of implants per site Complications at surgery Additional description of surgical characteristics that could be important


15

Data collection form 6. Orthodontic characteristics* Entry Description and character of the information Time point of loading and definition of time point Direction of loading Magnitude of force of loading at start Increase of magnitude of force during treatment Continuous or intermittent loading Duration of loading Intervals of activation Use of power arms Same orthodontist Type of tooth movement (intrusion, mesialization etc.) En masse or single tooth movement Number of teeth attached to implant Additional description of orthodontic characteristics that could be important

Points of interest *Indicate page number and column in article Data collection form 7. Implant maintenance characteristics*

Entry Description and character of the information Type of home care in general Antibiotics Anti-inflammatory drugs Use of Chlorhexidine rinses and duration Oral hygiene instruction Assessment of mobility Additional description of the implant maintenance characteristics that could be important

Points of interest *Indicate page number and column in article Data collection form 8. Outcomes*

Entry Description and character of the information Definition of success Success rates Osseo-integration Variables that influence success rates Summary of adverse effects Implant fracture (at insertion, during treatment, at removal) Inflammation Pain and discomfort Toleration Easily removed Anesthetic at removal Healing after removal Biologic damage Implant mobility Other complications Additional description of the outcomes that could be important


16

References

1. Roberts WE, Marshall KJ, Mozsary PG. Rigid endosseous implant utilized as anchorage to protract molars and close an atrophic extraction site. Angle Orthod. 1990;60:135-52.

2. Wehrbein H, Merz BR. Aspects of the use of endosseous palatal implants in orthodontic therapy. J Esthet Dent. 1998;10:315-24.

3. Gray JB, Steen ME, King GJ, Clark AE. Studies on the efficacy of implants as orthodontic anchorage. Am J Orthod Dentofacial Orthop. 1983;83:311-7.

4. Roberts WE, Smith RK, Zilberman Y, Mozsary PG, Smith RS. Osseous adaptation to continuous loading or rigid endosseous implants. Am J Orthod Dentofacial Orthop. 1984;86:95-111.

5. Roberts WE, Helm FR, Marshall KJ, Gongloff RK. Rigid endosseous implants for orthodontic and orthopedic anchorage. Angle Orthod. 1989;59:247-256.

6. Ödman J, Lekholm U, Jemt T, Brånemark PI, Thilander B. Osseointegrated titanium implants: a new approach in orthodontic treatment. Eur J Orthod. 1988;10:98-105.

7. Carano A, Melsen B. Implants in orthodontics. Interview. Prog Orthod. 2005;6:62-9. 8. Ohmae M, Saito S, Morohashi T, Seki K, Qu H, Kanomi R, Yamasaki KI, Okano T, Yamada S, Shibasaki Y. A clinical

and histological evaluation of titanium mini-implants as anchors for orthodontic intrusion in the beagle dog. Am J Orthod Dentofacial Orthop. 2001; 119:489-97.

9. Cope JB. Temporary anchorage devices in orthodontics: A paradigm shift. Semin Orthod 2005;11:3-9. 10. Kanomi R. Mini-implant for orthodontic anchorage. J Clin Orthod. 1997;31:763-7. 11. Berens A, Wiechmann D, Rudiger J. L’ancrage intra-osseux en orthodontie à l’aide de mini-et de microvis.

International Orthodontics 2005;3:235-43. 12. Miyawaki S, Koyama I, Inoue M, Mishima K, Sugahara T, Takano-Yamamoto T. Factors associated with the stability

of titanium screws placed in the posterior region for orthodontic anchorage. Am J Orthod Dentofacial Orthop. 2003;124:373-8.

13. Costa A, Raffaini M, Melsen B. Miniscrews as orthodontic anchorage: a preliminary report. Int J Adult Orthodon Orthognath Surg. 1998;13:201-9.

14. Freudenthaler JW, Haas R, Bantleon HP. Bicortical titanium screws for critical orthodontic anchorage in the mandible: a preliminary report on clinical applications. Clin Oral Implants Res. 2001;12:358-63.

15. Fritz U, Ehmer A, Diedrich P. Clinical suitability of titanium miniscrews for orthodontic anchorage-preliminary experiences. J Orofac Orthop. 2004;65:410-8.

16. Heymann GC, Tulloch JF. Implantable devices as orthodontic anchorage: A review of current treatment modalities. J Esthet Restor Dent. 2006;18:68-80.

17. Papadopoulos MA, Tarawneh F. The use of miniscrew implants for temporary anchorage in orthodontics: A comprehensive review. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2007;103:e6-e15.

18. Cornelis MA, Scheffler NR, De Clerck HJ, Tulloch JF, Behets CN. Systematic review of experimental use of temporary skeleletal anchorage devices in orthodontics. Am J Orthod Dentofacial Orthop. 2007;131(4 Suppl): S52-58.

19. Mah J, Bergstrand F. Temporary anchorage devices: a status report. J Clin Orthod. 2005; 39:132-6. 20. Cheng SJ, Tseng IY, Lee JJ, Kok SH. A prospective study of the risk factors associated with failure of mini-implants

used for orthodontic anchorage. Int J Oral Maxillofac Implants. 2004;19:100-6. 21. Liou EJ, Pai BC, Lin JC. Do miniscrews remain stationary under orthodontic forces? Am J Orthod Dentofacial

Orthop. 2004;126:42-7. 22. Motoyoshi M, Hirabayashi M, Uemura M, Shimizu N. Recommended placement torque when tightening an

orthodontic mini-implant. Clin Oral Implants Res. 2006;17:109-14. 23. Thiruvenkatachari B, Pavithranand A, Rajasigamani K, Kyung HM. Comparison and measurement of the amount of

anchorage loss of the molars with and without the use of implant anchorage during canine retraction. Am J Orthod Dentofacial Orthop. 2006;129:551-4

24. Park HS, Jeong SH, Kwon OW. Factors affecting the clinical success of screw implants used as orthodontic anchorage. Am J Orthod Dentofacial Orthop. 2006;130:18-25.

25. Tseng YC, Hsieh CH, Chen CH, Shen YS, Huang IY, Chen CM. The application of mini-implants for orthodontic anchorage. Int J Oral Maxillofac Surg. 2006;35:704-7.

26. Chen CH, Chang CS, Hsieh CH, Tseng YC, Shen YS, Huang IY, Chen CM. The use of microimplants in orthodontic anchorage. J Oral Maxillofac Surg. 2006;64:1209-13.

27. Kravitz ND, Kusnoto B, Tsay TP, Hohlt WF. The use of temporary anchorage devices for molar intrusion. J Am Dent Assoc. 2007;138:56-64.

28. Lin JC, Liou EJ, Yeh CL, Evans CA. A comparative evaluation of current orthodontic miniscrew systems. World J Orthod. 2007;8:136-44.

29. Skeggs RM, Benson PE, Dyer F. Reinforcement of anchorage during orthodontic brace treatment with implants or other surgical methods. Cochrane Database Syst Rev. 2007 Jul 18;(3):CD005098.

30. Wehrbein H, Göllner P. Miniscrews of palatal implants for skeletal anchorage in the maxilla: comparative aspects for decision making. World J Orthod. 2008;9:63-73.

17

31. Arcuri C, Muzzi F, Santini F, Barlattani A, Giancotti A. Five years of experience using palatal mini-implants for orthodontic anchorage. J Oral Maxillofac Surg. 2007;65:2492-7.

32. Prabhu J, Cousley RRJ. Current products and practice. Bone anchorage devices in orthodontics. J Orthod. 2006;33:288-307.

33. Ohashi E, Pecho OE, Moron M, Lagravere MO. Implant vs screw loading protocols in orthodontics. Angle Orthod. 2006;76;721-7.

34. Huan LH, Shotwell JL, Wang HL. Dental implants for orthodontic anchorage. Am J Orthod Dentofacial Orthop. 2005;127:713-22.

35. Higgins JPT, Green S, (editiors). Cochrane Handbook for Systematic Reviews of Interventions 5.0.0. (updated February 2008). The Cochrane Collaboration, 2008. Available from www.cochrane-handbook.org.

36. Turpin DL. CONSORT and QUOROM guidelines for reporting randomized clinical trials and systematic reviews. Am J Orthod Dentofacial Orthop. 2005;128:681-5.

37. Luzi C, Verna C, Melsen B. A prospective clinical investigation of the failure rate of immediately loaded mini-implants used for orthodontic anchorage. Prog Orthod. 2007;8:192-201.

38. Kuroda S, Sugawara Y, Deguchi T, Kyung HM, Takano-Yamamoto T. Clinical use of miniscrew implants as orthodontic anchorage: success rates and postoperative discomfort. Am J Orthod Dentofacial Orthop. 2007;131:9-15.

http://www.cochrane-handbook.org/

18

Protocol 2

Insertion torque and orthodontic mini-implants: protocol for a systematic review of the artificial bone

literature This is the non-edited protocol for the manuscript: Meursinge Reynders R, Ronchi L, Ladu L, Van Etten-Jamaludin F, Bipat S. Insertion torque and orthodontic mini-implants: a systematic review of the artificial bone literature. Proc Inst Mech Eng H. 2013 Nov;227(11):1181-202.

19

Introduction Excessive torque forces applied during the insertion of orthodontic mini-implants (OMIs) have been associated with causing micro-damage and consequently necrosis of the surrounding bone1-4. This condition could ultimately compromise the stability of these devices. To avoid such trauma it is necessary to understand which variables control maximum insertion torque (MIT) values. Insertion torque (IT) is defined as the amount of torque required to overcome the frictional force between the screw and the bone during insertion procedures5. Variables that affect MIT can be divided in: patient, implant, location, and surgery related factors. These variables have been tested in studies on human, animal, cadaver, and artificial bone6-14. Confounding factors such as variability of the insertion procedures can affect outcomes particularly in the former two groups of studies15. Axial load, direction of insertion, stripping of bone, and stability of the operator and subjects are difficult to control in human and animal research. Such studies therefore do not meet the specifications of ASTM International for testing insertion torque of medical bone screws5. Lack of uniformity of human, animal, and cadaver bone can be another confounding factor5,13,15,16. This heterogeneity was confirmed by wide ranges of IT values in these types of bone9-12,15,17,18. The task force of the ASTM International also found both raw and fresh frozen bovine and porcine bone not suitable for comparative testing of bone screws19. On the other hand, the uniformity and consistent properties of artificial bone makes it an ideal material for IT tests19. Research has shown that certain densities of rigid polyurethane foams possess mechanical properties that are in the range of human bone19-21. To reduce the risks of these confounding factors, this systematic review will focus exclusively on variables that influence MIT values in artificial bone. The consistent characteristics of this material also reduce the risk of inter-study heterogeneity, which is important when comparing outcomes. Currently no reviews have been conducted on this topic and a systematic review will therefore be appropriate. The objectives of this broad-spectrum review will be to identify variables associated with maximum insertion torque values during the insertion of OMIs into artificial bone and to quantify such associations. Adverse effects such as implant or bone fractures and cracks will also be assessed.

Methods The methodology of this paper will follow the guidelines of ASTM Standards F543-07ε1 and F1839-08

ε1, The Cochrane Handbook for Systematic Reviews of Interventions, the CONSORT, and PRISMA statements5,19,22-26. A flow diagram for the protocol of this systematic review will be presented in figure 1. ASTM Standards F543-07ε1 and F1839-08 ε1 will be consulted for background information5,19. The first designation provides performance considerations and standard test methods for metallic medical bone screws5. The second document covers rigid unicellular polyurethane foam as a standard material for such tests19. Terminology used in this protocol will be presented in the Appendix.

20

Figure 1. Flow diagram for a protocol of a systematic review25 GRADE: Grading of Recommendations, Assessment, Development and Evaluation

21

Criteria for considering studies for this review Types of studies, test materials, and interventions To address the objectives of this systematic review selection criteria will be defined for the types of studies, test materials, and interventions. Eligibility will not be based on the outcomes of these studies27. Selection criteria include28,29:

• Studies that recorded maximum insertion torque (MIT) values during the insertion of OMIs. • Only laboratory studies that used polyurethane foams as artificial bone will be considered.

These materials were chosen by ASTM international for IT tests because they exhibit similar mechanical properties as human cancellous bone19-21.

• Only studies which used electronic torque sensors for the recordings of IT values will be included.

• OMIs with a diameter smaller than 2.5 mm will be included. This criteria was chosen because larger implants would not classify for specific orthodontic indications (e.g., interradicular positioning)29.

Types of outcome measures This systematic review will address a broad scope review question and multiple outcomes on implant, test block, and insertion procedures related factors will be assessed. Maximum insertion torque will be selected to express these outcomes. MIT is the highest IT value registered during the insertion procedure of OMIs and is presented in Ncm. The following criteria will be set for testing an association between MIT and implant, test block, or insertion procedures related factors:

• Associations will only be considered when based on sample sizes of minimally 5 OMIs. • Associations will be rejected when the article presents direct proof that one or more

confounding factors were not controlled. Lack of information about the possibility of confounding will not be sufficient to reject an association.

• Only associations that presented their p-value will be considered. Adverse effects of the insertion procedure will also be assessed. For this analysis implant fractures and cracks or fractures in the test blocks will be recorded.

Search methods for the identification of studies Our search strategy will aim for high sensitivity and will accept low precision30. To find eligible studies the following electronic data bases will be searched through February 24 2012: Google Scholar Beta, Pubmed (Medline), Embase (Ovid), CENTRAL, Science Direct, Scopus, Web of Science, LILACS, and AJOL30. Grey literature databases will also be consulted which include: OpenGrey, HMIC, NTIS. A librarian (FVEJ), who is specialized in computerized searches of health science publications at the Academic Medical Center of the University of Amsterdam will be consulted for the examination of these databases. Search strategies will be customized for each database and will always be copied and pasted and never re-typed as this can introduce errors30 (See Appendix). To avoid inappropriate exclusion a wide variety of search terms will be combined, which include the following subject headings and keywords: orthodontics, torque, implant, mini implant, micro implant, microimplant, screw, mini screw, miniscrew, micro screw, microscrew, and temporary anchorage device. For each individual search engine the appropriate characters will be used to truncate or explore search terms. Nouns, adjectival, singular and plural forms of all keywords will be inserted30. To limit inadvertent exclusion of eligible articles we will not apply search filters and the Boolean NOT operator. As an adjunct to searching electronic databases, manual searches will also be performed in the following journals: The American Journal of Orthodontics & Dentofacial Orthopedics, The Angle

22

Orthodontist, The European Journal of Orthodontics, The Journal of Orthodontics , The Journal of Clinical Orthodontics, Seminars in Orthodontics, World Journal of Orthodontics, and The International Journal of Adult Orthodontics and Orthognathic surgery. References of review articles that will be identified by these search methods will be screened for pertinent papers30. In addition, references of each eligible paper will also be screened for articles that will not be found by the electronic search engines and manual searches. Risk of publication bias will be further controlled, because no language restrictions will be applied in our search strategy.

Selecting studies To improve precision and reduce the risk of introducing publication bias, the selection of studies will be performed independently by three authors (RMR, LR, LL)31,32. These investigators are experts on this topic and experienced systematic reviewers. The quality of the methodology and the statistics of this paper will be guaranteed by the fifth author (SB). Assessments of inter-examiner agreement will be conducted according to the protocol presented in the Cochrane Handbook for Systematic Reviews of Interventions, which is described under here31. To reduce the chances of missing eligible studies, pilot tests will be performed on samples of reports prior to conducting the search process. This preliminary testing will be done to improve the quality of the eligibility criteria, to guarantee that these criteria could be applied consistently, and to improve inter-examiner agreement31. The selection process will then be conducted according to the following protocol: (1) all titles and abstracts will be examined to remove obviously irrelevant reports; (2) the full text of potentially relevant papers will be retrieved. To avoid inappropriate exclusion, ambiguous articles will also be included; 3) multiple reports of the same study will be linked to avoid biases through duplicate publications; (4) all selected studies will be subsequently reviewed and examined for compliance with the selection criteria; (5) investigators will be contacted for clarification in the case of doubt concerning appropriate eligibility of studies; (6) Then, final decisions will be made on eligibility31,33. Full agreement on the inclusion of studies will be recorded, indicating a kappa of 1.0. Selection procedures will be presented in a flow diagram and reasons for exclusion of all full text articles will be explained.

Data collection Checklists of items to consider for the data collection of the selected studies will be prepared. They include detailed information about the source, eligibility, methods, artificial bone, interventions, confounding factors, outcomes, results, and miscellaneous variables that could affect outcomes31. These collection forms will be first pilot tested for their validity and modifications will be made to improve their quality. Then data of each individual study will be extracted and transferred to “characteristics of studies and findings” tables31. Data extraction procedures will also be first pilot tested. Subgroup analyses will be conducted when a study provides sufficient data to perform such assessments. To minimize errors and reduce potential biases all data extraction procedures will be done independently by three authors (Authors RR, LR, LL). Disagreements between authors will be resolved by rereading and discussion till full consensus will be reached. Data collection forms are presented in the appendix of this protocol.

Assessment of risk of bias of outcomes The validity of the selected studies will be scored through an analysis of the risk of bias in the results. Judgements will be made on the magnitude, reliability, and the direction of bias e.g. the extent to

23

which systematic error could have overestimated or underestimated the true effect of the intervention34. The following protocol for the risk of bias assessment will be applied: (1) specific criteria will be set for the analysis of articles. These criteria will be based on ASTM Standards F543-07ε1 and F1839-08 ε1; (2) each article will be analysed according to these guidelines and findings will be listed in tables; (3) these outcomes will then be assessed for their risk of bias5,19. The following types of systematic error will be scored: performance, detection, attrition, reporting, and publication bias34. Performance bias will be divided in four subgroups: implant, test block, insertion procedures, and operator related bias. Judgments will be made for each entry and will be rated as: ‘low’, ’high’, or ‘unclear’ risk of bias. A domain will only be scored as ‘high’ risk when the article presented direct evidence that an entry was at risk of confounding. Lack of information or uncertainty over potential risk of systematic error will be rated as ‘unclear’34. Judgments of bias will not be used as eligibility criteria for the inclusion of studies, but will be applied solely to explain possible differences in the results. Tables in the appendix further elaborate on the risk of bias assessments. Assessment of the quality of evidence The quality of evidence to the extent that one can be confident that proposed associations between specific variables and MIT can be trusted will be assessed according to the GRADE approach35. Judgments on quality will be graded as ‘high’, ‘moderate’, ‘low’, or ‘very low’. These assessments will involve considerations of the risk of bias analysis, directness of evidence, heterogeneity, precision of effect estimates, and risk of publication bias36. Quality ratings will be made for each proposed association with MIT36. Only associations of high or moderate quality will be considered.

Measures of treatment effect In the case that treatment effects will be summarized in a meta-analysis, these measures will be selected post-hoc, e.g., risk ratios and odds ratios etc.37.

Data synthesis A meta-analysis will be performed in the case of: (1) low risk of bias in the individual studies; (2) consistent direction of outcomes between eligible studies; (3) low publication and/or reporting bias; (4) a high number of selected studies; (5) homogeneity between studies in the types of participants, interventions, outcomes etc.34,37,38. Funnel plots will only be displayed when more than 10 studies will present an association between a particular variable and MIT. For fewer studies the power of this test is too low to distinguish chance from real asymmetry33.

Contacting authors Emails will be sent to the authors of the eligible studies when additional information on their research study will be necessary. To avoid the risk of “email discrimination”, because certain email addresses could have been identified as “spam mail” by the receiving internet provider, we will also sent letters from different email addresses.

24

Differences between the protocol and the review All differences between the protocol and the final systematic review will be presented together with a rationale for these changes and their potential for introducing bias39.

25

Appendix Definitions of terminology

Terminology in the orthodontic literature

Terminology according to the ASTM5

Definition

Screw head Screw head The end of the screw which is opposite of the tip and from which the means of inserting the screw is coupled5. Screw heads of orthodontic screws frequently contain an orthodontic attachment.

Screw neck Screw shank Unthreaded portion of the screw excluding the head5. Not all orthodontic mini-implants have necks.

Thread runout Thread runout The intersection of the screw thread with either the screw shaft or screw head5.

Screw tip Screw tip The end (apex) of the screw which is opposite of the head Screw length The overall length of the screw measured from the screw head to

the screw tip5. Screw length Thread length

The length of the threaded portion of the screw, measured from the thread runout to the screw tip5.

Core (inner diameter) Core diameter (minor diameter)

The smallest diameter of the threaded portion of the screw measured at the thread root5.

Diameter (outer diameter)

Thread diameter (major diameter)

The largest diameter of the threaded portion of the screw measured over the thread crests5.

Pitch Pitch The length between the thread crests5. Flute A vertical groove in the threads of the screw that facilitates clearing

of bone debris40. Self-drilling screw Self-tapping screw A screw that has any number of flutes at its tip which are intended

to cut the screw’s thread form into the bone upon insertion5. When turned, these screws create their own thread.

Pre-drilling screw Non-tapping screw A screw with a tip that does not contain a flute5. When turned, these screws do not create their own thread and require predrilling.

Cylindrical screw The diameter of the core remain constant over the core length. Conical screw The diameter of the core increase from the tip towards the head of

the screw. Pilot hole Pilot hole The hole drilled into the bone into which the screw tip is inserted5. Starter pilot hole Superficial pilot hole. Full length pilot hole Pilot hole for the entire screw length. Self-drilling insertion technique

Technique for the insertion of the screw that does not require a full length pilot hole

Pre-drilling insertion technique

Technique for the insertion of the screw that requires a pilot hole for the entire screw length. In the orthodontic literature this technique is frequently defined as the self-tapping technique.

Axial force (N) Self-tapping force and Axial force (N)

ASTM5 defines this item as ‘The amount of axial force required to engage the self-drilling features of self-drilling style screws’. In this article this definition will be applied for both self-drilling and pre-drilling screws.

Insertion speed Test speed The revolutions per minute (rpm) that are applied by the motor-driven torque wrench to the head of the screw5.

Insertion depth Insertion depth The entire length of the threaded portion (length) of the screw as inserted into the test block5.

Insertion torque

Insertion torque The amount of torque required to overcome the frictional force between the screw and the material used for testing while driving the screw into the material5.

Maximum insertion torque

The maximum torque value recorded from the beginning to the end of the insertion process of OMIs15. MIT is expressed in Newton centimetre (Ncm).

Angular momentum Insertion torque values are plotted against time and the angular momentum is represented by the steepness of this insertion torque curve15. The angular momentum is expressed in Newton centimetre seconds (Ncms).

26

Search strategy: Subject headings and keywords: orthodontics, torque, implant, mini implant, micro implant, microimplant, screw, mini screw, miniscrew, micro screw, microscrew, and temporary anchorage device $

Source of records Search strategy Google Scholar Beta (orthodontics OR orthodontic OR orthodontist OR orthodontists) (success OR successes OR

stability OR stabilities) (implant OR implants OR “mini implant” OR “mini implants” OR “micro implant” OR “micro implants” OR microimplant OR microimplants OR screw OR screws OR “mini screw” OR “mini screws” OR miniscrew OR miniscrews OR “micro screw” OR “micro screws” OR microscrew OR microscrews OR “temporary anchorage device” OR “temporary anchorage devices”)

PubMed (Medline) (orthodont*) AND (torque*) AND (implant* OR mini implant* OR micro implant* OR microimplant* OR screw* OR mini screw* OR miniscrew* OR micro screw* OR microscrew* OR temporary anchorage device*)

Embase (Ovid) (orthodont*) AND (torque*) AND (implant* OR mini implant* OR micro implant* OR microimplant* OR screw* OR mini screw* OR miniscrew* OR micro screw* OR microscrew* OR temporary anchorage device*)

CENTRAL Orthodont* Science Direct (orthodont*) AND (torque*) AND (implant* OR mini implant* OR micro implant* OR

microimplant* OR screw* OR mini screw* OR miniscrew* OR micro screw* OR microscrew* OR temporary anchorage device*)

Scopus (orthodont*) AND (torque*) AND (implant* OR mini implant* OR micro implant* OR microimplant* OR screw* OR mini screw* OR miniscrew* OR micro screw* OR microscrew* OR temporary anchorage device*)

Web of Science (orthodont*) AND (torque*) AND (implant* OR mini implant* OR micro implant* OR microimplant* OR screw* OR mini screw* OR miniscrew* OR micro screw* OR microscrew* OR temporary anchorage device*)

LILACS (orthodont*) AND (torque*) AND (implant* OR mini implant* OR micro implant* OR microimplant* OR screw* OR mini screw* OR miniscrew* OR micro screw* OR microscrew* OR temporary anchorage device*)

AJOL orthodontics , orthodontic , orthodontist, orthodontists were all used as single keywords Grey literature orthodontics , orthodontic , orthodontist, orthodontists were all used as single keywords Hand searching The American Journal of Orthodontics & Dentofacial Orthopedics, The Angle Orthodontist,

The European Journal of Orthodontics, The Journal of Orthodontics , The Journal of Clinical Orthodontics, Seminars in Orthodontics, World Journal of Orthodontics, and The International Journal of Adult Orthodontics and Orthognathic surgery.

References of eligible studies and review articles

References will be retrieved from the respective journals

$ For each individual search engine the appropriate characters were used to truncate or explore search terms. All search strategies were copied and pasted from the original without re-typing as this can introduce errors30.

27

Data collection forms Data collection form 1. Source and eligibility31,41*

Entry Description and character of the information Authors, title, journal of article Source of article Language of the article Name of reviewer Trial register Eligible/not eligible Purpose Comments

*Locate where information on each item can be found, e.g., Page 12 column 3. Data collection form 2. Implant related factors*

Entry Description and character of the information Number of implants Implant type (company) Implant characteristics Implant material Diameter (D) Length (L)

Core diameter Surface finish Sterilized Design (self-drilling, pre-drilling) Core form (cylindrical or conical) Thread characteristics Pitch dimensions Flute in tip Flute in core

*Locate where information on each item can be found, e.g., Page 12 column 3. Data collection form 3. Test block related factors*

Entry Description and character of the information Type of bone Trade name Composition Density Tensile strength Compression strength Shear strength Characteristics of test block Thickness of test block Dimensions of the test block Cortical bone thickness

*Locate where information on each item can be found, e.g., Page 12 column 3.

28

Data collection form 4. Test apparatus related factors* Entry Description and character of the information Testing fixture Should not deflect or deform under loading5.

Torquing device and recorder Able to continuous recordings of torque values and calibrated to measure the values encountered during the test5.

Bushing Should be incorporated in the design of the IT apparatus and worn bushings shall be discarded and replaced5.

Test block clamp Should be incorporated in the design of the IT apparatus and should not deform the test block5.

*Locate where information on each item can be found, e.g., Page 12 column 3. Data collection form 5. Surgery related factors*

Entry Description and character of the information Insertion technique (self-drilling or self-tapping)

Pilot holes for predrilling technique (diameter in mm and depth)

Pre-drilling device for pre-drilling pilot holes Pre-drilling speed for pre-drilling pilot holes Quality of the pre-drilling pilot hole drill Pilot holes for self-drilling technique (diameter in mm and depth)

Predrilling device for self-drilling pilot holes Predrilling speed for self-drilling pilot holes Quality of the self-drilling pilot hole drill Distance between screws (mm) Insertion torquing device (for example a hand-driven torquing ratchet)

Insertion force or load (kg) Insertion angulation (degrees) Insertion speed (rpm) Insertion time (seconds) Number of rotations during insertion process Insertion depth (mm) Maxium insertion torque (Ncm) Insertion torque registered during entire screwing process ? Insertion angular momentum (increase in torque per second (Ncms)

Friction Bone and head of screw ? Stripping or stalling

*Locate where information on each item can be found, e.g., Page 12 column 3. Data collection form 6. Outcomes*

Entry Description and character of the information Maximum insertion torque (MIT) (Ncm) Association with MIT Adverse effects Comments


29

Classification scheme for the assessment of risk of bias31,34

Type of bias Description of bias Performance bias Refers to systematic error in the care that is provided, or in exposure to factors other than the

interventions of interest that can affect the results of a research study. Performance bias covers the period from the start of the intervention and finishes at the start of measuring outcomes.

Performance bias implants

Refers to systematic error in implant related factors that can affect the results of a research study.

Performance bias test block

Refers to systematic error in test block related factors that can affect the results of a research study.

Performance bias insertion procedures

Refers to systematic error in insertion procedures related factors that can affect the results of a research study.

Performance bias operator

Refers to systematic error in the blinding of the operator that can affect the results of a research study.

Detection bias Refers to systematic error in the assessment and analysis of outcomes that can affect the results of a research study. Detection bias covers the period from the start of measuring outcomes and finishes at the completion of the data analysis.

Attrition bias Refers to systematic error in dealing with incomplete outcome data that can affect the results of a research study. Attrition bias covers the period from the start of randomization and finishes at the completion of the data analysis.

Reporting bias Refers to selective reporting of some or all of the outcomes that can affect the results of a research study. Reporting bias covers the period of writing the report.

Publication bias Refers to the selective publication of a research study. Publication bias covers the period after the completion of the report of a research study and finishes with the publication or non-publication of the article.

Assessment of risk of bias Performance bias mini-implant

Entry Description and character of the information Risk of bias score* Screw identification Could the type, characteristics, and dimensions of the screw

related variables cause bias?

Screw chemical composition Could chemical composition related variables cause bias? Screw Surface finish Could surface finish related variables cause bias? Mini-implant comparisons/confounding

Could uncontrolled variables cause bias?

Overall risk of bias score Overall risk of bias score based on the various mini-implant related entries

Key rationale for bias score Explain the key rationale for the risk of bias score *Magnitude of judgment on risk of bias: ‘high’, ‘low’ or ‘unclear’. Unclear risk of bias: The Cochrane Collaboration assigns ‘unclear’ risk of bias if: (1) insufficient detail is reported of what happened in the study; or (2) what happened in the study is known, but the risk of bias is unknown; or (3) an entry is not relevant to the study at hand34.

30

Performance bias test block Entry Description and character of the information Risk of bias score* Test block material description Could the type, and characteristics of the test block material

cause bias?

Test block dimensions Could the dimensions of the test block cause bias? The smallest dimensions of the testblock shall be 10 x the diameter under test5.

Test block chemical composition

Could the chemical composition of the test block cause bias?

Test block physical characteristics

Could the physical characteristics of the test block cause bias?

Test block quality Could the test block quality, e.g., homogeneity, cause bias? Overall risk of bias score Overall risk of bias score based on the various test block

related entries

Key rationale for bias score Explain the key rationale for the risk of bias score *Magnitude of judgment on risk of bias: ‘high’, ‘low’ or ‘unclear’. Unclear risk of bias: The Cochrane Collaboration assigns ‘unclear’ risk of bias if: (1) insufficient detail is reported of what happened in the study; or (2) what happened in the study is known, but the risk of bias is unknown; or (3) an entry is not relevant to the study at hand34. Performance bias insertion procedures

Entry Description and character of the information Risk of bias score* Testing fixture Could the characteristics of the texting fixture cause bias?

Note: the testing fixture should not deflect or deform under loading5.

Torqueing device and recorder Could the characteristics of the torqueing device and recorder cause bias? Note: Assess whether the torqueing device and recorder were able to continuous recordings of torque values and calibrated to measure the values encountered during the test5.

Bushing Could the characteristics of the bushing cause bias? Note: Bushing should be incorporated in the design of the IT apparatus and worn bushings shall be discarded and replaced5.

Test block clamp Could the characteristics of the test block clamp cause bias? Note: The test block clamp should be incorporated in the design of the IT apparatus and should not deform the test block5.

Maximum insertion torque (MIT)

Was MIT recorded? Note: Not recording MIT could cause bias.

Continuous recording of insertion torque (IT)

Was IT recorded continuously? Note: Not recording IT continuously could cause bias.

Axial load Was the axial load described? Note: Axial loads higher than 1.14 kgf could cause bias.

Insertion depth Was the insertion depth recorded? Specifications on pilot hole Were specifications on pilot hole drilling described? Spacing between screws Was the distance between screws described? Note: The

distance between screws should be at a minimum distance of 5 x the diameter of the screw.

Insertion speed Was the insertion speed described? Friction between head of the screw and bone?

Was friction between the head of the screw and bone described? Note: Friction between the heads of the screw and bone should be avoided.

Overall risk of bias score Overall risk of bias score based on the various insertion procedures related entries


31

Performance bias operator Entry Description and character of the information Risk of bias score* Blinding Could blinding of the operator issues cause bias? Other operator issues Could other operator related issues cause bias? Overall risk of bias score Overall risk of bias score based on the various operator

related entries

Key rationale for bias score Explain the key rationale for the risk of bias score *Magnitude of judgment on risk of bias: ‘high’, ‘low’ or ‘unclear’. Unclear risk of bias: The Cochrane Collaboration assigns ‘unclear’ risk of bias if: (1) insufficient detail is reported of what happened in the study; or (2) what happened in the study is known, but the risk of bias is unknown; or (3) an entry is not relevant to the study at hand34. Assessment of other risks of bias

Type of bias Description and character of the information Risk of bias score* Detection bias Refers to systematic error in the assessment and analysis of

outcomes that can affect the results of a research study. Detection bias covers the period from the start of measuring outcomes and finishes at the completion of the data analysis.

Key rationale for bias score Explain the key rationale for the risk of bias score Attrition bias Refers to systematic error in dealing with incomplete

outcome data that can affect the results of a research study. Attrition bias covers the period from the start of randomization and finishes at the completion of the data analysis.

Key rationale for bias score Explain the key rationale for the risk of bias score Reporting bias Refers to selective reporting of some or all of the outcomes

that can affect the results of a research study. Reporting bias covers the period of writing the report.

Key rationale for bias score Explain the key rationale for the risk of bias score Publication bias Refers to the selective publication of a research study.

Publication bias covers the period after the completion of the report of a research study and finishes with the publication or non-publication of the article.


32

References

1. Motoyoshi M, Hirabayashi M, Uemura M, Shimizu N. Recommended placement torque when tightening an orthodontic mini-implant. Clin Oral Implants Res. 2006;17:109-14.

2. Lee NK, Baek SH. Effects of the diameter and shape of orthodontic mini-implants on microdamage to the cortical bone. Am J Orthod Dentofacial Orthop. 2010;138:e1-8.

3. Wawrzinek C, Sommer T, Fischer-Brandies H. Microdamage in cortical bone due to the overtightening of orthodontic microscrews. J Orofac Orthop. 2008;69:121-34.

4. Togni F, Baras F, Ribas Mde O, Taha OM. Histomorphometric analysis of bone tissue repair in rabbits after insertion of titanium screws under different torque. Acta Cir Bras. 2011;26:261-6

5. ASTM Standard F543-07ε1, Standard specification and test methods for metallic medical bone screws. ASTM International, West Conshohocken, PA, USA, 2007 (Editorially corrected August 2009), www.astm.org.

6. Motoyoshi M, Yoshida T, Ono A, Shimizu N. Effect of cortical bone thickness and implant placement torque on stability of orthodontic mini-implants. Int J Oral Maxillofac Implants. 2007;22:779-84.

7. Motoyoshi M, Uemura M, Ono A, Okazaki K, Shigeeda T, Shimizu N. Factors affecting the long-term stability of orthodontic mini-implants. Am J Orthod Dentofacial Orthop. 2010;137:588.e1-5.

8. Salmória KK, Tanaka OM, Guariza-Filho O, Camargo ES, de Souza LT, Maruo H. Insertional torque and axial pull-out strength of mini-implants in mandibles of dogs. Am J Orthod Dentofacial Orthop. 2008;133:790.e15-22.

9. Kim SH, Lee SJ, Cho IS, Kim SK, Kim TW. Rotational resistance of surface-treated mini-implants. Angle Orthod. 2009;79:899-907.

10. Wilmes B, Rademacher C, Olthoff G, Drescher D. Parameters affecting primary stability of orthodontic mini-implants. J Orofac Orthop. 2006;67:162-74.

11. Wilmes B, Ottenstreuer S, Su Y-Y, Drescher D. Impact of implant design on primary stability of orthodontic mini-implants. J Orofac Orthop. 2008;69:42-50.

12. Wilmes B, Drescher D. Impact of insertion depth and predrilling diameter on primary stability of orthodontic mini-implants. Angle Orthod. 2009;79:609-14.

13. Song YY, Cha JY, Hwang CJ. Mechanical characteristics of various orthodontic mini-screws in relation to artificial cortical bone thickness. Angle Orthod. 2007;77:979-85.

14. Lim SA, Cha JY, Hwang CJ. Insertion torque of orthodontic miniscrews according to changes in shape, diameter and length. Angle Orthod. 2008;78:234-40.

15. Meursinge Reynders RA, Ronchi L, Ladu L, van Etten-Jamaludin F, Bipat S. Insertion torque and success of orthodontic mini-implants: A systematic review. Am J Orthod Dentofacial Orthop. 2012;142:596-614.e5.

16. Brinley CL, Behrents R, Kim KB, Condoor S, Kyung HM, Buschang PH. Pitch and longitudinal fluting effects on the primary stability of miniscrew implants. Angle Orthodontist. 2009;79:1156-1161.

17. Arismendi JA, Ocampo ZM, Morales M, Gonzalez FJ, Jaramillo PM, Sanchez A. Evaluation of stability of mini implants as bony anchorage for upper molar intrusion. Revista Facultad de Odontologia Universidad de Antioquia 2007;19:59-73.

18. Suzuki EY, Suzuki B. Placement and removal torque values of orthodontic miniscrew implants. Am J Orthod Dentofacial Orthop. 2011;139:669-78.

19. ASTM Standard F1839-08 ε1, Standard specification for rigid polyurethane foam for use as a standard material for testing orthopaedic devices and instruments. ASTM International, West Conshohocken, PA, USA. 2008. (Editorially corrected August 2009). www.astm.org.

20. Szivek JA, Thomas M, Benjamin JB. Technical Note - Characterization of a synthetic foam as a model for human cancellous bone. J Appl Biomater. 1993;4:269-272.

21. Szivek JA, Thomas M, Benjamin JB. Characterization of three formulations of a synthetic foam as models for a range of human cancellous bone types. J Appl Biomater. 1995;6:125-128.

22. Higgins JPT, Green S, (editiors). Cochrane Handbook for Systematic Reviews of Interventions. Chichester (UK): John Wiley & Sons, 2008.

23. Turpin DL. Updated CONSORT and PRISMA documents now available. Am J Orthod Dentofacial Orthop. 2010;137:721-2.

24. CONSORT 2010. www.consort-statement.org 25. PRISMA 2009. www.prisma-statement.org 26. Moher D, Liberati A, Tetzlaff J, Altman DG; PRISMA Group. Preferred reporting items for systematic reviews and

meta-analyses: the PRISMA statement. PLoS Med. 2009 Jul 21;6(7):e1000097. 27. O’Connor D, Green S, Higgins JPT. Chapter 5: Defining the review question and developing criteria for including

studies. In: Higgins JPT, Green S, (editiors). Cochrane Handbook for Systematic Reviews of Interventions. Chichester (UK): John Wiley & Sons, 2008.

28. Chen Y, Kyung HM, Zhao WT, Yu WJ. Critical factors for the success of orthodontic mini-implants: a systematic review. Am J Orthod Dentofacial Orthop. 2009;135:284-91.

29. Reynders R, Ronchi L, Bipat S. Mini-implants in orthodontics: as systematic review of the literature. Am J Orthod Dentofacial Orthop. 2009;135:564.e1-19.

http://www.ncbi.nlm.nih.gov/pubmed/18538236?ordinalpos=91&itool=EntrezSystem2.PEntrez.Pubmed.Pubmed_ResultsPanel.Pubmed_DefaultReportPanel.Pubmed_RVDocSum








http://www.astm.org/

http://www.consort-statement.org/

http://www.prisma-statement.org/



33

30. Lefebvre C, Manheimer E, Glanville J. Chapter 6: Searching for studies. In: Higgins JPT, Green S, (editors). Cochrane Handbook for Systematic Reviews of Interventions. Chichester (UK): John Wiley & Sons, 2008.

31. Higgins JPT, Deeks JJ. Chapter 7: Selecting studies and collecting data. In: Higgins JPT, Green S, (editors). Cochrane Handbook for Systematic Reviews of Interventions. Chichester (UK): John Wiley & Sons, 2008.

32. Borenstein M, Hedges LV, Higgins JPT, Rothstein HR. Chapter 30: Publication bias. In: Borenstein M, Hedges LV, Higgins JPT, Rothstein HR, (editors). Introduction to Meta-Analysis. Chichester (UK): John Wiley & Sons, 2009.

33. Sterne JAC, Egger M, Moher D. Chapter 10: Addressing reporting biases. In: Higgins JPT, Green S, (editors). Cochrane Handbook for Systematic Reviews of Interventions. Chichester (UK): John Wiley & Sons, 2008.

34. Higgins JPT, Altman DG (editors). Chapter 8: Assessing risk of bias in included studies. In: Higgins JPT, Green S, (editiors). Cochrane Handbook for Systematic Reviews of Interventions. Chichester (UK): John Wiley & Sons, 2008.

35. GRADE Working Group 2004. GRADE Working Group. Grading quality of evidence and strength of recommendations. BMJ. 2004;328:1490-4.

36. Schünemann HJ, Oxman AD, Vist GE, Higgins JPT, Deeks JJ, Glasziou P, Guyatt GH. Chapter 12: Interpreting results and drawing conclusions. In: Higgins JPT, Green S (editors), Cochrane Handbook for Systematic Reviews of Interventions. Chichester (UK): John Wiley & Sons, 2008.

37. Deeks JJ, Higgins JPT, Altman DG. Chapter 9: Analysing data and undertaking meta-analyses. In: Higgins JPT, Green S, (editiors). Cochrane Handbook for Systematic Reviews of Interventions. Chichester (UK): John Wiley & Sons, 2008.

38. Borenstein M, Hedges LV, Higgins JPT, Rothstein HR. Chapter 40: When does it make sense to perform a meta-analysis? In: Borenstein M, Hedges LV, Higgins JPT, Rothstein HR, (editors). Introduction to Meta-Analysis. Chichester (UK): John Wiley & Sons, 2009.

39. Higgins JPT, Green S. Chapter 4: Guide to the contents of a Cochrane protocol and review. In: Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions. Chichester (UK): John Wiley & Sons, 2008.

40. Mazzocca AD, Caputo AE, Browner BD, Mast JW, Mendes MW. Principles of internal fixation. In Browner BD, Levine AM, Jupiter JB, Trafton PG (Eds). Skeletal Trauma (3rd edition). Philadelphia, Saunders 2003;195-204.

41. Reeves BC, Deeks JJ, Higgins JPT, Wells GA . Chapter 13: Including non-randomized studies. In: Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions. Chichester (UK): John Wiley & Sons, 2008.

34

Protocol 3

Insertion torque and success of orthodontic mini-implants: protocol for a systematic review

This is the non-edited protocol for the manuscript: Meursinge Reynders RA, Ronchi L, Ladu L, van Etten-Jamaludin F, Bipat S. Insertion torque and success of orthodontic mini-implants: a systematic review. Am J Orthod Dentofacial Orthop. 2012;142(5):596-614.e5.

35

Abstract Background: Orthodontic mini implants (OMIs) are used as stationary anchorage during orthodontic treatment, but these devices can loosen, become mobile or even migrate. Excessive torque forces applied during the insertion of these devices have been associated with instability and failure of these devices. Reports on ‘safe’ maximum insertion torque (MIT) values have been controversial. Objectives: To evaluate the effects of specific (MIT) values on the success of OMIs. Search strategy: We will search till February 24 2012 without language restrictions: Google Scholar Beta, Pubmed (Medline), Embase (Ovid), CENTRAL, Science Direct, Scopus, Web of Science, LILACS, and AJOL, and Grey Literature. We will conduct manual searches of topic specific journals, check reference lists, and contact researchers and organizations in the field. Selection criteria: We will select randomized controlled trials (RCT) and non-randomized comparative studies (NRS) that compare success rates of OMIs inserted with a MIT of 5-10 Ncm with those inserted with MIT values beyond this range. Data collection and analysis: Three review authors will independently select eligible trials, extract data, and assess methodological quality. This study will measure Risk (RR) and Odds ratios (OR) with 95% confidence intervals (CI) and when appropriate undertake a meta-analysis. Desired characteristics of studies 1) to find low publication bias; 2) identify a high number of large RCT of high quality, high precision, and a large effect size; 3) find low heterogeneity and a consistent direction of effects of interventions between studies. Desired outcomes 1) MIT of 5-10 Ncm is associated with higher success rates of OMIs compared to OMIs inserted with a MIT beyond this range; 2) no adverse effects. Impact on the health service and future treatment of patients Less suffering for patients, lower costs, more efficient treatment, more reliable treatment outcomes, shorter treatment times, less adverse effects.

Plain Language Summary Insertion torque values for stable orthodontic implants Background and rationale Most orthodontic treatments require anchorage to prevent unwanted tooth movement. OMIs have been introduced as promising stationary anchorage devices, but these devices can loosen, become mobile or even migrate. Consequences of these adverse outcomes include: reoperation, increased costs, and risk of trauma to nerves, dental roots and other vital structures. Failure rates of OMIs vary around 20%. It has been suggested that excessive torque forces applied during the insertion of these devices, can cause necrosis of the surrounding bone thereby compromising their success. It is therefore necessary to understand at what levels torque strains remain physiologic and can guarantee the stability of these implants. Several clinical studies have associated specific torque levels (5-10 Ncm) with higher success rates, but conflicting outcomes have also been reported. Because of this disagreement and because currently no reviews have addressed this issue, a systematic review (SR) was deemed appropriate. Desired outcomes and impact on the health service and future treatment of patients In this SR we hope to find: 1) a high number of reliable studies with large sample sizes and similar designs; 2) studies that present a strong association between a specific MIT value and success of OMIs; 3) no adverse effects. Lower failure rates of OMIs could lead to reduced suffering of patients, lower costs, more efficient orthodontic treatment, more reliable treatment outcomes, shorter treatment times, and less adverse effects.

36

Flow diagram of the protocol The protocol for undertaking a systematic review is presented in a flow diagram (Figure 1) and is based on The Cochrane Handbook for Systematic Reviews of Interventions and the PRISMA statement1-3.

Figure 1: Protocol for a systematic review2-6

STEP A: Background and objectives Step 1) Background and rationale for undertaking a systematic review for the PICO question. Step 2) Objectives (primary outcomes and at least one adverse effect).

STEP B: Finding eligible studies Step 1) Criteria for considering studies for this review:

a) Study characteristics (types of studies, participants, interventions and duration, and outcome measures).

b) Report characteristics (language of publications, publication status, and year of publication).

Step 2) Information sources (electronic databases, grey literature, hand searching, reference searching, corresponding with researchers etc.). Step 3) Search (development of the search strategy). Step 4) Study selection (strategy for study selection, e.g. pilot testing of the selection process and independent selection, in duplicate, and obtaining information from researchers). Step 5) Create a flow diagram for searching studies. Step 6) Present eligible and excluded studies and list reasons for exclusion.

STEP C: Data collection Step 1) Data collection process (strategy for data collection, e.g. pilot testing of data collection forms, independent extraction, in duplicate, and obtaining information from researchers). Step 2) Data items (source, eligibility, design, participants, interventions, outcomes, duration, results, and miscellaneous data).

37

STEP D: Data analysis Step 1) Dealing with risk of bias in individual studies (strategy for the assessment of bias within studies and across outcomes). Step 2) Summary measures (principal summary measures, e.g. risk ratio with confidence intervals for each outcome). Step 3) Dealing with unit of analysis issues (assess unit of analysis issues according to the level at which randomization occurred). Step 4) Dealing with missing data (strategies for dealing with missing participants, statistics etc.). Step 5) Dealing with heterogeneity (identifying sources of heterogeneity and strategies for dealing with it). Step 6) Dealing with publication bias (identifying sources of publication bias and strategies for dealing with it).

STEP G: Data synthesis Step 1: Synthesis of results (strategy for synthesizing the results, including criteria for undertaking a meta-analysis, type of meta-analysis, and measures of consistency). Step 2: Subgroup analysis (list all planned subgroup or meta-regression analyses). Step 3: Sensitivity analysis (list all planned sensitivity analyses).

STEP F: Produce a summary of findings table for each primary study

STEP E: Assessment of the quality (GRADE) for each body of evidence Step 1) List and define parameters that can affect the GRADE rating, e.g. the design of a study, risk of bias, directness of evidence, precision of results, heterogeneity, unit of analysis issues, and reporting bias. Step 2) Define rating parameters for the GRADE assessment (high, moderate, low, or very low). Step 3) ‘Initial’ quality assessment (GRADE). Step 4) Assess factors that may increase or decrease the level of quality. Step 5) ‘Final’ quality assessment (GRADE).

STEP H: Presenting the results, discussion and conclusions

38

Step A: Background and objectives Description of the condition Many variables that affect the stability of orthodontic mini-implants (OMIs) are still poorly understood7-12. It has been suggested that excessive torque forces applied during the insertion of these devices, can cause necrosis of the surrounding bone and compromise their success11,13-15. It is therefore necessary to understand at what levels torque strains remain physiologic and can guarantee the stability of these implants. Most orthodontic treatments require anchorage to prevent unwanted tooth movement. OMIs have been introduced as promising stationary anchorage devices, but their outcomes are not always consistent7,16. They can loosen, become mobile or even migrate10,17,18. The six principal categories of variables that influence their success rates include: patient, implant, location, surgery, orthodontics, and implant maintenance related factors16,19. Each can be further subdivided and this paper will focus on insertion torque, which is a subgroup of surgery related factors. Description of the intervention Insertion torque (IT) results from frictional resistance between the screw thread and its surrounding bone and is a standard to evaluate mechanical stability20-23. Maximum insertion torque (MIT) is expressed in Ncm and is the maximum torque value recorded during the insertion of OMIs. Torque ratchets or torque sensors have been developed to measure these values and can inform the clinician instantly about the risk of implant fracture, the quality of the bone, root contact, excessive tightening and possibly stripping of the bone15. Digital sensors are recommended over mechanical devices because they can record consecutive IT levels at high frequency intervals. How the intervention might work Stability of implants can be divided into primary and secondary stability. The former is mechanical stabilization achieved immediately after insertion and the latter is attained when new bone forms at the implant interface24. To achieve initial stability, a certain level of IT is necessary11,25. Studies with dental implants have shown that increases in peak insertion torque can reduce the amount of micromotion and improve their success24,26. However, excessive stress to the bone can cause necrosis and local ischemia and might impede osseointegration and hence secondary stability11,14. This association was also suggested in various clinical studies in the orthodontic literature11,27-30. Animal studies have associated higher maximum insertion torque (MIT) values and overtightening of OMIs with fractures of the cortical bone13,15. The orthopaedic literature has shown that overtightening can damage and cause stripping of the bone, which can lead to a diminished holding strength with losses in pull-out strength up to 40-50%31,32. By modifying implant characteristics and surgical techniques the clinician can obtain desired MIT levels and therefore appropriate primary stability in sites with either stiff or fragile bone. This was confirmed in animal studies, which showed that changes in implant characteristics, predrilling or selfdrilling surgical techniques, and variations in the diameter of the pilot hole can significantly influence MIT values13,33,34. Why is it important to do this review Clinicians would like to know whether specific MIT recordings can help to improve success rates of OMIs. Further, if a range of ‘safe’ torque measures can be identified, they also want to learn which variables influence these values. Numerous studies on this topic have been conducted in human, animal, and laboratory studies11,13,20-22,28,35.

39

Many clinical papers have presented a MIT in the range of 5-10 Ncm as the “gold standard” and Google scholar recorded over 140 citations for one11 of these papers11,28,29. However conflicting outcomes have also been recorded in the orthodontic literature27,35,36. Because of this disagreement and because currently no reviews have addressed these issues, a systematic review was deemed appropriate. This paper has extracted just one (MIT) factor that could affect the stability of OMIs to make it more narrow in scope and therefore more wieldy. This choice was also supported by the outcomes of two recent broad scope systematic reviews8,16.

Objectives The objectives for this systematic review are summarized in the following PICO question: Is the application of MIT values in the range of 5-10 Ncm (Intervention) during the insertion of OMIs in patients (Participants) that require maximum anchorage during orthodontic treatment associated with higher success rates of OMIs (Outcomes) than those inserted with MIT values beyond this range (Comparison) ?

STEP B: Finding eligible studies

Criteria for considering studies for this review Eligibility criteria will be separated into 2 components: study characteristics and report characteristics2.

Study characteristics The study characteristics for the eligibility criteria are divided in: types of studies, participants, interventions and duration, and outcomes and are listed under here. Types of studies

• Human research studies with minimum samples sizes of 10 will be considered. Animal studies, laboratory studies, technique articles, case reports, opinion papers, reviews, and in vitro studies are excluded.

• Randomized controlled trial (RCT) will be our preferred choice of research design. • Non-randomized studies (NRS) with a low or moderate risk of bias will also be assessed37. This

decision was based on the justifications of the Cochrane Handbook for systematic reviews of interventions and The Oxford 2011 Levels of Evidence Working Group37,38. Their rationales for including NRS in a systematic review were: (1) high quality NRS could produce a better unbiased effect size compared to low quality RCTs; (2) randomized controlled trials (RCT) could be unavailable for ethical reasons; (3) NRS place the validity of the current literature in perspective and show the need for future research; (4) findings of a review of NRS may be helpful in designing subsequent studies; (5) NRS could reveal potential unexpected or rare harms of interventions37,39; and (6) the validity of NRS may be upgraded for demonstrating a large treatment effect6,40,41.

• If moderate or high quality RCTs will be identified, NRS will not be consulted for their treatment effect, but only for additional information on adverse effects of the intervention37.

40

Types of participants • Patients of either sex, in any ethnic, socioeconomic, or age group, or in any setting in need of

stationary anchorage during treatment with fixed orthodontic appliances were included. Types of interventions and duration

• Interventions that insert OMIs for stationary anchorage in the maxilla and or mandible of orthodontic patients followed by a phase of orthodontic loading.

• Interventions that measure MIT values during the insertion of OMIs. MIT represents primary stability and is defined as the maximum torque value recorded from the beginning to the end of the insertion process of OMIs20. MIT is expressed in Ncm and is recorded with either mechanical or electronic sensors.

• Interventions that apply forces for more than 120 days and measure their success are included42. This arbitrary time frame was chosen because most orthodontic objectives cannot be completed in less than 4 months.

• Interventions that use OMIs with a diameter smaller than 2.5 mm. This limit was chosen because larger implants would not classify for specific orthodontic indications (e.g., interradicular positioning). Articles on mini-plates will be excluded, because of their different biomechanical characteristics.

Types of outcome measures Primary and secondary outcomes as well as adverse effects will be assessed43. Eligibility will be established irrespective of the outcomes measured or reported43. Primary outcomes Success will be selected as the primary outcome for our PICO question. The following characteristics are defined for this parameter:

• An OMI will be considered successful when it can be loaded with orthodontic forces and fulfil its anchorage objectives during a minimum time period of 4 months.

• Success will be selected as the primary outcome and will be expressed in ratios (number of successful implants/total number of inserted implants).

• OMIs that were lost or had become unusable are considered failures. This group also includes implants that had fractured at insertion or during orthodontic treatment.

One adverse effect, fractured implants, will also be included as a primary outcome and is expressed in ratios (number of fractured implants/total number of inserted implants). Secondary outcomes: Three types of secondary outcomes will be assessed:

• Subjective stability of OMIs • Objective stability of OMIs • Variables that influence MIT values

Subjective stability of OMIs Stability of OMIs can either be measured subjectively by a clinician or objectively with various measuring devices44-46. Failure, immobility, mobility, and displacement will be used as parameters to classify subjective stability and were defined according to a recent systematic review by Reynders16:

• Success without mobility (score 0): implants with no clinical detectable mobility that could fulfil all necessary orthodontic anchorage objectives.

• Success with mobility (score 1): implants that had become mobile but could still fulfil all necessary orthodontic anchorage objectives.

• Success with displacement (score 2): implants that had become displaced but could still fulfil all necessary orthodontic anchorage objectives.

41

• Failure (score 3): implants that were lost or had become unusable, including those that had become inoperative because of excessive tissue proliferation that could not be reversed by simple excision; implants that had caused irreversible biologic damage; implants that could not be used because of the risk of causing irreversible biologic damage; and implants that fractured at placement, during orthodontic treatment, or at the removal of the screw.

• Not specified success (score NSS): the type of success of implants was not specified and included scores 0, 1, and 2 (Reynders 2009: 564e2)16.

Objective Stability of OMIs Objective stability of OMIs is measured at their removal with mechanical or digital instruments. Instrument specific values will be recorded, e.g. removal torque or resonance frequency values.

Variables that influence MIT values Variables that might influence MIT values of OMIs will be classified under the following headings: implant, patient, location, and surgery related factors. Associations between MIT and these parameters will be tested according to the following criteria:

• An association with MIT will be only considered if it is based on samples sizes of 10 implants or more.

• A proposed association with MIT will be rejected when the article presents direct proof that one or more influencing variables are not controlled. Lack of information about confounding factors will not be sufficient to reject an association.

• Only associations that will present their p values or confidence intervals are considered.

Adverse effects Adverse effects of insertion procedures of OMIs will be assessed according to the guidelines presented in a recent systematic review16. Adverse effects include: implant fracture at insertion, biologic damage, inflammation, and pain and discomfort. Reynders16 defined biologic damage according to the following parameters:

• No biologic damage (score 0): no biologic damage had occurred and no correcting dental procedures were necessary.

• Reversible biologic damage (Score 1): biologic damage that is completely reversible with simple dental procedures, including removal of hyperplastic tissue and fractured mini-implants that could be removed without causing irreversible damage.

• Irreversible biologic damage (score 2): biologic damage that is not completely reversible with simple dental procedures, including tooth, nerve, sinus, and blood vessel damage; fractured mini-implants that could not be removed; and need for orthognathic surgery caused by uncontrolled biomechanics with mini implants.

• Not specified biologic damage (score NSBD): biologic damage was described, but the type of damage was not identified.

• Post implant biologic damage (score PIBD): biologic damage caused by treatment with mini-implants, but it occurred or was found after the removal of the screw (Reynders 2009: 564e2)16.

42

Reynders16 defined inflammation according to the following parameters: Inflammation was measured either within the first month of implant placement or beyond this time limit and was defined as:

• No inflammation: (score 0): no signs of inflammation during the entire period of treatment with mini-implants

• Temporary inflammation (score 1): inflammation confined to the first month. • Continuing inflammation (score 2): Inflammation lasted longer than the first month. • Not specified inflammation (NSI): its duration was not specified (Reynders 2009: 564.e2)16.

Pain and discomfort were measured within the first 2 weeks after implantation or beyond and were defined according to Reynders16:

• No pain or discomfort (score 0): no pain or discomfort were present during the entire treatment period with OMIs.

• Moderate pain or discomfort (score 1): moderate pain or discomfort was noted within the first 2 weeks.

• Severe pain or discomfort (score 2): Severe pain or discomfort was noted within the first 2 weeks.

• Continuing pain or discomfort (score 3): pain lasted longer than 2 weeks. • Not specified pain (score NSP): the presence of pain and discomfort was described, but

their quality or duration were not specified (Reynders 2009:564.e3)16. These assessments of pain were based on the McGill Pain Questionnaire47.

Report characteristics Report eligibility criteria include: language of publications, publication status, and the year of publication2. Language of publications No language restrictions will be applied in the search strategy and pertinent articles will be translated and reviewed. Including non-English literature is essential, because Japanese and Korean journals have published extensively on OMIs8,16. Publication status Publication status refers to the inclusion of unpublished research and abstracts e.g. those extracted from the grey literature2. This material will be included when we will be able to retrieve full articles from the authors. These non-peer-reviewed papers will then be peer-reviewed independently by our 3 topic experts and assessed for eligibility. Systematic differences have been identified between outcomes of published research and those in the grey literature48-51. Such differences will be assessed through sensitivity analyses. Year of publication To prevent bias, literature searches will be conducted to the most recent date and should not be done through dates prior to 1997, the year of the introduction of OMIs52.

43

Information sources To find eligible studies we will apply the following protocol:

• We will consult the following electronic data bases through February 24, 2012: Google Scholar Beta, Pubmed (Medline), Embase (Ovid), CENTRAL, Science Direct, Scopus, Web of Science, LILACS, and AJOL49,53,54.

• The fourth author (Author 4), a librarian who specialized in computerized searches of health science publications at the Academic Medical Centre of the University of Amsterdam, will assist with the examination of these databases and the development of the search strategy.

• To improve transparency and facilitate future updating of this systematic review, the start and end dates for searches in each of these databases will be recorded2.

• Eligible reports will also be searched in research registers, conference proceedings, and the grey literature, because restricting searches exclusively to electronic databases can introduce bias49,53,54. We will search for grey literature in Google Scholar Beta, Open Grey, The Health Management Information Consortium (HMIC), and The National Technical Information Service (NTIS).

• To analyse if the keywords cover all articles on OMIs, the following journals will be handsearched: The American Journal of Orthodontics & Dentofacial Orthopedics, The Angle Orthodontist, The European Journal of Orthodontics, The Journal of Orthodontics , The Journal of Clinical Orthodontics, Seminars in Orthodontics, World Journal of Orthodontics, and The International Journal of Adult Orthodontics and Orthognathic surgery.

• In addition, references found in each identified paper will be manually screened for articles that possibly will be missed by the electronic search engines. All manual and electronic searches will be solicited for review articles and references found in these reviews will be screened for relevant papers53.

• We will also contact researchers and subject specialists to identify ongoing or unpublished studies2.

Search

• Transparency and reproducibility of our search process will be our primary goal and our search strategy will aim for high sensitivity and will accept low precision53.

• To avoid inappropriate exclusion, a wide variety of search terms will be combined, which will include the following subject headings and keywords: orthodontics, torque, implant, mini implant, micro implant, microimplant, screw, mini screw, miniscrew, micro screw, microscrew, temporary anchorage device. For each individual search engine the appropriate characters will be used to truncate or explore search terms. Nouns, adjectival, singular and plural forms of all keywords will be inserted53. Search filters will be avoided and the Boolean NOT operator will not applied53.

• Search strategies for each electronic data base will be pilot tested. • Search strategies for each electronic database will be presented in a table.

Study selection

• To minimize the risk of missing or rejecting eligible studies, three topic experts (Authors 1, 2, and 3) will select studies55-57.

44

• All selection procedures will be performed independently by these reviewers. The fifth author (Author 5) will guarantee the soundness of the methodology and statistics of this systematic review.

• To control inter-examiner agreement (kappa statistics) we will follow the protocol presented in the Cochrane Handbook for Systematic Reviews of Interventions57. According to this protocol, pilot tests on samples of reports will be used to refine and clarify the eligibility criteria and ensure that these criteria can be applied consistently57. Then, all titles and abstracts will be examined to remove obviously irrelevant reports. The full text of potentially relevant papers will be retrieved and reviewed. Ambiguous articles will also be read to prevent inappropriate exclusion.

• Unpublished studies which present sufficient data for a complete peer-reviewing process and qualify as moderate or high quality (GRADE) are also considered40,58. In the case that full eligible papers of unpublished studies cannot be retrieved from the investigators, outcomes will only be presented in the discussion and will be used to put the outcomes of the selected studies in perspective.

• To prevent bias through duplicate publication, special attention will be paid to identify multiple reports of single research studies57. For this purpose we will check whether these studies had similar methodology sections, published similar data in different journals within a short time period, had a retrospective design, and included the same authors. However, authors in such publications do not always overlap59. In case of uncertainties, authors of articles that are suspected of multiple publications will be contacted for additional information.

• Disagreements between authors about eligibility of papers will be resolved by rereading, discussion, and if necessary by contacting the authors of the original studies for clarification2.

• Selection procedures will be illustrated in a flow diagram according to the PRISMA statement2,3. Excluded full text articles will be listed in a table with the reason for their exclusion.

• The following protocol for contacting authors will be applied: (1) Emails will be sent with reference to the article of interest. Compliments are made about their study and we explain that we are conducting a systematic review on their topic of research. In this email authors are asked if they are willing to provide some additional information on their investigation; (2) In the case of a positive or a negative answer or no reply, a second email will be sent with questions regarding this missing information; (3) If after 2 mails no answers are received, the same letters will be sent from different email addresses; (4) If authors would still not respond to these letters, co-workers at, e.g. universities, will be searched and subsequently contacted.

STEP C: Data collection

Data collection process To reduce the risk of bias and introducing mistakes during data collection, the following protocol will be installed:

• Prior to conducting this systematic review, multiple data collection forms will be designed and pilot tested for their validity.

• These procedures and all data extraction will be done independently by the three topic experts.

• Outcome measures will be recorded according to the format presented by the authors. Transformation will be performed in a subsequent phase for the summary of findings tables57.

45

• When only effect estimates will be available, the generic inverse variance method will be applied to use these data in a meta-analysis. This application will only be accepted if effect estimates are accompanied by measures such as 95% confidence intervals, standard error or an exact P value57.

• In case of uncertainties about the flow of participants through a study, flow diagrams of individual studies will be drawn60.

• In the case of extracting data from multiple reports of the same study, data will be first extracted from each report separately and information will be subsequently combined across multiple data collection forms57.

• Disagreements between reviewers will be resolved by rereading and discussion. • Investigators of the selected studies will be contacted in case of uncertainty or inability to

extract all necessary information from the selected papers or in case of disagreement between reviewers57. If possible individual patient data (IPD) will be sought directly from the authors57.

• All persisting disagreements between reviewers will be reported in the review.

Data items

• Data items, which present any information about a study, including the source, eligibility, design, participants, interventions, outcomes, duration, results, and miscellaneous data will be collected in our data extraction forms57. Interventions will be further subdivided in: implant, location, surgery, orthodontics, and implant maintenance related factors. Although orthodontics and implant maintenance related factors do not influence insertion torque values, they will be recorded because they can influence success rates.

• Data that can facilitate the assessment of heterogeneity between studies and the risk of various biases will also be extracted.

• To guarantee the transparency of our data extraction procedures, we will list all data collection forms with a description of each extracted item in the appendix (Data collection forms I-XIII). We will also score the availability or quality of information for each item as ‘clear’, ‘not described’, ‘incomplete’, or ‘not applicable’. The definitions of these terms are presented in data collection form I.

• When items will be added or modified in these data collection forms after starting the review process, we will report these changes and their rationales2.

STEP D: Data analysis

Dealing with risk of bias in individual studies To ascertain the validity of the outcomes of the selected studies, the risk of bias for an outcome within a study (across domains) and across outcomes will be assessed.

• The Cochrane risk of bias tool will be modified specifically for our PICO question and will be pilot tested by the three topic experts. Risk of bias will be assessed independently by these reviewers and inter-examiner agreement will be calculated2,61.

• Critical judgments will be made for the following domains: selection, performance, detection, attrition, reporting, and across outcomes biases61. Performance bias will be further subdivided

46

in: implant, location, surgery, orthodontics, implant maintenance, outcome assessment, operator, and duration related systematic errors.

• To guarantee transparency of these judgments, definitions of each type of bias will be presented in table I. Assessments will be made on the magnitude, reliability, and the direction of bias. Judgments for the magnitude of systematic error will be defined as ‘high’, ‘low’ or ‘unclear’ risk of bias61. The Cochrane Collaboration61 assigns the latter score if: 1) Insufficient detail is reported of what happened in the study; 2) what happened in the study is known, but the risk of bias is unknown; 3) an entry is not relevant to the study at hand (2008d:197)61.

• Domains of bias will be only scored as ‘high’ risk when the article presents direct evidence that an entry is at risk of systematic error.

• Judgments for scoring the reliability and the direction of bias are defined in table I. • If multiple outcomes will be identified, specific tables will be developed for the assessment of

risk of bias across outcomes. • The analysis of risk of bias will be used as: (1) a possible explanation for the differences in

results; (2) a parameter to assess the validity of outcomes; (3) a reason for undertaking certain subgroup or sensitivity analyses; (4) a rationale for not undertaking a meta-analysis (Deeks 2008)4.

• The assessment of risk of bias will not be used as a threshold for the inclusion of studies.

47

Table I. The assessment of risk of bias within studies (across domains) and across outcomes4,61

Type of bias Description of bias Selection bias Refers to the impact of systematic error caused by the type of sample selection, e.g. random sequence

generation, allocation concealment, description of sample characteristics, and other selection related confounding factors on outcomes.

Performance bias participants

Refers to the impact of systematic error caused by, e.g. the number, age, physical status, dental status, and other participants related confounding factors on outcomes.

Performance bias implants

Refers to the impact of systematic error caused by, e.g. the description, quality, and other implant related confounding factors on outcomes.

Performance bias location

Refers to the impact of systematic error caused by, e.g. the site of insertion, character of the mucosa, exposure and other location related confounding factors on outcomes.

Performance bias surgery

Refers to the impact of systematic error caused by, e.g. the experience of the operator, the torqueing device, flap/flapless technique, distance between screws, direction and speed of insertion, self-drilling/pre-drilling insertion technique, starter and full length pilot holes, depth of insertion, axial load, stripping, and other surgery related confounding factors on outcomes.

Performance bias orthodontics

Refers to the impact of systematic error caused by, e.g. the type of orthodontic movement, timing and duration of force application, the magnitude, type and direction of the force, and other orthodontics related confounding factors on outcomes.

Performance bias implant maintenance

Refers to the impact of systematic error caused by, e.g. protocols for antibiotics, chlorhexidine rinses, oral hygiene, control of peri-implantitis and mobility and other implant maintenance related confounding factors on outcomes.

Performance bias outcome assessment

Refers to the impact of systematic error caused by the character and quality of outcome assessments, e.g. subjective or objective methods of assessments, and other outcome assessment related confounding factors on outcomes.

Performance bias operator

Refers to the impact of systematic error caused by the operator(s), e.g. blinding on outcomes.

Performance bias duration

Refers to impact of systematic error caused by the duration of the intervention, e.g. the time of follow-up or a precise description of the duration of the study on outcomes.

Detection bias assessors

Refers to the impact of systematic error caused by the outcome assessor(s), e.g. blinding, number of assessors, inter-observer agreement, and other detection related confounding factors on outcomes.

Attrition bias Refers to the impact of systematic error caused by incomplete outcome data, e.g. completeness of sample, follow-up and data on outcomes.

Reporting bias Refers to the impact of systematic error caused by the type of reporting, e.g. selective or incorrect reporting on outcomes.

Across outcomes bias

Refers to differences in the impact of systematic error on different types of outcomes. This assessment is only applicable when more than one outcome is assessed.

Judgments on the risk of bias were made for the magnitude, reliability, and the direction of bias. Magnitude of judgment on risk of bias: ‘high’, ‘low’ or ‘unclear’.* Reliability of judgment on risk of bias: ’reliable’, ‘unreliable’ or ‘unclear’** Direction of judgment on risk of bias: ‘under-estimating’ ,‘over-estimating’ the treatment effect’ or ‘unclear’** *Unclear: The Cochrane Collaboration61 assigns ‘unclear’ risk of bias if: (1) insufficient detail is reported of what happened in the study; or (2) what happened in the study is known, but the risk of bias is unknown; or (3) an entry is not relevant to the study at hand (2008:197)61. **Unclear: Indicating either lack of information or uncertainty on this judgement.

48

Summary measures Risk and odds ratios are chosen as the principal summary measures for our dichotomous primary outcomes. These ratios are calculated from the success ratios in the different experimental groups (number of successful implants/total number of inserted implants).

Dealing with unit of analysis issues Unit of analysis issues will be assessed according to the level at which randomization occurs and will be analysed for each specific study design4. Unit of analysis issues could arise in studies of long duration with repeated observations, e.g. assessments of stability of OMIs at different times of follow-up4.

Dealing with missing data

• Electronic searches with high sensitivity will be conducted to identify pertinent papers53. • The specific reasons for missing data will be assessed62. • Original investigators will be contacted to obtain missing data according to our communication

with authors protocol. • Imputation techniques for missing standard deviations will be avoided wherever possible63. • The methods and assumptions for dealing with missing data will be explicitly presented, e.g.

clearly explain whether it will be assumed that data were missing at random or not63. Analyses of sensitivity will be conducted to assess how reasonable changes in the methods and assumptions can affect the results63.

• A statistician will be consulted for choosing appropriate statistical models for dealing with missing data63.

• Our choice of a sensitivity analysis e.g best-case and worst case scenarios, an available case analysis, or an Intention-to-treat (ITT) analysis will depend on the numbers and proportions and the specific reasons for missing data in the selected studies63.

• The ITT analysis will not be applied for the assessment of adverse effects63. • The possible impact of missing data on outcomes will be presented in the summary of findings

table and in the discussion63.

Dealing with heterogeneity Prior to handling the data and synthesizing the results, we will assess possible sources of variability in outcomes between studies and will apply our protocol for dealing with heterogeneity. Variations within a study are also analyzed, because additional heterogeneity can be introduced when for example different types of OMIs are used within a trial arm.

Potential sources of heterogeneity To facilitate the assessment of heterogeneity, tables are created that list clinical, methodological, and other sources of heterogeneity (Tables II-IV)4,64.

49

• Clinical heterogeneity refers to variability in the types of patients, interventions, outcomes, duration, and coexistent conditions (Table II)4,64. Interventions related sources of clinical heterogeneity are further subdivided in implant, location, surgery, orthodontics, and implant maintenance related sources (Table III).

• Methodological heterogeneity refers to differences in effect sizes between studies caused by the design, conduct and analysis, and size of the study4,64. These sources of bias are further described in Table IV. To assess systematic error as a potential source of heterogeneity we will apply our risk of bias tool in Table I.

• Other sources of heterogeneity can include, publication bias, funding, the cost of treatment, and the country where the study was conducted65.

Table II. Potential sources of clinical heterogeneity

Potential source of heterogeneity

Description

Type of patients Sources of heterogeneity that relate to ethnicity, socio-economic status, physical status, dental status, sex, age, country, setting, co-existent conditions, and other patient specific characteristics.

Type of interventions Sources of heterogeneity that relate to implant, location, surgery, orthodontics, and implant maintenance related factors.

Type of outcomes Sources of heterogeneity that relate to clinical success measures, subjective stability measures, objective stability measures, unit of measurement, or other types of outcomes related factors.

Duration of the intervention

Sources of heterogeneity that relate to assessment of outcomes, e.g. at the end of force application, at the end of orthodontic treatment, or at other time points of follow-up.

Coexistent conditions Sources of heterogeneity that relate to, e.g. past interventions, co-interventions etc. Table III. Potential sources of interventions related clinical heterogeneity

Potential source of heterogeneity

Description

Implant related sources

Sources of heterogeneity that relate to, e.g. the implant type, form, material, diameter, length, surface finish, drilling design, thread characteristics, pitch dimensions, flute in tip, flute in core, being sterilized by an implant company or not, and other implant related variables.

Location related sources

Source of heterogeneity that relate to, e.g. the implant site , distance to root, assessment of bone condition, bone thickness, keratinized/non keratinized mucosa, mucosal thickness, exposed/unexposed implant, and other location related factors.

Surgery related sources

Sources of heterogeneity that relate to, e.g. the experience of the operator, one or multiple operators, flap or flapless surgery, self-drilling or pre-drilling insertion technique, dimensions of the pilot holes for the self-drilling technique, pre-drilling device for the self-drilling pilot holes, pre-drilling speed for the self-drilling pilot holes, dimensions of the pilot holes for the pre-drilling technique, pre-drilling device for the pre-drilling pilot holes, pre-drilling speed for the pre-drilling pilot holes, distance between the screws, insertion depth, friction between the head of the screw and the bone, the direction of insertion, insertion speed, axial insertion force, insertion with insertion torqueing device, type of torqueing device, stripping of the bone, mono or bicortical anchorage, and other surgery related factors.

Orthodontics related sources

Sources of heterogeneity that relate to, e.g. the type of orthodontic movement, time of force application, force magnitude, type of force, duration of force application, and the direction of the force.

Implant maintenance related sources

Sources of heterogeneity that relate to, e.g. the medication, chlorhexidine, oral hygiene, peri-implantitis, and mobility protocol.

50

Table IV. Potential sources of methodological and other sources of heterogeneity Potential source of heterogeneity

Description

Design of the study Sources of heterogeneity related to the type of selection procedures, e.g. random sequence generation, quasi random sequence generation, other types of sequence generation, concealment of allocation, and other selection related variables.

Conduct and analysis of the study

Sources of heterogeneity related to the conduct and analysis of the study, e.g. selection bias, performance bias, detection bias , attrition bias, reporting bias, and scientific misconduct.

Study size Sources of heterogeneity related to study size. Publication bias Sources of heterogeneity related to publication bias. Funding Sources of heterogeneity related to the source of funding. Cost of treatment Sources of heterogeneity related to the cost of treatment. Country Sources of heterogeneity related to the country where the research study was conducted.

Strategies for dealing with heterogeneity Although there is no consensus on how to deal with heterogeneity in a systematic review, the following step by step approach will be applied4,64,66-68:

(1) To reduce bias we specify a priori which potential sources of heterogeneity we want to assess. These factors include: specific insertion torque values, the design of the study, the diameter and length of the implant, the duration of the intervention, the location of the implant, and the time of force application.

(2) After data extraction, tables are created for each source of heterogeneity and data are controlled for correct data extraction;

(3) Original investigators will be contacted if clarification on certain data will be necessary; (4) Data extraction tables are evaluated to assess whether it made sense to apply statistics or not,

because heterogeneity can be so excessive that undertaking a meta-analysis could mislead the reader69. The simultaneous presence of major variations in study design, screw diameter, and follow-up between studies could be an example for not pooling outcomes in a meta-analysis;

(5) When it is decided to measure heterogeneity, the following statistics will be used: Q, p, T2, T 4,66,70,71. Because clinical and methodological diversity between studies should be expected a measure of inconsistency among studies, the I2

statistic, is also calculated66,71,72. Using these statistics should be done with caution, because they have low power when the total number of studies is small67. A drawback of using the I2 statistic is that its value increases when the sample sizes of the include studies increase73. Further, statistical heterogeneity can be present with or without clinical heterogeneity and methodological heterogeneity4,64,70. In addition, poor precision could mask true heterogeneity70;

(6) In the presence of minimal, moderate or substantial heterogeneity (I2 in the range of 0-90%), a meta-analysis will be conducted to produce either a summary effect or to analyse the dispersion of outcomes4,74-76. In the presence of excessive heterogeneity (higher than 90%), the systematic review will be presented in a narrative format, because presenting a meta-analysis under those conditions can only mislead the reader73,75;

(7) The following statistics are available to further explore heterogeniety: (a) assessing the causes of heterogeneity through subgroup analyses and meta-regression; (b) ignoring heterogeneity by using the fixed-effect meta-analyses; (c) conducting a random-effects meta-analysis to

51

explore unexplained heterogeneity; (d) changing the type of effect measure; and (e) excluding studies4. Each of these strategies should be applied with considerable concern74.

Dealing with publication bias Publication bias is a type of systematic error across studies that can affect the summary effect2. Clarke77 defines publication bias as “the selective reporting of some or all of the results of a research study”. Potential sources of publication bias The impact of publication bias in the orthodontic literature has not been studied extensively78. We therefore refer to ‘potential’ sources of publication bias. These sources are listed in table V and include: outcome, limited or incomplete reporting of findings, funding source, location, time lag, and multiple publication related bias54. Descriptions of each entry are also presented in this table. Table V. Potential sources of publication bias

Potential source of publication bias

Description

Outcome related bias The selective reporting of some or all of the results of a research study depends on the nature and the direction of the results.

Limited or incomplete reporting of findings related bias

The selective reporting of some or all of the results of a research study depends on the limited or incomplete reporting of findings.

Funding source related bias The selective reporting of some or all of the results of a research study depends on the funding source.

Location related bias The selective reporting of some or all of the results of a research study depends on the location of the study e.g. impact factor, specialty, language, and country of the journal, the health field, the accessibility of the research findings in electronic databases, and citation.

Time lag bias The time between the completion of a research study and the publication of some or all of the results depends on the nature and direction of the results of this study.

Multiple publication bias Duplicate or multiple publications of some or all of the results of a single research study.

Strategies for dealing with publication bias Strategies for dealing with publication bias refer to the application of a variety of procedures: (1) the development of a the systematic review protocol and avoiding post hoc changes; (2) the search strategy for the identification of studies; (3) selection of studies; (4) data collection and analysis; (5) communication with authors; and (6) recruiting investigators. These strategies are further explained in in Table VI.

52

Table VI. Strategies for dealing with publication bias

Strategy for dealing with publication bias

Description

Development of a protocol

Refers to controlling publication bias through the development of a systematic review protocol without prior content knowledge of the existing studies on the PICO question.

Search strategy for the identification of studies

Refers to controlling publication bias by applying specific search strategies for the identification of studies e.g. consulting a librarian, pilot testing of the search strategy, applying no language restrictions, conducting extensive searches, including relevant search periods, and other measures to obtain relevant studies.

Selecting studies Refers to controlling publication bias by applying specific measures for selecting studies e.g. pilot testing of the selection procedures, independent selection of studies by 2 or more topic experts, assess ambiguous articles, assess presence of multiple publications, include only full articles, peer review full grey literature articles, conduct subgroup analyses of articles from the grey literature, translate non-English articles, and contact authors when necessary etc.

Data collection and analysis

Refers to controlling publication bias through specific data collection and analysis procedures e.g. designing data extraction forms specifically to identify and assess sources of publication bias.

Communication with authors

Refers to controlling publication bias through contacting authors in order to obtain e.g. full articles, individual patient data, and information on multiple publications of single studies etc.

Recruiting investigators

Refers to controlling publication bias through recruiting investigators as collaborators in the systematic review.

Publication bias can be explored to a certain extent through a variety of statistical methods54,65,79. We plan to conduct funnel plots for the identification and quantification of publication bias54,79. Funnel asymmetry will be only assessed when at least 10 eligible articles are identified, because with fewer studies the power of this test is too low to separate real asymmetry from chance alone54.

STEP E: Assessment of the quality (GRADE) for each body of evidence For the assessment of the quality of a body of evidence we will apply the GRADE approach, which was also adopted by the Cochrane Collaboration6,40,58. We will apply the GRADE rating protocol according to the following step by step approach: (1) List and define parameters that can affect the GRADE rating, e.g. the design of a study, risk of bias, directness of evidence, precision of results, heterogeneity, unit of analysis issues, and reporting bias; (2) Define rating parameters for the GRADE assessment (high, moderate, low, or very low) (table VII)40; (3) Score these quality parameter for each individual outcome; (4) Assess factors that may increase or decrease the level of quality. Factors that can decrease the quality of a body of evidence include: limitations of the design of a study, high risk of bias, indirectness of evidence, imprecisions of results, unexplained heterogeneity, unit of analysis issues, and a high probability of publication bias. Factors that can upgrade the body of evidence include: a large treatment effect, or when all possible biases would affect the magnitude of a treatment effect in the same direction4,6,38,40,75; and (5) Score a ‘final’ quality rating (GRADE). We will only consider outcomes of high or moderate quality.

53

Table VII. Quality of evidence (GRADE) and definitions40 High Further research is unlikely to change our confidence in the estimate of effect. Moderate

Further research is likely to have an important influence on our confidence in the estimate of effect and may change the estimate.

Low Further research is very likely to have an important influence on our confidence in the estimate of effect and is likely to change the estimate.

Very low Any estimate of effect is uncertain.

STEP F: Summary of findings table for each primary study Our summary of findings table for each primary study will provide the following information: participants, settings, interventions, comparison, outcomes, adverse effects, risk of bias, and the quality of evidence80. Outcomes will be further subdivided to provide information on: risk ratios, confidence intervals, number of participants, quality of the evidence, and comments.

STEP G: Data synthesis Synthesis of results

• After undertaking steps D, E, and F of our protocol (figure 1) we will assess whether to conduct a meta-analysis or not. Strategies for dealing with these issues are based on: (I) the risk of bias for an outcome within a study (across domains) and across outcomes; (II) summary measures (III) unit of analysis issues; (IV) missing data; (V) heterogeneity; (VI) publication bias; and (VII) the quality (GRADE) for each body of evidence.

• Justifying criteria for conducting a meta-analysis are illustrated in another flow diagram, which was based on The Cochrane Handbook for Systematic Reviews of Interventions, the handbook “Introduction to Meta-Analysis, and the PRISMA statement (Figure 2)2-5,74.

54

Figure 2. Flow diagram for our protocol for undertaking a meta-analysis1,74

STEP A: Justifying a meta-analysis Step 1) Establish criteria for justifying a meta-analysis: (1) low risk of bias in eligible studies (high or moderate quality GRADE); (2) consistent direction of effect sizes (treatment effect) across the range of studies; (3) low reporting bias; (4) a high number of eligible studies; (5) low heterogeneity between studies. Step 2) Assess criteria 1-5 prior to undertaking a meta-analysis. Step 3) Measure the summary effect or analyze the dispersion of outcomes in a meta-analysis or do not undertake a meta-analysis. Step 4) In the case that a meta-analysis is not justified, the systematic review will be presented in a narrative format.

STEP B: Undertaking a meta-analysis Step 1) Calculate the appropriate effect measures of the intervention for each primary study, e.g. risk or odds ratio. Step 2) Calculate the weight average of the intervention effect of all eligible primary studies. Step 3) Assess, based on a priori defined criteria, whether to conduct a fixed- effects or a random-effects meta-analysis or both.

STEP C: Conducting subgroup and sensitivity analyses Step 1) Assess the justifying criteria for undertaking a meta-analysis 1-4. Step 2) Consider that a small number of studies is acceptable for the fixed-effect, but not for the random-effect model meta-analysis. Step 3) Explore heterogeneity (justifying criterion 5) and if indicated conduct subgroup analyses and meta-regressions for the a priori defined potential sources of heterogeneity. Step 4) Conduct sensitivity analyses to assess the effects of procedural decisions on findings.

STEP E: Present findings Step 1) Present a forest plot with the summary effect size when the heterogeneity between studies is low. Step 2) Present only the dispersion in observed effects when the heterogeneity between studies is high. Step 3) Present additional tables and figures

55

• Criteria for performing a meta-analysis of the summary effect size are based on statistical, clinical and methodological assessments . Such a synthesis is conducted in the case of: (1) low risk of bias in eligible studies; (2) consistent effect sizes (treatment effect) across the range of studies; (3) low reporting bias; (4) a high number of eligible studies; (5) low heterogeneity between studies4,61,75.

• In the presence of minimal, moderate or substantial heterogeneity (I2 in the range of 0-90%), a meta-analysis will be conducted to produce either a summary effect or to analyse the dispersion of outcomes4,76. In the presence of excessive heterogeneity, the systematic review will be presented in a narrative format, because presenting a meta-analysis can only mislead the reader69,75.

• We will assess whether a fixed-effects or a random-effects meta-analysis of dichotomous data should be conducted or both.

• A fixed-effect meta-analysis will be conducted when the effects of interventions in both magnitude and direction will be the same in all eligible studies. Random effects models will be applied when intervention effects are related but not identical and follow some distribution4.

Subgroup analysis

• Heterogeneity will be explored through subgroup analyses and meta-regression4. These analyses assess whether summary effects vary according to specific characteristics of the included studies2. Pre-specified subgroups include: specific insertion torque values, the design of the study, the diameter and length of the implant, the duration of the intervention, the location of the implant, and the time of force application.

• For each subgroup analysis we will state whether an analysis was pre-specified or was decided post hoc.

• Subgroup analyses will only be performed if the study itself or IPD provide sufficient data for such assessments4. In the case of insufficient data, authors of eligible studies are contacted to obtain IPD.

• Meta-regressions will be performed to assess if the size of the effects of interventions are related to different characteristics of the trials. These variables include: allocation sequence concealment, blinding of outcome assessors etc. Meta-regression are only considered when at least 10 eligible articles are identified for the characteristic that will be modelled4.

• A statistician will be consulted for both subgroup analyses and meta-regressions. • Interpretation and application of the findings from subgroup analyses and meta-regressions

will be done with caution4.

Sensitivity analysis

• Sensitivity analyses will be applied to assess the impact of decisions regarding eligibility criteria, searching for studies, type of data, methods of analysis etc. on outcomes4.

• Sensitivity analyses will be conducted for: the decisions to include NRS, grey literature, mechanical torque sensors, studies with high risk of bias, and studies with follow-ups longer than 12 months. Analyses on the impact of including studies from the grey literature are also explored, because systematic differences have been identified between outcomes of published research and those in the grey literature48-51.

• When decisions to undertake sensitivity analyses are made post hoc, it will be stated and the rationales are explained.

56

• If sensitivity analyses will show that particular decisions can influence the findings of the systematic review, trial authors will be contacted for verification and to obtain IPD. In the case that this procedure will not be successful, findings will be interpreted with caution4.

Differences between the protocol and the review Full description of any changes of methods from the protocol to the review will be presented. We will describe when these decisions were taken, the rationale for the changes and their potential for introducing bias5.

57

Data collection forms Data collection form I. Character of information and risk of bias judgements

Entry Description Character of the information

The character of the information was scored for each item in the data collection forms and was defined as: C: Clear: the information is described and complete. ND: Not described: the information is missing. IC: Incomplete: some information is presented and some is missing; ‘partially clear’. NA: Not applicable.

Risk of bias Judgments on the risk of bias were made for the magnitude, reliability, and the direction of bias. Magnitude of judgment on risk of bias: ‘high, low or unclear’. Reliability of judgment on risk of bias:’ reliable, unreliable or unclear’. Direction of judgment on risk of bias: ‘under-estimating, over-estimating the treatment effect or unclear’. The Cochrane Collaboration61 assigns ‘unclear’ risk of bias when: 1) insufficient detail is reported of what happened in the study; 2) what happened in the study is known, but the risk of bias is unknown; 3) an entry is not relevant to the study at hand (2008:197)61.

Data collection form II. Source and eligibility37,57

Entry Description and character of the information Authors, title, journal of article

List authors, title, and journal.

Source of article Describe how the article was retrieved, e.g. Medline, grey literature, hand searching of review articles etc.

Language of the article Describe in which language the article was published. Name of reviewer Report the name of the reviewer. Trial register Describe if trial was registered, under what number and in which register. Eligible/not eligible Confirm eligibility. Explain reason for exclusion. Purpose Copy the objectives of the paper according to the authors.

58

Data collection form III. Study design37,57 Entry Description and character of the information Date of the study Describe when the study was started and completed. Study design Describe the type of study design, e.g. randomized controlled trial, case-control study,

cohort, splitmouth etc. ? Sequence generation 1) Random sequence generation ? Describe the type of randomization, e.g.

computerized. 2) Non-random sequence generation ? A) Quasi random sequence generation ? Allocation based on, e.g. alternation, date of birth, day of admission etc. B) Other types of sequence generation ? Allocation based on:

a) Preference related factors, e.g. preferences of researchers, operators, and or patients.

b) Logistics related factors, e.g. organizational issues. c) Operator related factors, e.g. experience (learning curve). d) Patient related factors, e.g. health and dental status. e) Location related factors, e.g. anatomy of the implant site. f) Treatment related factors, e.g. orthodontic treatment objectives and

biomechanics. g) Implant maintenance related factors, e.g. oral hygiene. h) Outcome related factors, e.g. the stability of the implant.

Allocation concealment Concealment of allocation. Describe blinding of patients, surgeon, loading operator, personnel, and outcome assessors.

Risk of selection bias Refers to the effect of systematic error in one or more of the above described entries on outcomes. Score the magnitude, reliability, and the direction of bias (Data collection form I). Explain the rationale for these scores.

Approved by ethical board

Describe if the research study was approved by an ethical board and describe the components of this ethical board.

Study design critique Describe strengths and weaknesses of the chosen study design. Data collection form IV. Participants57

Entry Description and character of the information Number Present the total number of patients and the number of patients per subgroup. Ethnicity Which ethnic group(s) ? If applicable, present the number of patients per subgroup. Socio-economic status Describe the socio-economic status of the patient(s)? If applicable, present the number

of patients per subgroup. Physical status Describe variables, e.g. uncontrolled diabetes, osteoporosis, smoking, pharmacological

treatment etc. If applicable, present the number of patients per subgroup. Dental status Describe variables, e.g. periodontal disease, loss of teeth etc. If applicable, present the

number of patients per subgroup. Sex Present the divisions of the sexes in the research study. If applicable, present the

number of patients per subgroup. Age List age in years and months with standard deviations and/or ranges. If applicable,

present the number of patients per subgroup. Additional patient characteristics

Describe additional characteristics, e.g. university or high school students, dental students etc. If applicable, present the number of patients per subgroup.

Country Describe in which country(ies) the research study was conducted. If applicable, present the number of patients per subgroup.

Setting Describe the setting of the research study, e.g. dental school. If applicable, present the number of patients per subgroup.

Co-existent conditions Describe possible past interventions or co-interventions etc. If applicable, present the number of patients per subgroup.

Performance bias participants

Refers to the effect of systematic error in one or more of the above described entries on outcomes. Score the magnitude, reliability, and the direction of bias (Data collection form I). Explain the rationale for these scores.

59

Data collection form V. Interventions— implants57

Entry Description and character of the information Number of implants Present the total number of implants and if applicable the number of implants per

subgroup. Implant type Present the type of implant (code number) with the name of the company(ies) and if

applicable the number of implants per subgroup. Implant form Describe the form of the screw: tapered/cylindrical/or a combination of forms.

Describe these specifics for each part of the screw. Implant material Describe the chemical make-up of the screw. Diameter (D) Length (L)

Describe the thread diameter and core diameter in mm. Describe the thread length and body length in mm.

Surface finish Describe the type of surface finish. Drilling design Describe if the screw has a self-drilling or a pre-drilling design. Thread characteristics Describe the width and angles of the crest of the screw. Pitch dimensions Describe the dimensions of the pitch of the screw. Flute in tip Describe if the tip of the screw is fluted or not.

Describe the length of the flute. Flute in core Describe if the core of the screw is fluted or not.

Describe the length of the flute. Sterilized Describe if the screws are sterilized by the company or by the operator. Performance bias implants


Data collection form VI. Interventions— location57

Entry Description and character of the information Implant site Describe the exact insertion site of the implant. Distance to root Describe the distance between the implant and the root of a tooth in mm. Assessment of bone condition

Describe if any quality assessment of the bone e.g. radiographically was undertaken prior to surgery.

Bone thickness Describe the thickness of the cortical and trabecular bone in mm. Keratinized/non keratinized mucosa

Describe if the implant was placed in the keratinized or non keratinized mucosa.

Mucosal thickness Describe the thickness of the mucosa in mm. Exposed/non exposed Describe if the implant was left exposed or non exposed under the mucosa after surgery. Performance bias location


60

Data collection form VII. Interventions— surgery57 Entry Description and character of the information Experience of the operator Describe the number of mini-implants inserted by the operator prior to this

study. One or multiple operators Describe if all surgeries were conducted by the same operator and if not how

many implants were placed by each operator. Flap or flapless surgery Describe if surgery was flap or flapless. Self-drilling or pre-drilling insertion technique

Describe if surgery was conducted according to the self-drilling or pre-drilling insertion technique.

Pilot holes for the self-drilling technique

Describe if pilot holes were drilled for the self-drilling technique and describe diameter and depth in mm.

Pre-drilling device for the self-drilling pilot holes

Describe the type of device that was used for the self-drilling pilot holes.

Pre-drilling speed for the selfdrilling pilot holes

Describe the pre-drilling speed in rpm (rounds per minute) for the self-drilling pilot holes.

Pilot holes for the pre-drilling technique

Describe if pilot holes were drilled for the predrilling technique and describe the diameter and depth in mm.

Pre-drilling device for the pre-drilling pilot holes

Describe the type of device that was used for the pre-drilling pilot holes.

Pre-drilling speed for the predrilling pilot holes

Describe the pre-drilling speed in rpm (rounds per minute) for the predrilling pilot holes.

Distance between the screws In the case of multiple screws, describe the distances between the screws in mm.

Insertion depth Describe the insertion depth in mm. Friction between the head of the screw and the bone

Describe if the screw head touched the bone and had caused friction. Such a contact could have caused stripping of the bone.

The direction of insertion Describe the angulation of insertion of the screw in degrees. Insertion speed Describe the insertion speed in rpm (rounds per minute). Axial insertion force Describe the axial insertion force in kg. Insertion with insertion torqueing device

Describe if the insertion was conducted with a torqueing device.

Type of torqueing device Describe the type of the insertion torqueing device and describe if it was a mechanical or a digital device.

Stripping of the bone Describe if stripping occurred (rotation of the screw without penetration of the bone).

Mono or bicortical anchorage Describe if the anchorage was mono or bicortical. Performance bias Surgery


61

Data collection form VIII. Interventions— orthodontics57 Entry Description and character of the information Type of orthodontic movement

Describe the type of orthodontic tooth movement, e.g. intrusion, extrusion, bodily movement etc.

Time of force application

Describe the time of force application in days, e.g. at the time of surgery or days or weeks after the insertion of the mini-implants.

Force magnitude Describe the magnitude of the force in grams that was applied to the mini-implant and present possible changes in the magnitude of force during treatment.

Type of force Describe the type of force that was applied to the mini-implants, e.g. intermittent or continuous forces or both. For the former and latter type of force application, describe the duration and the force magnitude of each activation.

Duration of force application

Describe the duration of the force application in weeks or days with the standard deviation.

Direction of force Describe the direction of the force application, e.g. perpendicular, extrusive etc. Performance bias orthodontics


Data collection form IX. Interventions— implant maintenance57

Entry Description and character of the information Medication protocol Describe if medication pre and/or post-surgery was prescribed. Include name of

medication, dosage, frequency, and duration. Chlorhexidine protocol Describe if chlorhexidine rinses were prescribed post-surgery. Include the

concentration, frequency, and duration. Oral hygiene protocol Describe if an oral hygiene protocol was prescribed and if so describe the details,

frequency, and duration. Peri-implantitis protocol

Describe if patients were controlled for peri-implantitis post-surgery and if so describe the details, frequency, and duration.

Mobility protocol Describe if the mobility of the mini-implants was assessed post-surgery and if so describe the details, frequency, and duration.

Performance bias implant maintenance


62

Data collection form X. Primary and secondary outcomes and duration57 Entry Description and character of the information Clinical success measures

Describe the definition of “clinical success measures” according to the authors, e.g. the ability of an implant to fulfill the anchorage objectives of pre-established tooth movements or the ability of an implant to fulfill its anchorage objectives during a specific pre-established period.

Unit of measurement Describe the unit of measurement of clinical success measures, e.g. clinical success rates in percentages or ratios (successful implants/total number of inserted implants).

Subjective stability measures*

Describe the definition of “subjective stability measures” according to the authors, e.g. no mobility, mobility, mobility and displacement, not specified mobility etc.

Unit of measurement Describe the unit of measurement of subjective stability measures, e.g. subjective stability rates in percentages or ratios (successful implant stability/total number of inserted implants).

Objective stability measures

Describe the type of recordings of “objective stability measures”, e.g. recordings with removal torque or resonance frequency sensors.

Unit of measurement Describe the unit of measurement of objective stability measures, e.g. Newton centimeter (Ncm).

Duration of the intervention

Describe how the duration of the intervention was defined, e.g. duration of force application, duration of the implant in situ, duration of orthodontic treatment etc.

Time point of outcome assessment

Describe the time point of follow-up of the outcome assessment, e.g. after days, weeks, or months.

Outcome related bias Refers to the effect of systematic error in one or more of the above described outcome related entries on outcomes. Score the magnitude, reliability, and the direction of bias (Data collection form I). Explain the rationale for these scores.

Duration related bias Refers to the effect of systematic error in one or more duration related entries on outcomes. Score the magnitude, reliability, and the direction of bias (Data collection form I). Explain the rationale for these scores.

*Subjective stability measures: Score 0: Success without mobility Score 1: Success with mobility Score 2: Success with displacement Score 3: Failure NSS: Not specified success (includes scores 0-2)

63

Data collection form XI. Adverse effects57 Entry Description and character of the information Biologic damage Describe how biologic damage is scored:

• No biologic damage (score 0): no biologic damage had occurred and no correcting dental procedures were necessary.

• Reversible biologic damage (Score 1): biologic damage that is completely reversible with simple dental procedures, including removal of hyperplastic tissue and fractured mini-implants that could be removed without causing irreversible damage.

• Irreversible biologic damage (score 2): biologic damage that is not completely reversible with simple dental procedures, including tooth, nerve, sinus, and blood vessel damage; fractured mini-implants that could not be removed; and need for orthognathic surgery caused by uncontrolled biomechanics with mini implants.

• Not specified biologic damage (score NSBD): biologic damage was described, but the type of damage was not identified.

• Post implant biologic damage (score PIBD): biologic damage caused by treatment with mini-implants, but it occurred or was found after the removal of the screw (Reynders 2009: 564e2)16.

Implant fracture Describe for each subgroup the number of implants that fractured during their insertion and removal. Present the fracture ratio (the number of fractured implants/total number of implants).

Inflammation Describe how inflammation is scored: • No inflammation: (score 0): no signs of inflammation during the entire

period of treatment with mini-implants • Temporary inflammation (score 1): inflammation confined to the first

month. • Continuing inflammation (score 2): Inflammation lasted longer than the

first month. • Not specified inflammation (NSI): its duration was not specified

(Reynders 2009: 564.e2)16. Pain and discomfort Describe how pain and discomfort are scored:

• No pain or discomfort (score 0): no pain or discomfort were present during the entire treatment period with OMIs.

• Moderate pain or discomfort (score 1): moderate pain or discomfort was noted within the first 2 weeks.

• Severe pain or discomfort (score 2): Severe pain or discomfort was noted within the first 2 weeks.

• Continuing pain or discomfort (score 3): pain lasted longer than 2 weeks. • Not specified pain (score NSP): the presence of pain and discomfort was

described, but their quality or duration were not specified (Reynders 2009:564.e3)16.

64

Data collection form XII. Results57 Entry Description and character of the information Number of participants/ implants

Summarize division of experimental groups, e.g. number of patients or implants per trial arm or subgroup.

Sample size for each outcome Present the sample size for each outcome. Missing participants/ implants for each outcome

Present the missing participants or implants for each outcome.

Maximum insertion torque values

Present the maximum insertion torque values for each patient or implant subgroup.

Duration of interventions Present the duration of the intervention for each patient or implant subgroup. Subjective stability Present the success rates in percentages and in ratios (number of successful

implants/total number of inserted implants) for each patient or implant subgroup. Present also the subjective stability measures for each outcome.*

Objective stability Present objective stability values for each patient or implant subgroup. Association between maximum insertion torque values and success

Present association between maximum insertion torque values and success for each patient or implant subgroup.

GRADE rating ** Score GRADE rating for each association between maximum insertion torque and success and present the rationale for this rating.

Associations between variables and maximum insertion torque values

Present associations between, e.g. implant, patient, location, and surgery related factors and maximum insertion torque values for each patient or implant subgroup. Present also the rationale for rejecting any of these associations.

Adverse effects Present ratios of adverse effects of interventions, e.g. implant fracture ratios (fractured implants/total number of inserted implants). Score adverse effects according to ‘Data collection table XI’.

*Subjective stability measures: Score 0: Success without mobility Score 1: Success with mobility Score 2: Success with displacement Score 3: Failure NSS: Not specified success (includes scores 0-2) **Quality of evidence (GRADE)40: High: Further research is unlikely to change our confidence in the estimate of effect. Moderate: Further research is likely to have an important influence on our confidence in the estimate of effect and may change the estimate. Low: Further research is very likely to have an important influence on our confidence in the estimate of effect and is likely to change the estimate. Very low: Any estimate of effect is uncertain.

65

Data collection form XIII. Results57 Entry Description and character of the information Summary data for each intervention group: Primary outcomes

Study group Event (success) No event (failure)

Total

MIT (5-10 Ncm) MIT (Beyond 5-10 Ncm)

Risk ratio: Odds ratio: Risk difference: Standard error: Confidence interval: P value:

Summary data for each intervention group: Secondary outcomes, e.g. removal torque values

Study group Removal torque value (success)

Removal torque value (failure)

Total

MIT (5-10 Ncm) MIT (Beyond 5-10 Ncm)


Summary data for each intervention group: Implant fracture

Study group No fracture Fracture Total MIT (5-10 Ncm) MIT (higher than 10 Ncm) MIT (lower than 5 Ncm)


Data collection form XIV. Miscellaneous57

Entry Description and character of the information Funding source Describe the funding source and indicate if it is an implant company. Key conclusions of study authors Describe the main conclusions by the authors of the research study. Important comments by the study authors

Describe important comments by the authors of the research study.

References to other relevant studies

List references of other relevant studies that should be retrieved.

Key conclusions of the review authors

Describe the main conclusions by the review authors.

Major strength according to review authors

Describe the major points of strength of the research study according to the review authors.

Major weaknesses according to review authors

Describe the major points of weakness of the research study according to the review authors.

Contacting authors Indicate whether authors should be contacted and indicated what additional information is necessary.

Discuss with statistician Describe points that need to be discussed with a statistician. Points of interest Describe points of interest, e.g. the description of a new type of torque

sensor. Future studies Describe ideas for future studies suggested by the authors or reviewers. Other comments by review authors Any additional comment that has not been covered.

66

References

1. Higgins JPT, Green S, (editors). Cochrane Handbook for Systematic Reviews of Interventions. Chichester (UK): John Wiley & Sons, 2008.

2. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gøtzsche PC, Ioannidis JP, Clarke M, Devereaux PJ, Kleijnen J, Moher D. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. PLoS Med. 2009 Jul 21;6(7):e1000100. Epub 2009 Jul 21.

3. Moher D, Liberati A, Tetzlaff J, Altman DG; PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med. 2009 Jul 21;6(7):e1000097.

4. Deeks JJ, Higgins JPT, Altman DG. Chapter 9: Analysing data and undertaking meta-analyses. In: Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions. Chichester (UK): John Wiley & Sons, 2008.

5. Higgins JPT, Green S. Chapter 4: Guide to the contents of a Cochrane protocol and review. In: Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions. Chichester (UK): John Wiley & Sons, 2008.

6. Schünemann HJ, Oxman AD, Vist GE, Higgins JPT, Deeks JJ, Glasziou P, Guyatt GH. Chapter 12: Interpreting results and drawing conclusions. In: Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions. Chichester (UK): John Wiley & Sons, 2008.

7. Cha JY, Kil JK, Yoon TM, Hwang CJ. Miniscrew stability evaluated with computerized tomography scanning. Am J Orthod Dentofacial Orthop. 2010;137:73-9.

8. Chen Y, Kyung HM, Zhao WT, Yu WJ. Critical factors for the success of orthodontic mini-implants: a systematic review. Am J Orthod Dentofacial Orthop. 2009;135:284-91.

9. Cheng SJ, Tseng IY, Lee JJ, Kok SH. A prospective study of the risk factors associated with failure of mini-implants used for orthodontic anchorage. Int J Oral Maxillofac Implants. 2004;19:100-6.

10. Liou EJ, Pai BC, Lin JC. Do miniscrews remain stationary under orthodontic forces ? Am J Orthod Dentofacial Orthop. 2004;126:42-7.

11. Motoyoshi M, Hirabayashi M, Uemura M, Shimizu N. Recommended placement torque when tightening an orthodontic mini-implant. Clin Oral Implants Res. 2006;17:109-14.

12. Park HS, Jeong SH, Kwon OW. Factors affecting the clinical success of screw implants used as orthodontic anchorage. Am J Orthod Dentofacial Orthop. 2006;130:18-25.

13. Lee NK, Baek SH. Effects of the diameter and shape of orthodontic mini-implants on microdamage to the cortical bone. Am J Orthod Dentofacial Orthop. 2010;138:e1-8.

14. Meredith N. Assessment of implant stability as a prognostic determinant. Int J Prosthodont. 1998;11:491-501. 15. Wawrzinek C, Sommer T, Fischer-Brandies H. Microdamage in cortical bone due to the overtightening of

orthodontic microscrews. J Orofac Orthop. 2008;69:121-34. 16. Reynders R, Ronchi L, Bipat S. Mini-implants in orthodontics: a systematic review of the literature. Am J Orthod

Dentofacial Orthop. 2009;135:564.e1-19. 17. Mortensen MG, Buschang PH, Oliver DR, Kyung HM, Behrents RG. Stability of immediately loaded 3- and 6-mm

miniscrew implants in beagle dogs--a pilot study. Am J Orthod Dentofacial Orthop. 2009;136:251-9. 18. Wang YC, Liou EJ. Comparison of the loading behaviour of self-drilling and predrilled miniscrews throughout

orthodontic loading. Am J Orthod Dentofacial Orthop. 2008;133:38-43. 19. Gapski R, Wang HL, Mascarenhas P, Lang NP. Critical review of immediate implant loading. Clin Oral Implants Res

2003;14:515-27. 20. Kim SH, Lee SJ, Cho IS, Kim SK, Kim TW. Rotational resistance of surface-treated mini-implants. Angle Orthod.

2009;79:899-907. 21. Lim SA, Cha JY, Hwang CJ. Insertion torque of orthodontic miniscrews according to changes in shape, diameter

and length. Angle Orthod. 2008;78:234-40 22. Salmória KK, Tanaka OM, Guariza-Filho O, Camargo ES, de Souza LT, Maruo H. Insertional torque and axial pull-

out strength of mini-implants in mandibles of dogs. Am J Orthod Dentofacial Orthop. 2008;133:790.e15-22. 23. Ueda M, Matsuki M, Jacobsson M, Tjellström A. Relationship between insertion torque and removal torque

analyzed in fresh temporal bone. Int J Oral Maxillofac Implants. 1991;6:442-7. 24. Trisi P, Perfetti G, Baldoni E, Berardi D, Colagiovanni M, Scogna G. Implant micromotion is related to peak

insertion torque and bone density. Clin Oral Implants Res. 2009;20:467-71. 25. Ivanoff CJ, Sennerby L, Lekholm U. Influence of initial implant mobility on the integration of titanium implants.

An experimental study in rabbits. Clin Oral Implants Res. 1996;7:120-7. 26. Ottoni JM, Oliveira ZF, Mansini R, Cabral AM. Correlation between placement torque and survival of single tooth

implants. Int J Oral Maxillofac Implants. 2005;20:769-76. 27. Chaddad K, Ferreira AF, Geurs N, Reddy MS. Influence of surface characteristics on survival rates of mini-

implants. Angle Orthod. 2008;78:107-13. 28. Motoyoshi M, Matsuoka M, Shimizu N. Application of orthodontic mini-implants in adolescents. Int J Oral

Maxillofac Surg. 2007;36:695-9. 29. Motoyoshi M, Yoshida T, Ono A, Shimizu N. Effect of cortical bone thickness and implant placement torque on

stability of orthodontic mini-implants. Int J Oral Maxillofac Implants. 2007;22:779-84.











67

30. Motoyoshi M, Uemura M, Ono A, Okazaki K, Shigeeda T, Shimizu N. Factors affecting the long-term stability of orthodontic mini-implants. Am J Orthod Dentofacial Orthop. 2010;137:588.e1-5.

31. Cleek TM, Reynolds KJ, Hearn TC. Effect of screw torque level on cortical bone pull out strength. J Orthop Trauma. 2007;21:117-23.

32. Lawson KJ, Brems J. Effect of insertion torque on bone screw pullout strength. Orthopedics. 2001;24:451-4. 33. Chen Y, Shin HI, Kyung HM. Biomechanical and histological comparison of self-drilling and self-tapping

orthodontic microimplants in dogs. Am J Orthod Dentofacial Orthop. 2008;133:44-50. 34. Okazaki J, Komasa Y, Sakai D, Kamada A, Ikeo T, Toda I, Suwa F, Inoue M, Etoh T. A torque removal study on the

primary stability of orthodontic titanium screw mini-implants in the cortical bone of dog femurs. Int J Oral Maxillofac Surg. 2008;37:647-50.

35. Arismendi JA, Ocampo ZM, Morales M, Gonzalez FJ, Jaramillo PM, Sanchez A. Evaluation of stability of mini implants as bony anchorage for upper molar intrusion. Revista Facultad de Odontologia Universidad de Antioquia 2007;19:59-73.

36. Suzuki EY, Suzuki B. Placement and removal torque values of orthodontic miniscrew implants. Am J Orthod Dentofacial Orthop. 2011;139:669-78.

37. Reeves BC, Deeks JJ, Higgins JPT, Wells GA . Chapter 13: Including non-randomized studies. In: Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions. Chichester (UK): John Wiley & Sons, 2008.

38. OCEBM Levels of Evidence Working Group. “The Oxford 2011 Levels of Evidence”. Oxford Centre for Evidence-Based Medicine. [online] Available from: http://www.cebm.net/index.aspx?o=5653 (accessed September 9th 2012).

39. Pandis N. The evidence pyramid and introduction to randomized controlled trials. Am J Orthod Dentofacial Orthop. 2011;140:446-7.

40. Guyatt GH, Oxman AD, Vist GE, Kunz R, Falck-Ytter Y, Alonso-Coello P, Schünemann HJ. GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. Working Group. BMJ. 2008;336:924-6.

41. Guyatt G, Jaeschke R, Prasad K, Cook DJ. Chapter 14: Summarizing the evidence. In: Guyatt G, Rennie D, Meade MO, Cook DJ (editors). Users’ guides to the medical literature. New York:McGraw Hill Medical, 2008.

42. Luzi C, Verna C, Melsen B. A prospective clinical investigation of the failure rate of immediately loaded mini-implants used for orthodontic anchorage. Prog Orthod. 2007;8:192-201.

43. O’Connor D, Green S, Higgins JPT. Chapter 5: Defining the review question and developing criteria for including studies. In: Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions. Chichester (UK): John Wiley & Sons, 2008.

44. Lim HJ, Eun CS, Cho JH, Lee KH, Hwang HS. Factors associated with initial stability of miniscrews for orthodontic treatment. Am J Orthod Dentofacial Orthop. 2009;136:236-42.

45. Song YY, Cha JY, Hwang CJ. Mechanical characteristics of various orthodontic mini-screws in relation to artificial cortical bone thickness. Angle Orthod. 2007;77:979-85.

46. Wu J, Bai YX, Wang BK. Biomechanical and histomorphometric characterizations of osseointegration during mini-screw healing in rabbit tibiae. Angle Orthod. 2009;79:558-563.

47. Melzack R. The McGill Pain Questionnaire: Major properties and scoring methods. Pain. 1975; 1: 277-299. 48. Egger M, Jüni P, Bartlett C, Holenstein F, Sterne J. How important are comprehensive literature searches and the

assessment of trial quality in systematic reviews ? Empirical study. Health Technology Assessment 2003;7:1. 49. Hopewell S, McDonald S, Clarke M, Egger M. Grey literature in meta-analyses of randomized trials of health care

interventions. Cochrane Database Syst Rev. 2007 Apr 18;(2):MR000010. 50. McAuley L, Pham B, Tugwell P, Moher D. Does the inclusion of grey literature influence estimates of intervention

effectiveness reported in meta-analyses? Lancet. 2000 Oct 7;356(9237):1228-31. 51. Toma M, McAlister FA, Bialy L, Adams D, Vandermeer B, Armstrong PW. Transition from meeting abstract to full-

length journal article for randomized controlled trials. JAMA. 2006 Mar 15;295(11):1281-7. 52. Kanomi R. Mini-implant for orthodontic anchorage. J Clin Orthod. 1997;31(11):763-7. 53. Lefebvre C, Manheimer E, Glanville J. Chapter 6: Searching for studies. In: Higgins JPT, Green S (editors). Cochrane

Handbook for Systematic Reviews of Interventions. Chichester (UK): John Wiley & Sons, 2008. 54. Sterne JAC, Egger M, Moher D. Chapter 10: Addressing reporting biases. In: Higgins JPT, Green S (editors).

Cochrane Handbook for Systematic Reviews of Interventions. Chichester (UK): John Wiley & Sons, 2008. 55. Borenstein M, Hedges LV, Higgins JPT, Rothstein HR. Chapter 30: Publication bias. In: Borenstein M, Hedges LV,

Higgins JPT, Rothstein HR (editors). Introduction to Meta-Analysis. Chichester (UK): John Wiley & Sons, 2009. 56. Edwards P, Clarke M, DiGuiseppi C, Pratap S, Roberts I, Wentz R. Identification of randomized controlled trials in

systematic reviews: accuracy and reliability of screening records. Stat Med. 2002 Jun 15;21(11):1635-40. 57. Higgins JPT, Deeks JJ. Chapter 7: Selecting studies and collecting data. In: Higgins JPT, Green S (editors).

Cochrane Handbook for Systematic Reviews of Interventions. Chichester (UK): John Wiley & Sons, 2008. 58. GRADE Working Group 2004. GRADE Working Group. Grading quality of evidence and strength of

recommendations. BMJ. 2004;328:1490-4. 59. Gøtzsche PC. Multiple publication of reports of drug trials. Eur J Clin Pharmacol. 1989;36(5):429-32. 60. Schulz KF, Altman DG, Moher D, for the CONSORT Group. CONSORT 2010 Statement: updated guidelines for

reporting parallel group randomised trials. BMJ. 2010;340:c332





http://www.cebm.net/index.aspx?o=5653



http://www.ncbi.nlm.nih.gov/pubmed?term=%22Hopewell%20S%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/pubmed?term=%22McDonald%20S%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/pubmed?term=%22Clarke%20M%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/pubmed?term=%22Egger%20M%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/pubmed?term=McAuley%20L%5BAuthor%5D&cauthor=true&cauthor_uid=11072941

http://www.ncbi.nlm.nih.gov/pubmed?term=Pham%20B%5BAuthor%5D&cauthor=true&cauthor_uid=11072941

http://www.ncbi.nlm.nih.gov/pubmed?term=Tugwell%20P%5BAuthor%5D&cauthor=true&cauthor_uid=11072941

http://www.ncbi.nlm.nih.gov/pubmed?term=Moher%20D%5BAuthor%5D&cauthor=true&cauthor_uid=11072941

http://www.ncbi.nlm.nih.gov/pubmed?term=McAuley%20L%20Pham%20B%20Tugwell%20P

http://www.ncbi.nlm.nih.gov/pubmed?term=Toma%20M%5BAuthor%5D&cauthor=true&cauthor_uid=16537738

http://www.ncbi.nlm.nih.gov/pubmed?term=McAlister%20FA%5BAuthor%5D&cauthor=true&cauthor_uid=16537738

http://www.ncbi.nlm.nih.gov/pubmed?term=Bialy%20L%5BAuthor%5D&cauthor=true&cauthor_uid=16537738

http://www.ncbi.nlm.nih.gov/pubmed?term=Adams%20D%5BAuthor%5D&cauthor=true&cauthor_uid=16537738

http://www.ncbi.nlm.nih.gov/pubmed?term=Vandermeer%20B%5BAuthor%5D&cauthor=true&cauthor_uid=16537738

http://www.ncbi.nlm.nih.gov/pubmed?term=Armstrong%20PW%5BAuthor%5D&cauthor=true&cauthor_uid=16537738

http://www.ncbi.nlm.nih.gov/pubmed?term=toma%20m%20mcalister%20fa%20bialy%20l%20adams%20d

http://www.ncbi.nlm.nih.gov/pubmed?term=Kanomi%20R%20Mini-implant%20for%20orthodontic%20anchorage.%20Journal%20of%20Clinical%20Orthodontics%2031%3A%20763%E2%80%937.



http://www.ncbi.nlm.nih.gov/pubmed?term=G%C3%B8tzsche%20PC%5BAuthor%5D&cauthor=true&cauthor_uid=2666138


68

61. Higgins JPT, Altman DG. Chapter 8: Assessing risk of bias in included studies. In: Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions. Chichester (UK): John Wiley & Sons, 2008.

62. Higgins JPT, White IR, Wood AM. Imputation methods for missing outcome data in meta-analysis of clinical trials. Clinical Trials 2008;5:225-239.

63. Higgins JPT, Deeks JJ, Altman DG. Chapter 16: Special topics in statistics. In: Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions. Chichester (UK): John Wiley & Sons, 2008.

64. West SL, Gartlehner G, Mansfield AJ, Poole C, Tant E, et al. Comparative Effectiveness Review Methods: Clinical Heterogeneity. Agency for Healthcare Research and Quality. Publication No. 10-EHC070-EF. September 2010.

65. Rothstein HR, Sutton AJ, Borenstein M (editors). Publication Bias in Meta-Analysis: Prevention, Assessment, and Adjustments. Chichester (UK): John Wiley & Sons, 2009.

66. Higgins JPT, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ. 2003;327:557-560.

67. Schroll JB, Moustgaard R, Gøtzsche PC. Dealing with substantial heterogeneity in Cochrane reviews. Cross-sectional study. BMC Med Res Methodol. 2011 Feb 24;11:22.

68. Sutton AJ, Abrams KR, Jones DR, Sheldon TA, Song F. Chapter 3: Assessing between study heterogeneity. In: Sutton AJ, Abrams KR, Jones DR, Sheldon TA, Song F (editors). Methods for Meta-Analysis in Medical Research. Chichester (UK): John Wiley & Sons, 2000.

69. O'Rourke K, Detsky AS. Meta-analysis in medical research: strong encouragement for higher quality in individual research efforts. J Clin Epidemiol. 1989;42(10):1021-4

70. Borenstein M, Hedges LV, Higgins JPT, Rothstein HR. Chapter 16: Identifying and quantifying heterogeneity. In: Borenstein M, Hedges LV, Higgins JPT, Rothstein HR (editors). Introduction to Meta-Analysis. Chichester (UK): John Wiley & Sons, 2009.

71. Higgins JPT, Thompson SG. Quantifying heterogeneity in a meta-analysis. Statistics in Medicine. 2002;21:1539-1558.

72. Higgins JP. Commentary: Heterogeneity in meta-analysis should be expected and appropriately quantified. Int J Epidemiol. 2008;37(5):1158-60.

73. Rücker G, Schwarzer G, Carpenter JR, Schumacher M. Undue reliance on I(2) in assessing heterogeneity may mislead. BMC Med Res Methodol. 2008 Nov 27;8:79.

74. Borenstein M, Hedges LV, Higgins JPT, Rothstein HR. Introduction to Meta-Analysis. Borenstein M, Hedges LV, Higgins JPT, Rothstein HR (editors). Chichester (UK): John Wiley & Sons, 2009.

75. Borenstein M, Hedges LV, Higgins JPT, Rothstein HR. Chapter 40: When does it make sense to perform a meta-analysis? In: Borenstein M, Hedges LV, Higgins JPT, Rothstein HR (editors). Introduction to Meta-Analysis. Chichester (UK): John Wiley & Sons, 2009.

76. Borenstein M, Hedges LV, Higgins JPT, Rothstein HR. Chapter 1. How a meta-analysis works. In: Borenstein M, Hedges LV, Higgins JPT, Rothstein HR (editors). Introduction to Meta-Analysis. Chichester (UK): John Wiley & Sons, 2009.

77. Email correspondence between Mike Clarke and Reinhart Meursinge Reynders on the definition of Publication Bias. August 11 2012.

78. Koletsi D, Karagianni A, Pandis N, Makou M, Polychronopoulou A, Eliades T. Are studies reporting significant results more likely to be published ? Am J Orthod Dentofacial Orthop 2009;136:632.e1-5.

79. Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ. 1997 Sep 13;315(7109):629-34.

80. Schünemann HJ, Oxman AD, Higgins JPT, Vist GE, Glasziou P, Guyatt GH. Chapter 11: Presenting results and ‘Summary of findings’ tables. In: Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions. Chichester (UK): John Wiley & Sons, 2008.

http://effectivehealthcare.ahrq.gov/

http://effectivehealthcare.ahrq.gov/

http://www.ncbi.nlm.nih.gov/pubmed?term=Schroll%20JB%5BAuthor%5D&cauthor=true&cauthor_uid=21349195

http://www.ncbi.nlm.nih.gov/pubmed?term=Moustgaard%20R%5BAuthor%5D&cauthor=true&cauthor_uid=21349195

http://www.ncbi.nlm.nih.gov/pubmed?term=G%C3%B8tzsche%20PC%5BAuthor%5D&cauthor=true&cauthor_uid=21349195

http://www.ncbi.nlm.nih.gov/pubmed?term=Dealing%20with%20substantial%20heterogeneity%20in%20Cochrane%20reviews.%20Cross-sectional%20study

http://www.ncbi.nlm.nih.gov/pubmed?term=O'Rourke%20K%5BAuthor%5D&cauthor=true&cauthor_uid=2809651

http://www.ncbi.nlm.nih.gov/pubmed?term=Detsky%20AS%5BAuthor%5D&cauthor=true&cauthor_uid=2809651

http://www.ncbi.nlm.nih.gov/pubmed?term=Meta-analysis%20in%20medical%20research%3A%20strong%20encouragement%20for%20higher%20quality%20in%20individual%20research%20efforts

http://www.ncbi.nlm.nih.gov/pubmed?term=Higgins%20JP%5BAuthor%5D&cauthor=true&cauthor_uid=18832388

http://www.ncbi.nlm.nih.gov/pubmed?term=International%20Journal%20of%20Epidemiology%202008%3B37%3A1158%E2%80%931160

http://www.ncbi.nlm.nih.gov/pubmed?term=International%20Journal%20of%20Epidemiology%202008%3B37%3A1158%E2%80%931160

http://www.ncbi.nlm.nih.gov/pubmed?term=R%C3%BCcker%20G%5BAuthor%5D&cauthor=true&cauthor_uid=19036172

http://www.ncbi.nlm.nih.gov/pubmed?term=Schwarzer%20G%5BAuthor%5D&cauthor=true&cauthor_uid=19036172

http://www.ncbi.nlm.nih.gov/pubmed?term=Carpenter%20JR%5BAuthor%5D&cauthor=true&cauthor_uid=19036172

http://www.ncbi.nlm.nih.gov/pubmed?term=Schumacher%20M%5BAuthor%5D&cauthor=true&cauthor_uid=19036172

http://www.ncbi.nlm.nih.gov/pubmed?term=undue%20reliance%20on%20in%20assessing%20heterogeneity

http://www.ncbi.nlm.nih.gov/pubmed?term=Egger%20M%5BAuthor%5D&cauthor=true&cauthor_uid=9310563

http://www.ncbi.nlm.nih.gov/pubmed?term=Davey%20Smith%20G%5BAuthor%5D&cauthor=true&cauthor_uid=9310563

http://www.ncbi.nlm.nih.gov/pubmed?term=Schneider%20M%5BAuthor%5D&cauthor=true&cauthor_uid=9310563

http://www.ncbi.nlm.nih.gov/pubmed?term=Minder%20C%5BAuthor%5D&cauthor=true&cauthor_uid=9310563


69

Protocol 4

Insertion torque recordings for the diagnosis of contact between orthodontic mini-implants and

dental roots: protocol for a systematic review This is the non-edited protocol for the manuscript: Meursinge Reynders RA, Ladu L, Ronchi L, Di Girolamo N, De Lange J, Roberts N, Plüddemann A. Insertion torque recordings for the diagnosis of contact between orthodontic mini-implants and dental roots: A systematic review. Syst Rev. 2016 Mar 31;5:50. This non-edited protocol was published as: Meursinge Reynders RA, Ladu L, Ronchi L, Di Girolamo N, De Lange J, Roberts N, Plüddemann A. Insertion torque recordings for the diagnosis of contact between orthodontic mini-implants and dental roots: protocol for a systematic review. Syst Rev. 2015 Apr 2;4(1):39.

70

Abstract Background: Hitting a dental root during the insertion of orthodontic mini-implants (OMIs) is a common adverse effect of this intervention. This condition can permanently damage these structures and can cause implant instability. Increased torque levels (index test) recorded during the insertion of OMIs may provide a more accurate and immediate diagnosis of implant-root contact (target condition) than radiographic imaging (reference standard). An accurate index test could reduce or eliminate X-ray exposure. These issues, the common use of OMIs, the high prevalence of the target condition, and because most OMIs are placed between roots warrant a systematic review. We will assess 1) the diagnostic accuracy and the adverse effects of the index test, 2) whether OMIs with root contact have higher insertion torque values than those without, and 3) whether intermediate torque values have clinical diagnostic utility. Methods: The Preferred Reporting Items for Systematic review and Meta-Analysis Protocols (PRISMA-P) 2015 statement was used as a the guideline for reporting this protocol. Inserting implants deliberately into dental roots of human participants would not be approved by ethical review boards and adverse effects of interventions are generally underreported. We will therefore apply broad spectrum eligibility criteria, which will include clinical, animal and cadaver models. Not including these models could slow down knowledge translation. Both randomized and non-randomized research studies will be included. Comparisons of interest and subgroups are pre-specified. We will conduct searches in MEDLINE and more than 40 other electronic databases. We will search the grey literature and reference lists and hand-search ten journals. All methodological procedures will be conducted by three reviewers. Study selection, data extraction and analyses, and protocols for contacting authors and resolving conflicts between reviewers are described. Designed specific risk of bias tools will be tailored to the research question. Different research models will be analysed separately. Parameters for exploring statistical heterogeneity and conducting meta-analyses are pre-specified. The quality of evidence for outcomes will be assessed through the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach. Discussion: The findings of this systematic review will be useful for patients, clinicians, researchers, guideline developers, policymakers, and surgical companies.

Keywords Diagnostic test accuracy, Implant, Screw, Root contact, Root proximity, Insertion torque, Orthodontics, Systematic review

71

Background Orthodontic mini-implants (OMIs) are used to provide anchorage during orthodontic tooth movement. Contact between these devices and dental roots during implant insertion is a common problem, because inter-radicular spaces are narrow1-6. Such contacts have been associated with damage of these structures and increased implant failure rates7-9. Because most OMIs are placed between dental roots and the prevalence of implant-root contact is high, an accurate test for the diagnosis of this target condition is indicated3,5,10. Target condition being diagnosed To control the reciprocal forces of tooth movement, orthodontists need some form of anchorage. It is usually obtained by connecting these forces to groups of teeth in the same or the opposing jaw or through the use of intra-or extra-oral removable appliances. However, these treatment mechanics may still cause anchorage loss, have a limited area of application, or depend on patient collaboration11. OMIs are not conditioned by most of these limitations, but they need to be inserted surgically. A 2014 survey of orthodontists in the USA by the Journal of Clinical Orthodontics identified that 90% of OMIs are inserted at inter-radicular sites10. Contact between OMIs and dental roots is one of the risks of this intervention and is the target condition of this systematic review. This condition can cause extensive damage to the roots7,12. Studies in maxillofacial surgery have reported similar problems with intermaxillary fixation screws13-15. The quality of the healing of the root injury as a result of the target condition varies and damage involving the dental pulp is less likely to repair completely7,12,16. Additional root damage can occur during orthodontic treatment, because implants are not stable and they can migrate towards dental roots17-19. This issue is particularly important, because contact between implants and the periodontal ligament was identified in 65.2% of consecutively inserted OMIs20. Close contact between OMIs and dental roots has also been associated with increased failure rates of these devices3,9,21-23. A recent systematic review identified three times higher failure rates in OMIs with root contact compared with those placed away from adjacent roots8. The diameters of the most commonly used OMIs vary between 1.2 and 2 mm8. Only little space for error is possible, because inter-radicular distances of at least 3 mm have been recommended for safe placement of OMIs. Such dimensions are only available in limited areas of the dental arches (Figure 1) 4,6. Specific surgical and radiographic positioning techniques have been developed to avoid root contact during implant insertion24-27. These methods are accurate, but they require additional radiographs and complex and expensive surgical guides3,28,29.

72

Figure 1. Inter-radicular distances in the maxillary arch and 1.5 mm (diameter) orthodontic mini-implants* *Quattro implants PSM Medical Solutions; Tuttlingen, Germany Various studies have assessed the prevalence of the target condition. Cho et al.2 scored the target condition in 21.3% of the implant insertions for inexperienced operators and in 13.5% for experienced operators. Kim et al.3 scored 30% for implant-root contact, and Motoyoshi et al.5 identified a prevalence of 17.1% for screws that were inserted with the self-drilling technique and 20.5% for those that were pre-drilled. This high prevalence of the target condition strengthens the importance of accurately diagnosing implant-root contact. Reference standard In current practice, X-rays are used to measure interradicular distances prior to implant placement (Figure 2). Additional radiographs are taken to diagnose implant-root contact during implant insertion and at the completion of this procedure. The latter assessment is the reference standard of the current diagnostic pathway. No other reference standards are currently used to diagnose the target condition. Studies that have used either two-dimensional or three-dimensional images are both eligible for our clinical questions, but the latter type is more accurate, because it also provides information on the third dimension30,31. Although X-rays are often considered the ‘gold’ standard of discriminative power, they can still produce false positive and false negative outcomes32,33. However, the main disadvantage of this method is the exposure to radiation. This is particularly a problem for three-dimensional radiographs, which produce higher radiation exposure than two-dimensional imaging34.

73

Figure 2. Current diagnostic pathway for assessing implant-root contact The index test The recording of torque values during the insertion of OMIs does not have these shortcomings and is the index test for our clinical questions. The American Society for Testing and Materials (ASTM International) defines insertion torque as ‘the amount of torque required to overcome the frictional force between the screw and the material used for testing while driving the screw into the material’35. This index test was chosen after having addressed the following preliminary questions36:

1. Is the prevalence of the target condition sufficiently high to order the diagnostic test? 2. Is there an effective treatment for the target condition? 3. Could the test introduce a change in the management strategies that is beneficial to the

patients? 4. Are patients undergoing the new test expected to be better off than those who do not from a

health perspective? 5. Are the costs and cost-effectiveness analyses of the new test expected to justify

implementation? All five questions were answered with a ‘Yes’. Advantages of the index test could include the following: (1) increasing the certainty of diagnosing root contact. In various animal models, increased insertion torque values have been associated with the target condition16,37,38; (2) reducing or eliminating the use

74

of the current reference standard and therefore the exposure to X-rays; and (3) contributing to the decision-making process, because the continuous recordings of torque during the entire insertion process could inform the clinician at which specific time point the OMI touches the root. A sudden steep increase in torque values could indicate root contact37. Potential disadvantages associated with the index test could include increased costs, a steep learning curve, a longer diagnostic pathway, pain and discomfort, and unforeseen adverse effects. However, most of these disadvantages are not applicable, because the index test does not: (1) prolong the procedure; (2) require additional learning (3) or introduce pain and discomfort, because torque values are measured during the current standard insertion process with sensors that are built into the screwdrivers. The index test will probably also reduce costs in the long run, because taking less or no radiographs will shorten the duration and lower the costs of the current diagnostic pathway. Furthermore, the purchasing price of the diagnostic instrument is relatively low (±€1,000), and this device can also be used for other diagnostic purposes, for example, measuring implant stability9,39. The test under review could be a candidate for a role as a ‘replacement’, a ‘triage’, or as an ‘add-on’ test40,41. It could replace the reference standard if the index test is, for example, more accurate, cheaper, faster, and causing less adverse effects (for example, less exposure to radiation) than the reference standard. Both intermediate and final radiographs can then be avoided. A role as a triage test (minimize false negatives) to minimize the use of the more expensive and invasive reference standard could be indicated, for example, in young cancer patients that have undergone radiotherapy. The index test could also be used as an add-on test to improve accuracy. The importance for conducting the index test was explained in the previous paragraphs. Scoping searches of the literature did not identify any review that assessed the clinical utility of the index test. Systematic reviews that critically appraise the literature and address specific research questions on this topic are therefore indicated.

Objectives Primary objectives For the primary objectives of this systematic review, the diagnostic accuracy question is formulated according to the participants, index test, reference standard, target condition (PIRT) mnemonic42,43:

• In OMIs (participants or problem), what is the accuracy of the level of insertion torque values (index test) compared to radiography (reference standard) to distinguish those with and without implant-root contact (target condition).

Secondary objectives For the secondary objectives, we formulated the following questions:

• ‘Do OMIs with root contact have higher insertion torque values than those without this target condition?’ ‘We will also assess whether intermediate recordings of insertion torque values have clinical utility for the diagnosis of the target condition’.

75

Methods/design The guideline for conducting systematic reviews of diagnostic accuracy of the Diagnostic Test Accuracy Working Group of the Cochrane Collaboration and the Cochrane Handbook for Systematic reviews of Interventions will be adopted to address our research questions44,45. The Preferred Reporting Items for Systematic review and Meta-Analysis Protocols (PRISMA-P) 2015 statement is used as a the guideline for reporting this protocol46,47. Differences in methodology between research questions are explained for each of these items. Changes in the methods during the conduct of this systematic review from those outlined in this protocol will be fully reported.

Eligibility criteria Our eligibility criteria will be adapted to the particular character of our research questions, because we do not expect to find clinical studies in which implants are deliberately inserted into dental roots. Because ethical review boards do not approve such procedures, we will also include in vivo animal studies and cadaver models. Including these experimental models is important because: 1) animal and cadaver studies might provide additional information on the usefulness of conducting the index test; 2) they could provide information on how to design future research studies on our clinical question; 3) considering outcomes from animal studies avoids wasting valuable research information, financial resources, and duplication48,49. The importance of these issues was further stressed by Iain Chalmers, one of the founders of the Cochrane Collaboration, in a recent international symposium on systematic reviews in laboratory animal science50; and 4) not considering these studies would risk that knowledge creation on this topic would remain at a standstill. These issues are further strengthened in the context of the high prevalence of the target condition, the risk of biologic damage of the interventional procedure, the instability of implants with the target condition, and the underreporting of adverse effects of interventions7-9,20,51,52. Actually, one could reason that it would be unethical not to include experimental studies when only limited numbers of clinical studies will be identified. This systematic review is not registered in the PROSPERO database, because its inclusion criteria cover only studies on human participants53. To avoid inappropriate exclusion of relevant articles, we will aim for more broad-scope inclusion criteria that are sufficiently specific and still cover all research objectives54. Study designs

• Our preferred research design will be studies that randomize participants to either creating or not creating implant-root contact. We do not expect to find randomized studies because (1) this research design is rare in diagnostic accuracy studies and (2) of the ethical reasons outlined in the previous section.

• Non-randomized studies (NRS), that is, cohort (prospective and retrospective) and case-control study designs, that used the index test alone or in combination with the reference standard and compared outcomes on participants with or without the target condition are eligible55,56. The decision to include NRS in systematic reviews is based on the justifications presented by the Non-Randomized Studies Methods Group of the Cochrane Collaboration and the Oxford 2011 Levels of Evidence Working Group56-58. These rationales include the following: (1) high-quality NRS could produce less biased outcomes compared to low-quality randomized controlled trials (RCTs); (2) RCTs are not always needed; (3) ethical reasons could make RCTs unfeasible; (4) NRS could reveal the deficiencies of the current literature and show the need of additional research; (5) findings of these NRS could help in designing such studies; (6) NRS could reveal unexpected, rare, or long-term harmful effects of interventions56,58,59; and (7)

76

large magnitude of treatment effects, the presence of a dose response gradient, or plausible residual confounding could upgrade the validity of NRS60,61.

• Cadaver and animal studies will also be included. The choice for including such study models was explained in the previous section.

• Non-primary studies such as ‘editorials’, ‘view point publications’, ‘case reports’, ‘reviews’, and studies that used computer simulated models, for example, finite element analyses, are excluded.

• Eligibility will not be based on ‘design labels’ only, because studies can be poorly indexed or labels can be used inconsistently by authors56,62.

• Outcomes of different study designs are assessed separately. Participants

• Adults and adolescents of 12 years and older of either sex, in any ethnicity, setting, or socio-economic group, no medical or dental history, in need of stationary anchorage during orthodontic treatment with fixed appliances are our preferred participants to include.

• For the animal studies, only monkeys and dogs are eligible, because these animal models are mostly used in orthodontic research and their dentitions closely resemble those of humans. For the cadaver studies, both human and animal models will be eligible. For each research model, outcomes will be assessed separately.

• Participants that had undergone a previous surgical intervention or the index test in the same area, for example, placement of an additional OMI or re-implantation of OMIs in the same site, are excluded, because such procedures could affect outcomes.

Interventions

• The target condition of interest is contact between OMIs and dental roots, which is currently diagnosed with the reference standard. This target condition refers to single or multiple implant-root contacts with or without root penetration5,37,38.

• The recordings of insertion torque values during the insertion of OMIs in inter-radicular areas of the maxillary and mandibular alveolar bone will be the index test under investigation.

Outcomes

• Insertion torque values for OMIs with and without root contact will be eligible for both primary and secondary research questions. These outcomes are collected as reported in the original studies47. Only torque units that are convertible to Newton centimeter (Ncm) are eligible.

• Recordings with either mechanical or digital torque drivers are eligible for the primary research question. We decided to include both types of index tests, because the Diagnostic Test Accuracy Working Group of the Cochrane Collaboration recommends avoiding index test criteria that are too narrow54. Subsequent analyses could then assess differences between these subgroups.

• For the secondary research objective, we will also assess whether sudden steep increases in torque values during the implant insertion process were identified. Such an outcome could indicate implant-root contact. Only recordings with digital torque sensors are eligible, because mechanical sensors cannot record continuous torque values and can therefore not assess how torque measures change during implant insertion.

• Eligibility will not be influenced by the outcome of the test, that is, negative outcomes are considered as important as positive outcomes.

• Both two- and three-dimensional radiographs are eligible as the reference test for the primary and secondary research questions.

• For the secondary research questions, histology will also be eligible as the reference standard. • Because the degree of error between reference standards might differ, their outcomes will be

assessed separately in subgroups.

77

Timing

• Torque recordings at various time points are eligible. Variations in these domains between studies will be assessed in subgroup analyses.

• The time points of recording the reference standards have to be identical as those recorded for the index test.

Setting We will not apply restrictions by type of setting. Participants treated in either teaching or non-teaching settings are therefore eligible. Language No language restrictions will be applied and pertinent non-English articles will be translated.

Information sources Electronic searches Eligible studies will be searched in the period from 1 January 1997, the year of the introduction of orthodontic mini implants in orthodontics, onwards63. The following protocol will be applied to find eligible studies:

• General and subject-specific electronic databases will be consulted from PubMed (MEDLINE), Google Scholar Beta, Embase (Ovid), Science Direct, and Cochrane Central Register of Controlled Trials (CENTRAL)62,64,65.

• Additional studies will be searched through the ‘Related Articles’ feature in PubMed. • Databases for diagnostic studies: TRIP Database, NHS Evidence, and SUMSearch2 will also be

searched. • The following citation indexes will be searched: Science Citation Index, Scopus, and Web of

Science62,64. • The following national and regional databases will also be searched: African Index Medicus,

African Journals online (AJOL), Australasian Medical Index, Index Medicus for the Eastern Mediterranean Region, IndMED, KoreaMed, LILACS, Index Medicus for the South-East Asia Region (IMSEAR), and Western Pacific Region Index Medicus (WPRIM)62,64.

Searching other resources Grey literature Eligible reports are also searched in the grey literature, for example, research registers, conference proceedings, and university dissertations64-66. The following grey databases will be searched:

• General databases: Google Scholar Beta, Open Grey, The Health Management Information Consortium (HMIC), and The National Technical Information Service (NTIS)62.

• Dissertations and databases of theses: ProQuest Dissertations & Databases, Index to Theses in Great Britain and Ireland, and DissOnline62.

• Conference abstracts or proceedings: Conference Proceedings Citation Index (Web of Science), BIOSIS Citation Index, Meeting Abstracts, and ISI Proceedings62.

• Review databases: Meta-analyses van Diagnostisch Onderzoek (MEDION), Database of Abstracts of Reviews of Effects (DARE), Health Technology Assessment database (HTA), and Turning Research into Practice (TRIP)62.

• MEDLINE, Embase, and TRIP are searched for guidelines. In addition, evidence-based guidelines of the following organizations are consulted: Australian National Health and Medical Research Council, Canadian Medical Association, National Guideline Clearinghouse, National Library of Guidelines, New Zealand Guidelines Group, and NICE Clinical Guidelines62.

78

• Citations alerts in PubMed and Ovid62. Handsearching

• American Journal of Orthodontics and Dentofacial Orthopedics, Angle Orthodontist, Australian Journal of Orthodontics, European Journal of Orthodontics, International Journal of Adult Orthodontics and Orthognathic surgery, Journal of Clinical Orthodontics, Journal of Orthodontics, Journal of the World Federation of Orthodontics, Orthodontics, Orthodontics & Craniofacial Research, Progress in Orthodontics, and Seminars in Orthodontics.

Reference lists

• Each selected paper, all review articles, and guidelines will be screened manually for references of relevant articles that possibly are not identified in the searches of the electronic databases62,64.

Correspondence

• Subject specialists, authors of the selected articles, and researchers and manufacturers involved in our topic of interest will be contacted to identify unpublished or ongoing studies62,67.

Search strategy

• A librarian (NR) specialized in computerized searches of healthcare publications will assist with the development of the search strategy.

• A detailed protocol for developing this search strategy is presented in Additional file 162,64,68-71. • Pertinent search terms are summarized in Table 1. • Search strategies will be pilot tested for each database and subsequently fine-tuned62.

Examples are presented for the search strategy of MEDLINE and Google Scholar in Table 264,70. • We will list the search strategies for each database in a table together with the total number

of records retrieved and the search dates.

79

Study records Data management To reduce the risk of inter-examiner disagreement, we will adopt the protocol presented in the Cochrane Handbook for Systematic Reviews of Interventions72. A sample of articles will be first pilot tested to refine and clarify the eligibility criteria and to apply them consistently72. These procedures will be conducted by three review authors (RMR, LL, and LR). Selection process

• Three topic experts (RMR, LL, and LR) will independently select the studies. • Titles and abstracts will be screened for eligibility, and the full texts of potentially relevant

articles will be retrieved and subsequently reviewed. To avoid inappropriate exclusion, ambiguous articles will also be read.

• Unpublished research studies, for example, those extracted from the grey literature, which contain sufficient data to permit peer-reviewing will be reviewed independently by the topic experts. When data of unpublished studied are insufficient for adequate peer-reviewing, authors of these papers will be contacted. Our protocol for contacting authors is described in Additional file 273,74. When no additional data will be retrieved, outcomes of such reports are only used in the discussion to put effect estimates of eligible studies in perspective.

• Authors, suspect of multiple publications of the same research study will also be contacted (Additional file 2). Suspicion of multiple publications is based on the following characteristics: (1) studies with a retrospective design with similar methodology; (2) same authors in similar research studies; and (3) publication of similar findings in different journals within a short time span.

• In the case of disagreement on the eligibility of an article, the review authors will discuss the selection procedures, reread the paper, and if necessary, contact its authors67 (Additional file 2).

• A PRISMA flow diagram will illustrate the selection procedures and excluded articles will be presented in a table together with the rationale for their exclusion67,75.

80

Data collection process Data collection forms (see ‘Data items’ subsection) will be first pilot tested on their validity and subsequently fine-tuned. These procedures are also used as calibration exercises and are conducted independently by three experienced systematic reviewers (RMR, LL, and LR). Disagreements will be resolved through discussions. An arbitrator (NDG) will be consulted to adjudicate remaining disagreements. Data items

• The Standards for the Reporting of Diagnostic accuracy studies (STARD) checklist was consulted for the development of data extraction forms76,77. Data collection forms of previous systematic reviews on OMIs were also checked for pertinent items73,78,79. All data collection forms are tailored to our specific research questions.

• Pilot tested collection forms are presented in Additional file 3 and address both our primary and secondary research questions. These forms are prepared in a sequence that facilitates the application of the Quality Assessment of Diagnostic Accuracy Studies (QUADAS)-2 tool80. This instrument assesses risk of bias and concerns of applicability of outcomes of individual studies in diagnostic accuracy studies. If modifications of these forms will be necessary during the review process, we will report these changes and will explain them.

• Flow diagrams of participants will be created for each individual study. • Data from multiple publications of the same study will be first extracted from each article.

These data will be subsequently analysed for potential overlap. • All data extraction procedures will be conducted independently by three topic experts (RMR,

LL, and LR). • Disagreement between reviewers on extracted data will be resolved through discussions and

rereading. If necessary, authors will be contacted for clarification (Additional file 2). • An arbitrator (NDG) will be consulted to resolve remaining disagreements. • Persisting disagreements between reviewers will be reported. • We will apply the protocol presented by the Cochrane Handbook for Systematic Reviews of

Interventions for dealing with missing data, for example, missing outcomes, summary data, individuals, or study-level characteristics81. We will first apply our protocol for contacting authors (Additional file 2). When authors do not reply or are unable to provide us with this information, we will assess whether data were missing at random or not. The rationale and consequences for all assumptions and methods for dealing with missing data will be addressed in the discussion81. A statistician will be consulted for selecting an appropriate statistical model. All imputation techniques are avoided whenever possible. Sensitivity analyses will be conducted to assess how reasonable changes in assumptions affect results81.

Outcomes and prioritization To establish a threshold for test positivity, we conducted scoping searches of the literature. We identified one clinical and one animal study during these initial searches5,16. Insertion torque values of OMIs with root contact in self-drilling groups increased respectively 22.5% in human participants and 113% in adult beagles compared with implants without this target condition5,16. Based on these findings, we will define for our primary outcomes a hypothetical maximum insertion torque increase of 25% or more as a positive result of the index test and values inferior to this threshold as a negative outcome. Maximum insertion torque values for OMIs with and without root contact will therefore be recorded for the primary outcomes. Differences in the type of target condition, for example, with or without root penetration, and different time points and insertion depths for measuring these outcomes will be subdivided and assessed separately. The difference between maximum insertion torque recordings will be calculated for the secondary outcomes. We will also record whether sudden

81

steep increases in torque values were identified during the implant insertion process. Outcomes measures will be recorded in the original format as defined by the authors of the selected studies. These measures will be transformed to the effect estimate of this systematic review, that is, Ncm, after the completion of all data extraction procedures72.

Risk of bias individual studies For the primary research question, we will adopt the QUADAS-2 tool80. The content and the rating guidelines of the QUADAS-2 tool are tailored to our review question according to the protocol depicted in a flow diagram (Figure 3)82. If indicated, certain signaling questions were added and others were omitted. To standardize the application of assessment criteria, clear definitions for each criterion were established prior to conducting the review83. The revised tool was subsequently pilot tested on a small number of eligible studies. This process was conducted by the three topic experts (RMR, LL, and LR). A methodologist, AP, was consulted for assistance and to serve as a referee. The tailored and pilot tested QUADAS-2 tool is presented in Additional file 4. This tool will be subsequently applied to all eligible studies. A table will be used to display the QUADAS-2 assessments (Additional file 5)80,82. QUADAS-2 scores are not used to generate a summary ‘quality score’, because such scores can be problematic82,84,85.

82

For the secondary research questions, we will use the ‘Risk of bias tool’ of the Cochrane collaboration, because this instrument has been designed specifically for interventional questions86. Both the Cochrane and the QUADAS-2 tools use domain-based evaluations of systematic error, but their biases are defined differently and cover different parts of the research phases. The differences between these tools are illustrated in Figures 4 and 582,86. Figure 4 also shows how the various phases of the research study are covered by specific types of bias. Risk of bias will be scored as ‘High’, ‘Low’, or ‘Unclear’. The Cochrane Handbook for Systematic Reviews of Interventions assigns this latter judgement if: 1) ‘insufficient detail is reported of what happened in the study’; 2) ‘what happened in the study is known, but the risk of bias is unknown’; 3) ‘an entry is not relevant to the study at hand’. The assessments of risk of bias will be presented in a ‘Risk of bias summary figure’86.

83

84

The scoring for both the QUADAS-2 and the Cochrane Risk of Bias tools will be conducted by three review authors (RMR, LL, and LR). Disagreements will be resolved through discussions. In the case of persisting disagreements, the methodologist, AP, will serve as an arbitrator. If necessary, authors of selected studies will also be contacted for clarification87. Risk of bias scores will be used to assess their potential influence on the overall outcomes of the review and on the confidence in the cumulative estimate, for example, for the Grading of Recommendations Assessment, Development and Evaluation (GRADE) judgments47. Risk of bias scores will also be consulted during the decision-making whether to undertake a meta-analysis or not (see data synthesis).

Data synthesis Qualitative synthesis We will present and explain the characteristics and outcomes of the eligible studies in a narrative synthesis. This qualitative summary will be conducted whether or not a quantitative data synthesis will be considered appropriate. Characteristics of included studies will be presented first and outcomes are subsequently listed according to the order of our research questions47. This narrative summary will be completed with a series of tables: (1) characteristics of participants; (2) characteristics of selection procedures; (3) characteristics of the target condition, index test, and reference standard; (4) tabular presentation of the QUADAS-2 results; (5) tabular presentation of the Cochrane risk of bias assessment; (6) outcomes of the diagnostic accuracy tests; (7) insertion torque values with or without implant-root contact; (8) adverse effects of interventions; and (9) information obtained from contacted authors. Data analysis For the primary outcomes of each selected study, we will score our binary outcomes in 2 × 2 contingency tables, which present the outcomes of the test results and those of the reference standard. The specificity, sensitivity, the positive and negative predictive values of the index test, and the positive and negative likelihood ratios will be calculated for each of these studies. We will also calculate the number needed to diagnose and misdiagnose. The diagnostic odds ratio will be calculated when outcomes of various studies are synthesized in a meta-analysis. Besides the 25% threshold for test positivity, we will also assess thresholds at respectively 15%, 35%, 45%, and 55% and will present them in receiver operating characteristic (ROC) curves88. Q values will not be used in this systematic review, because they frequently give a wrong impression of accuracy88. For the secondary research questions, the mean insertion torque values with their standard deviation for OMIs with and without root contact will be presented for each selected study. The mean difference between these recordings will be calculated. These values will be reported along with the 95% confidence intervals. These effect measures will be presented in a forest plot. Clinical and experimental studies are presented in separate figures. Statistical tests will be carried out with Review Manager version 5.389. All intervention groups of multi-arm studies will be listed in the table ‘Characteristics of included studies’. Unit of analysis issues could arise according to the level at which randomization occurs or in studies with repeated recordings of insertion torque values90. These issues will be analysed for each specific study design, and our primary analysis will be per randomized individual47,90. A meta-analysis is conducted in the case of the following: 1) low risk of bias in the selected studies, 2) consistent outcomes across the various studies, 3) low publication bias, 4) a high number of eligible studies, and 5) low heterogeneity86,90,91. Our protocol for conducting a meta-analysis for the primary and secondary research questions is presented in Additional file 688-99.

85

Subgroup analysis and investigation of heterogeneity Sources of heterogeneity that could affect outcomes can be categorized as study design, participants, target condition, index test, reference standard, implant, location of insertion, surgery, and setting related factors78,79,88. In this systematic review, we plan to investigate the following potential sources of heterogeneity:

• Study design: randomized versus non-randomized studies • Study participants: participants between the ages of 12 and 18 versus participants of 18 years

and older • The type of implant • The type of implant insertion: pre-drilling versus self-drilling insertion techniques • The type of index test: digital versus mechanical torque sensor • The type of reference standard: two-dimensional versus three-dimensional radiographs • The type of target condition: implant-root contact with root penetration versus without root

penetration • The time point for measuring outcomes

Heterogeneity between research models, that is, clinical, animal, and cadaver studies, will not be assessed, because these models are analysed separately. The following protocol will be applied for dealing with heterogeneity90,94,100-102:

1. Tables are created for each of the a priori defined sources of heterogeneity. Data are extracted and compiled in these tables, and the correctness of these procedures is subsequently double-checked. For both research questions, we will assess the same sources of heterogeneity.

2. If necessary, authors of selected research studies are contacted for clarification of specific issues.

3. The data extraction tables and the forest plots are subsequently inspected to assess whether it makes sense to apply statistics and further explore heterogeneity.

Subgroup analyses and meta-regression will be used to investigate statistical heterogeneity90. A statistician will be consulted for both statistics. We will report whether a subgroup analysis was planned a priori or was undertaken post hoc. Subgroup analyses will only be performed if sufficient data are reported for such assessments90. Meta-regressions will be conducted to investigate whether effect sizes are associated with specific characteristics of the study, for example, randomized sequence generation, and blinding of personnel. These analyses are only undertaken when a minimum of ten studies can be modelled for specific characteristics90. The limitations of both subgroup analyses and meta-regressions are considered when interpreting these statistics103. Models for investigating statistical heterogeneity are presented in Additional file 688-99.

Sensitivity analysis Sensitivity analyses will assess whether decisions made during the undertaking of the systematic review did affect its findings. Decision nodes that are pre-specified for our sensitivity analysis include the following: (1) the inclusion of grey literature or unpublished studies, (2) mechanical versus digital torque sensors, (3) the threshold for test positivity of 25% or more, (4) the inclusion of unblinded studies, and (5) the definition of the reference standard88. Additional sensitivity analyses will be conducted when specific issues suitable for such an analysis will arise during the review process88. All post hoc decisions to undertake such analyses will be reported.

86

Meta-biases To assess the presence of reporting bias, we will assess whether protocols of trials are available and whether they were published prior to recruiting participants47. The Clinical Trial Register at the International Clinical Trials Registry Platform of the World Health Organization will be searched to identify such studies published after 1 July 2005104. We will evaluate whether outcomes that were planned in the protocols were actually reported on in the published studies. Selective reporting of outcomes in all the eligible studies is also assessed as well as bias as a result of the outcomes of smaller studies. We will measure whether the random effects model presents more beneficial outcomes for the smaller studies than the fixed effect estimate47. Outcomes with or without data obtained from contacted authors will also be compared. The effect of including grey literature or unpublished studies on outcomes will be assessed in sensitivity analyses. Funnel plots will be conducted to further explore reporting bias65,105. Asymmetry in the funnel plots will only be assessed when ten or more eligible studies are identified, because with fewer articles the power of this statistics is too low65. All procedures to assess the meta-biases will be conducted by three review authors (RMR, LL, and LR).

Confidence in cumulative estimate/assessment of the quality of evidence (GRADE) Judgments about the quality of the evidence for the primary research question are rated according to the GRADE protocol (Table 3)106-109. The GRADEpro software will be used for the completion of this protocol, and the QUADAS-2 scores are integrated into this assessment110. For the secondary question, we assessed whether OMIs with root contact have higher insertion torque values than those without this target condition. Because this question does not specifically addresses a health problem, it does not qualify for an assessment using the GRADE approach.

87

Discussion The strengths of this systematic review are as follows: 1) it will be conducted by experienced reviewers, who have produced several systematic reviews and commentaries on this research topic73,78,79,111,112; and 2) it will be based on extensive literature searches, which will aim for high sensitivity and will accept low precision64. A weakness of this systematic review could be the eligibility of non-randomized studies and the inclusion of animal and cadaver models. However, we presented numerous reasons for placing this issue in a different perspective. These models could provide important knowledge for clinical applications and future research studies. Not considering these studies could waste valuable information and money, could slow down knowledge creation on this topic, and could lead to unnecessary duplication of research48-50,56-58. As outlined above, it could also be viewed as unethical not to include these research models. This issue was further strengthened in the context of the 1) high prevalence of the target disorder and the associated biologic damage and implant instability7-9,20; 2) ethical limitations of designing clinical studies on this research topic; 3) overwhelming evidence that data on adverse effects of interventions are poorly reported in the literature51,52; 4) the wide application of this interventional procedure in both orthodontics and maxillofacial surgery10,13-15; and 5) the usefulness of the research information for a wide group of stakeholders, that is, clinicians, researchers, patients in need of orthodontic implants or intermaxillary fixation screws, guideline developers, policymakers, and companies that produce implants and surgical instruments. Proposals for possible future research studies will be presented, taking in consideration, the inadequacies of the selected studies, patient-important outcomes, the setting, costs, the learning curve for the operator, and other barriers to the implementation of this health technology.

Additional files Additional file 1: Protocol for the search strategy. A detailed protocol for developing this search strategy. Additional file 2: Protocol for contacting authors. Protocol and sample email for contacting authors. Additional file 3: Data collection forms. Pilot tested collection forms used for this protocol. Additional file 4: QUADAS-2 quality assessment tool tailored to the clinical question80,82. The tailored and pilot tested QUADAS-2 tool. Additional file 5: Tabular presentation for QUADAS-2 results of the selected studies80. Exemplary table for the presentation of the QUADAS-2 scores. Additional file 6: Protocols for conducting meta-analyses and assessing statistical heterogeneity. Protocol for conducting a metaanalysis for the primary and secondary research questions and models for investigating statistical heterogeneity.

88

Abbreviations OMIs: orthodontic mini implants; RCTs: randomized controlled trials.

Competing interests The authors declare that they have no competing interests.

Authors’ contributions RMR conceived and designed the study and wrote the protocol and is the guarantor. LL and LR helped with the pilot testing of the data extraction forms and the search strategies and the development of the tailored QUADAS-2 tool. NR assisted with the development of the search strategy, and NDG, JDL, and AP have edited the draft protocol. All authors read and approved the final protocol.

Author details 1Department of Oral and Maxillofacial Surgery, Academic Medical Center, University of Amsterdam, Meibergdreef 9, 1105 AZ Amsterdam, The Netherlands. 2Private practice of orthodontics, Via Matteo Bandello 15, 20123 Milan, Italy. 3Department of Veterinary Sciences, University of Bologna, Via Tolara di Sopra 50, Ozzano dell’Emilia (BO) 40064, Italy. 4Department of Oral and Maxillofacial Surgery, Academic Medical Center and Academisch Centrum Tandheelkunde Amsterdam (ACTA), University of Amsterdam, Meibergdreef 9, Amsterdam AZ 1105, The Netherlands. 5Bodleian Health Care libraries, University of Oxford, Cairns Library Level 3, John Radcliffe Hospital, Oxford OX3 9DU, UK. 6Department of Primary Care Health Sciences, Centre for Evidence-Based Medicine, University of Oxford, New Radcliffe House, 2nd floor, Jericho, Oxford OX2 6NW, UK.

89

References

1. Antoszewska J, Trześniewska P, Kawala B, Ludwig B, Park H-S. Qualitative and quantitative evaluation of root injury risk potentially burdening insertion of miniscrew implants. Korean J Orthod. 2011;41(2):112–20.

2. Cho UH, Yu W, Kyung HM. Root contact during drilling for microimplant placement. Effect of surgery site and operator expertise. Angle Orthod. 2010;80(1):130–6.

3. Kim SH, Kang SM, Choi YS, Kook YA, Chung KR, Huang JC. Cone-beam computed tomography evaluation of mini-implants after placement: is root proximity a major risk factor for failure? Am J Orthod Dentofacial Orthop. 2010;138(3):264–76.

4. Lee KJ, Joo E, Kim KD, Lee JS, Park YC, Yu HS. Computed tomographic analysis of tooth-bearing alveolar bone for orthodontic miniscrew placement. Am J Orthod Dentofacial Orthop. 2009;135(4):486–94.

5. Motoyoshi M, Uchida Y, Matsuoka M, Inaba M, Iwai H, Karasawa Y, et al. Assessment of damping capacity as an index of root proximity in self-drilling orthodontic mini-implants. Clin Oral Investig. 2014;18(1):321–6.

6. Schnelle MA, Beck FM, Jaynes RM, Huja SS. A radiographic evaluation of the availability of bone for placement of miniscrews. Angle Orthod. 2004;74:832–7.

7. Alves Jr M, Baratieri C, Mattos CT, Araújo MT, Maia LC. Root repair after contact with mini-implants: systematic review of the literature. Eur J Orthod. 2013;35(4):491–9.

8. Papageorgiou SN, Zogakis IP, Papadopoulos MA. Failure rates and associated risk factors of orthodontic miniscrew implants: a meta-analysis. Am J Orthod Dentofacial Orthop. 2012;142(5):577–95. e7.

9. Suzuki M, Deguchi T, Watanabe H, Seiryu M, Iikubo M, Sasano T, et al. Evaluation of optimal length and insertion torque for miniscrews. Am J Orthod Dentofacial Orthop. 2013;144(2):251–9.

10. Keim RG, Gottlieb EL, Vogels 3rd DS, Vogels PB. 2014 JCO study of orthodontic diagnosis and treatment procedures, part 1: results and trends. J Clin Orthod. 2014;48(10):607–30.

11. Skeggs RM, Benson PE, Dyer F. Reinforcement of anchorage during orthodontic brace treatment with implants or other surgical methods. Cochrane Database Syst Rev. 2007;18(3), CD005098.

12. Hembree M, Buschang PH, Carrillo R, Spears R, Rossouw PE. Effects of intentional damage of the roots and surrounding structures with miniscrew implants. Am J Orthod Dentofacial Orthop. 2009;135(3):280. e1-9; discussion 280–1. doi:10.1016/j.ajodo.2008.06.022.

13. Farr DR, Whear NM. Intermaxillary fixation screws and tooth damage. Br J Oral Maxillofac Surg. 2002;40(1):84–5. 14. Jones DC. The intermaxillary screw: a dedicated bicortical bone screw for temporary intermaxillary fixation. Br J

Oral Maxillofac Surg. 1999;37(2):115–6. 15. Widar F, Kashani H, Kanagaraja S, Dahlin C, Rasmusson L. A retrospective evaluation of iatrogenic dental root

damage with predrilled vs drill-free bone anchor screws for intermaxillary fixation. Dent Traumatol. 2012;28(2):127–31.

16. Brisceno CE, Rossouw PE, Carrillo R, Spears R, Buschang PH. Healing of the roots and surrounding structures after intentional damage with miniscrew implants. Am J Orthod Dentofacial Orthop. 2009;135(3):292–301.

17. Liou EJ, Pai BC, Lin JC. Do miniscrews remain stationary under orthodontic forces? Am J Orthod Dentofacial Orthop. 2004;126:42–7.

18. Nienkemper M, Handschel J, Drescher D. Systematic review of mini-implant displacement under orthodontic loading. Int J Oral Sci. 2014;6(1):1–6.

19. Wang YC, Liou EJ. Comparison of the loading behaviour of self-drilling and predrilled miniscrews throughout orthodontic loading. Am J Orthod Dentofacial Orthop. 2008;133:38–43.

20. Kau CH, English JD, Muller-Delgardo MG, Hamid H, Ellis RK, Winklemann S. Retrospective cone-beam computed tomography evaluation of temporary anchorage devices. Am J Orthod Dentofacial Orthop. 2010;137(2):166. e1-5.

21. El-Beialy AR, Abou-El-Ezz AM, Attia KH, El-Bialy AM, Mostafa YA. Loss of anchorage of miniscrews: a 3-dimensional assessment. Am J Orthod Dentofacial Orthop. 2009;136(5):700–7.

22. Kuroda S, Yamada K, Deguchi T, Hashimoto T, Kyung HM, Takano-Yamamoto T. Root proximity is a major factor for screw failure in orthodontic anchorage. Am J Orthod Dentofacial Orthop. 2007;131 Suppl 4:S68–73.

23. Wang ZD, Li QY, Wang L, Gu YJ. Comparative evaluation of two kinds of micro-implant system with different size. Hua Xi Kou Qiang Yi Xue Za Zhi. 2009;27(2):150–3.

24. Choi HJ, Kim TW, Kim HW. A precise wire guide for positioning interradicular miniscrews. J Clin Orthod. 2007;41(5):258–61.

25. Estelita S, Janson G, Chiqueto K, Janson M, de Freitas MR. Predictable drill-free screw positioning with a graduated 3-dimensional radiographic-surgical guide: a preliminary report. Am J Orthod Dentofacial Orthop. 2009;136(5):722–35.

26. Kim SH, Choi YS, Hwang EH, Chung KR, Kook YA, Nelson G. Surgical positioning of orthodontic mini-implants with guides fabricated on models replicated with cone-beam computed tomography. Am J Orthod Dentofacial Orthop. 2007;131(4 Suppl):S82–9.

27. Suzuki EY, Buranastidporn B. An adjustable surgical guide for miniscrew placement. J Clin Orthod. 2005;39(10):588–90.

28. Gahleitner A, Podesser B, Schick S, Watzek G, Imhof H. Dental CT and orthodontic implants: imaging technique and assessment of available bone volume in the hard palate. Eur J Radiol. 2004;51(3):257–62.

90

29. Marmulla R, Wörtche R, Mühling J, Hassfeld S. Geometric accuracy of the NewTom 9000 Cone Beam CT. Dentomaxillofac Radiol. 2005;34(1):28–31.

30. De Paula-Silva FW, Wu MK, Leonardo MR, da Silva LA, Wesselink PR. Accuracy of periapical radiography and cone-beam computed tomography scans in diagnosing apical periodontitis using histopathological findings as a gold standard. J Endod. 2009;35(7):1009–12.

31. Estrela C, Bueno MR, Leles CR, Azevedo B, Azevedo JR. Accuracy of cone beam computed tomography and panoramic and periapical radiography for detection of apical periodontitis. J Endod. 2008;34(3):273–9.

32. Chen YH, Chang HH, Chen YJ, Lee D, Chiang HH, Yao CC. Root contact during insertion of miniscrews for orthodontic anchorage increases the failure rate: an animal study. Clin Oral Implants Res. 2008;19(1):99–106.

33. Knottnerus JA, Buntinx F, Van Weel C. Chapter 1: general introduction: evaluation of diagnostic procedures. In: Knottnerus JA, Buntinx F, editors. The evidence of Clinical Diagnosis: Theory and Methods of Diagnostic Research. Oxford: Blackwell Publishing; 2009. p. 256–72.

34. US Food and Drug Administration (FDA). Dental cone-beam computed tomography: [http://www.fda.gov/Radiation-EmittingProducts/Radiation EmittingProductsandProcedures/MedicalImaging/MedicalX-Rays/ucm315011.htm]. Accessed 29 August 2014.

35. ASTM. Standard F543-07ε1: standard specification and test methods for metallic medical bone screws. PA, USA: ASTM International, West Conshohocken; 2007 (Editorially corrected August 2009).

36. Brozek JL, Akl EA, Alonso-Coello P, Lang D, Jaeschke R, Williams JW, et al. GRADE Working Group: grading quality of evidence and strength of recommendations in clinical practice guidelines. Part 1 of 3. An overview of the GRADE approach and grading quality of evidence about interventions. Allergy. 2009;64(5):669–77.

37. McEwan MB. An evaluation of the insertion torque of orthodontic miniscrew implants in relation to tooth root contact. Saint Luis (USA), Department of Dentistry: MSc thesis; 2012.

38. Wilmes B, Su YY, Sadigh L, Drescher D. Pre-drilling force and insertion torques during orthodontic mini-implant insertion in relation to root contact. J Orofac Orthop. 2008;69(1):51–8.

39. Motoyoshi M, Hirabayashi M, Uemura M, Shimizu N. Recommended placement torque when tightening an orthodontic mini-implant. Clin Oral Implants Res. 2006;17:109–14.

40. Hsu J, Brożek JL, Terracciano L, Kreis J, Compalati E, Stein AT, et al. Application of GRADE: making evidence-based recommendations about diagnostic tests in clinical practice guidelines. Implement Sci. 2011;6:62.

41. Bossuyt PM, Irwig L, Craig J, Glasziou P. Comparative accuracy: assessing new tests against diagnostic pathways. BMJ. 2006;332(7549):1089–92.

42. Haynes RB, You JJ. Chapter 2: the architecture of diagnostic research. In: Knottnerus JA, Buntinx F, editors. The evidence of Clinical Diagnosis: Theory and Methods of Diagnostic Research. Oxford: Blackwell Publishing; 2009. p. 256–72.

43. Thompson MT, Van den Bruel A. Chapter 3: asking an answerable clinical question. In: Thompson MT, Van den Bruel A, editors. Diagnostic tests toolkit. Chichester: Wiley-Blackwell; 2012. p. 16–7.

44. Deeks JJ, Wisniewski S, Davenport C. Chapter 4: guide to the contents of a Cochrane Diagnostic Test Accuracy Protocol. In: Deeks JJ, Bossuyt PM, Gatsonis C, editors. Cochrane handbook for systematic reviews of diagnostic test accuracy version 1.0.0. The Cochrane Collaboration, 2013. http://www.srdta.cochrane.org/.

45. Higgins JPT, Green S. Cochrane handbook for systematic reviews of interventions version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011. http://www.cochrane-handbook.org.

46. Moher D, Shamseer L, Clarke M, Ghersi D, Liberati A, Petticrew M, et al. Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols (PRISMA-P) 2015 statement. Syst Rev. 2015;4(1):1 [Epub ahead of print].

47. Shamseer L, Moher D, Clarke M, Ghersi D, Deceased LA, Petticrew M, et al. the PRISMA-P Group: Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015: elaboration and explanation. BMJ. 2015;349:g7647.

48. Hooijmans CR, Ritskes-Hoitinga M. Progress in using systematic reviews of animal studies to improve translational research. PLoS Med. 2013;10(7):e1001482.

49. Chalmers I, Glasziou P. Avoidable waste in the production and reporting of research evidence. Lancet. 2009;374(9683):86–9.

50. Chalmers I: discussion and concluding remarks by Sir Iain Chalmers during the First International Symposium Systematic Reviews in Laboratory Animal Science. February 9th 2012, Nijmegen, The Netherlands. [http://www. youtube.com/watch?v=j5RDdkU2yxE]

51. Ioannidis JP, Evans SJ, Gøtzsche PC, O’Neill RT, Altman DG, Schulz K, et al. CONSORT Group: better reporting of harms in randomized trials: an extension of the CONSORT statement. Ann Intern Med. 2004;141(10):781–8.

52. Loke YK, Price D, Herxheimer A. Chapter 14: adverse efffects. In: Higgins JPT, Green S. editors. Cochrane handbook for systematic reviews of interventions version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011. http://www.cochrane-handbook.org.

53. PROSPERO. Centre for reviews and dissemination. University of York, UK. http://www.crd.york.ac.uk/PROSPERO/ 54. Bossuyt PM, Leeflang MM. Chapter 6: developing criteria for including studies. In: Deeks JJ, Bossuyt PM, Gatsonis

C, editors. Cochrane handbook for systematic reviews of diagnostic test accuracy version 0.4 [updated September 2008]. The Cochrane Collaboration, 2008. http://www.srdta.cochrane.org/.

http://www.srdta.cochrane.org/



http://www.crd.york.ac.uk/PROSPERO/

http://www.srdta.cochrane.org/

91

55. Knottnerus JA, Muris JW. Chapter 3: assessment of the accuracy of diagnostic tests: the cross-sectional study. In: Knottnerus JA, Buntinx F, editors. The evidence of clinical diagnosis: theory and methods of diagnostic research. Oxford: Blackwell Publishing; 2009. p. 256–72.

56. Reeves BC, Deeks JJ, Higgins JPT, Wells GA. Chapter 13: including nonrandomized studies. In: Higgins JPT, Green S, editors. Cochrane handbook for systematic reviews of interventions version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011. http://www.cochrane-handbook.org.

57. OCEBM Levels of Evidence Working Group. The Oxford 2011 levels of evidence. Oxford Centre for Evidence-Based Medicine. http://www.cebm. net/index.aspx?o=5653. Accessed 9 September 2014.

58. Reeves BC, Higgins JPT, Ramsay C, Shea B, Tugwell P, Wells GA. An introduction to methodological issues when including non-randomised studies in systematic reviews on the effects of interventions. Research Synthesis Methods. 2013;4:1–11.

59. Pandis N. The evidence pyramid and introduction to randomized controlled trials. Am J Orthod Dentofacial Orthop. 2011;140:446–7.

60. Guyatt GH, Oxman AD, Sultan S, Glasziou P, Akl EA, Alonso-Coello P, et al. GRADE Working Group: GRADE guidelines: 9. Rating up the quality of evidence. J Clin Epidemiol. 2011;64(12):1311–6.

61. Schünemann HJ, Oxman AD, Vist GE, Higgins JPT, Deeks JJ, Glasziou P, et al. Chapter 12: interpreting results and drawing conclusions. In: Higgins JPT, Green S, editors. Cochrane handbook for systematic reviews of interventions version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011. http://www.cochrane-handbook.org.

62. De Vet HCW, Eisinga A, Riphagen II, Aertgeerts B, Pewsner D. Chapter 7: searching for studies. In: Deeks JJ, Bossuyt PM, Gatsonis C, editors. Cochrane handbook for systematic reviews of diagnostic test accuracy version 0.4 [updated September 2008]. The Cochrane Collaboration, 2008. http://www. srdta.cochrane.org/.

63. Kanomi R. Mini-implant for orthodontic anchorage. J Clin Orthod. 1997;31(11):763–7. 64. Lefebvre C, Manheimer E, Glanville J. Chapter 6: searching for studies. In: Higgins JPT, Green S, editors. Cochrane

handbook for systematic reviews of interventions version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011. http://www.cochrane-handbook.org.

65. Sterne JAC, Egger M, Moher D. Chapter 10: addressing reporting biases. In: Higgins JPT, Green S, editors. Cochrane handbook for systematic reviews of interventions version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011. http://www.cochrane-handbook.org.

66. Hopewell S, McDonald S, Clarke M, Egger M. Grey literature in meta-analyses of randomized trials of health care interventions. Cochrane Database Syst Rev. 2007, Apr 18;2, MR000010.

67. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gøtzsche PC, Ioannidis JP, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. PLoS Med. 2009;6(7):e1000100. Epub 2009 Jul 21.

68. Beynon R, Leeflang MM, McDonald S, Eisinga A, Mitchell RL, Whiting P, et al. Search strategies to identify diagnostic accuracy studies in MEDLINE and EMBASE. Cochrane Database Syst Rev. 2013 Sep 11;9, MR000022. doi:10.1002/14651858.MR000022.pub3.

69. Relevo R. Chapter 4: effective search strategies for systematic reviews of medical tests. AHRQ publication no. 12-EHC073-EF. In: Chang SM, Matchar DB, editors. Methods guide for medical test reviews (AHRQ publication no. 12-EHC017). Rockville (USA): Agency for Healthcare Research and Quality; 2012 [http://www.effectivehealthcare.ahrq.gov/search-for-guides-reviewsand- reports/?pageaction=displayproduct&productid=558].

70. Bramer WM, Giustini D, Kramer BMR, Anderson PF. The comparative recall of Google Scholar versus PubMed in identical searches for biomedical systematic reviews: a review of searches used in systematic reviews. Syst Rev. 2013;2:115.

71. National Library of Medicine (NLM) fact sheet. http://www.nlm.nih.gov/pubs/factsheets/errata.html. Accessed 4 December 2013.

72. Higgins JPT, Deeks JJ. Chapter 7: selecting studies and collecting data. In: Higgins JPT, Green S, editors. Cochrane handbook for systematic reviews of interventions version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011. http://www.cochrane-handbook.org.

73. Meursinge Reynders R, Ronchi L, Ladu L, Van Etten-Jamaludin F, Bipat S. Insertion torque and orthodontic mini-implants: a systematic review of the artificial bone literature. Proc Inst Mech Eng H. 2013;227(11):1181–202.

74. Atherton H. Is there value in making author contact when determining risk of bias in a Cochrane Systematic Review. http://community.cochrane.org/news/blog/there-value-making-author-contact-when-determining-risk-biascochrane-systematic-review. Accessed 10 December 2014.

75. Moher D, Liberati A, Tetzlaff J, Altman DG. PRISMA group: preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med. 2009;6(7):e1000097.

76. STARD 2014. Towards complete and accurate reporting of studies of diagnostic accuracy: the STARD initiative. Standards for Reporting of Diagnostic Accuracy. The STARD checklist. http://www.stard-statement.org/. Accessed 10 August 2010.

77. Bossuyt PM, Reitsma JB, Bruns DE, Gatsonis CA, Glasziou PP, Irwig LM, et al. Standards for Reporting of Diagnostic Accuracy: the STARD statement for reporting studies of diagnostic accuracy: explanation and elaboration. Clin Chem. 2003;49(1):7–18.


http://www.cebm/


http://www/



http://www.nlm.nih.gov/


http://community.cochrane.org/news/blog/there-value-making-author-contact-when-determining-risk-biascochrane-systematic-review.%20Accessed%2010%20December%202014

http://community.cochrane.org/news/blog/there-value-making-author-contact-when-determining-risk-biascochrane-systematic-review.%20Accessed%2010%20December%202014

http://www.stard-statement.org/

92

78. Meursinge Reynders RA, Ronchi L, Ladu L, Van Etten-Jamaludin F, Bipat S. Insertion torque and success of orthodontic mini-implants: a systematic review. Am J Orthod Dentofacial Orthop. 2012;142(5):596–614. e5.

79. Reynders R, Ronchi L, Bipat S. Mini-implants in orthodontics: a systematic review of the literature. Am J Orthod Dentofacial Orthop. 2009;135:564. e1-19.

80. QUADAS-2: A quality assessment tool for diagnostic accuracy studies: University of Bristol. http://www.bris.ac.uk/quadas/. Accessed 1 March 2014.

81. Higgins JPT, Deeks JJ, Altman DG. Chapter 16: special topics in statistics. In: Higgins JPT, Green S, editors. Cochrane handbook for systematic reviews of interventions version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011. http://www.cochrane-handbook.org.

82. Whiting PF, Rutjes AW, Westwood ME, Mallett S, Deeks JJ, Reitsma JB, et al. QUADAS-2 group: QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med. 2011;155(8):529–36.

83. Santaguida PL, Riley CM, Matchar DB, Santaguida PL, Riley CM, Matchar DB. Chapter 5: assessing risk of bias as a domain of quality in medical test studies. AHRQ publication no. 12-EHC073-EF. In: Chang SM, Matchar DB, editors. Methods guide for medical test reviews (AHRQ publication no. 12-EHC017). Rockville (USA): Agency for Healthcare Research and Quality; 2012 [http://www.effectivehealthcare.ahrq.gov/search-for-guides-reviewsand- reports/?pageaction=displayproduct&productid=558]

84. Jüni P, Witschi A, Bloch R, Egger M. The hazards of scoring the quality of clinical trials for meta-analysis. JAMA. 1999;282(11):1054–60.

85. Whiting P, Harbord R, Kleijnen J. No role for quality scores in systematic reviews of diagnostic accuracy studies. BMC Med Res Methodol. 2005;5:19.

86. Higgins JPT, Altman DG, Sterne JAC. Chapter 8: assessing risk of bias in included studies. In: Higgins JPT, Green S, editors. Cochrane handbook for systematic reviews of interventions version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011. http://www.cochrane-handbook.org.

87. Reitsma JB, Rutjes AWS, Whiting P, Vlassov VV, Leeflang MMG, Deeks JJ. Chapter 9: assessing methodological quality. In: Deeks JJ, Bossuyt PM, Gatsonis C, editors. Cochrane handbook for systematic reviews of diagnostic test accuracy version 1.0.0. The Cochrane Collaboration, 2009. http://srdta.cochrane.org/handbook-dta-reviews.

88. Macaskill P, Gatsonis C, Deeks JJ, Harbord RM, Takwoingi Y. Chapter 10: analysing and presenting results. In: Deeks JJ, Bossuyt PM, Gatsonis C, editors. Cochrane handbook for systematic reviews of diagnostic test accuracy version 1.0.0. The Cochrane Collaboration, 2010. http://srdta. cochrane.org/handbook-dta-reviews.

89. The Nordic Cochrane Centre. Review Manager (RevMan) [Computer program]: version 5.3. Copenhagen: the Nordic Cochrane Centre, The Cochrane Collaboration; 2014.

90. Deeks JJ, Higgins JPT, Altman DG. Chapter 9: analysing data and undertaking meta-analyses. In: Higgins JPT, Green S, editors. Cochrane handbook for systematic reviews of interventions version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011. http://www.cochrane-handbook.org.

91. Borenstein M, Hedges LV, Higgins JPT, Rothstein HR. Chapter 40: when does it make sense to perform a meta-analysis? In: Borenstein M, Hedges LV, Higgins JPT, Rothstein HR, editors. Introduction to Meta-Analysis. Chichester (UK): John Wiley & Sons; 2009.

92. Reitsma JB, Glas AS, Rutjes AW, Scholten RJ, Bossuyt PM, Zwinderman AH. Bivariate analysis of sensitivity and specificity produces informative summary measures in diagnostic reviews. J Clin Epidemiol. 2005;58:982–90.

93. Rutter CM, Gatsonis CA. A hierarchical regression approach to meta-analysis of diagnostic test accuracy evaluations. Stat Med. 2001;20:2865–84.

94. Higgins JPT, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ. 2003;327:557–60.

95. Borenstein M, Hedges LV, Higgins JPT, Rothstein HR. Chapter 16: identifying and quantifying heterogeneity. In: Borenstein M, Hedges LV, Higgins JPT, Rothstein HR, editors. Introduction to meta-analysis. Chichester (UK): John Wiley & Sons; 2009.

96. Higgins JPT, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med. 2002;21:1539–58. 97. Higgins JP. Commentary: heterogeneity in meta-analysis should be expected and appropriately quantified. Int J

Epidemiol. 2008;37(5):1158–60. 98. Borenstein M, Hedges LV, Higgins JPT, Rothstein HR. Chapter 1. How a meta-analysis works. In: Borenstein M,

Hedges LV, Higgins JPT, Rothstein HR, editors. Introduction to meta-analysis. Chichester (UK): John Wiley & Sons; 2009.

99. O’Rourke K, Detsky AS. Meta-analysis in medical research: strong encouragement for higher quality in individual research efforts. J Clin Epidemiol. 1989;42(10):1021–4.

100. Schroll JB, Moustgaard R, Gøtzsche PC. Dealing with substantial heterogeneity in Cochrane reviews. Cross-sectional study. BMC Med Res Methodol. 2011;11:22.

101. Sutton AJ, Abrams KR, Jones DR, Sheldon TA, Song F. Chapter 3: assessing between study heterogeneity. In: Sutton AJ, Abrams KR, Jones DR, Sheldon TA, Song F, editors. Methods for meta-analysis in medical research. Chichester (UK): John Wiley & Sons; 2000.

102. West SL, Gartlehner G, Mansfield AJ, Poole C, Tant E, et al. Comparative effectiveness review methods: clinical heterogeneity. Agency for Healthcare Research and Quality. Publication No. 10-EHC070-EF. September 2010. http://www.effectivehealthcare.ahrq.gov/ehc/products/93/533/Clinical_Heterogeneity_Revised_Report.pdf.

103. Oxman AD, Guyatt GH. A consumer’s guide to subgroup analyses. Ann Intern Med. 1992;116(1):78–84.



http://srdta.cochrane.org/handbook-dta-reviews

http://srdta/


http://www.effectivehealthcare.ahrq.gov/ehc/products/93/533/Clinical_Heterogeneity_Revised_Report.pdf

93

104. International Clinical Trials Registry Platform of the World Health Organization. http://apps.who.int/trialsearch/AdvSearch.aspx. Accessed January 2015.

105. Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ. 1997;315(7109):629–34.

106. Balshem H, Helfand M, Schünemann HJ, Oxman AD, Kunz R, Brozek J, et al. GRADE guidelines: 3. Rating the quality of evidence. J Clin Epidemiol. 2011;64(4):401–6.

107. The Grading of Recommendations Assessment, Development and Evaluation (GRADE) Working Group. GRADE working group. http://www. gradeworkinggroup.org. Accessed 10 August 2014.

108. Guyatt G, Oxman AD, Akl EA, Kunz R, Vist G, Brozek J, et al. GRADE guidelines: 1. Introduction-GRADE evidence profiles and summary of findings tables. J Clin Epidemiol. 2011;64(4):383–94.

109. Schünemann HJ, Oxman AD, Brozek J, Glasziou P, Jaeschke R, Vist GE, et al. GRADE working group: grading quality of evidence and strength of recommendations for diagnostic tests and strategies. BMJ. 2008;336(7653):1106–10.

110. GRADEpro (GRADEproGDT) software from the GRADE working group. http://www.gradepro.org. Accessed 10 August 2010.

111. Meursinge Reynders RA. Low quality evidence on the stability of orthodontic mini-implants. Evid Based Dent. 2013;14:78–80.

112. Meursinge Reynders RA. Moderate quality evidence that surgical anchorage more effective than conventional anchorage during orthodontic treatment. Evid Based Dent. 2014;15:108–9.

http://www/

94

Additional files

Additional file 1: Protocol for the search strategy. A detailed protocol for developing our search strategy

• We will aim for a broad-spectrum search strategy, because this systematic review addresses different review questions and because diagnostic accuracy studies are often poorly reported and lack appropriate subject headings62. The search strategies for the various databases will therefore be highly sensitive, i.e., low risk of missing relevant studies, but with low precision, i.e. high number of irrelevant papers62. We will not apply “methodological or other search filters” to avoid the exclusion of pertinent papers62,68. Diagnostic search filters are currently not reliable when searching studies for systematic reviews69.

• Search strategies for articles on diagnostic test accuracy typically use two sets of search terms: terms related to the index test(s) and those related to the target condition(s)62. However, the target condition “implant-root contact” of our clinical question has not been defined precisely in the orthodontic literature. We therefore subdivided the target condition in 2 sets of search terms: 1) those that refer to root contact and 2) those that refer to implants.

• The following protocol was used to find pertinent subject headings and key words : 1) conduct preliminary searches to retrieve some relevant articles or related reviews to avoid relying exclusively on controlled vocabulary terms alone69; 2) use the specific search tools of individual databases, e.g The “Related Articles” option in PubMed, to find additional relevant articles; 3) identify key concepts, e.g., the index test, the target condition, and the description of the patients in these papers; 4) create search terms through the identification of synonyms, related terms, acronyms, abbreviations, variant spellings, and subject headings for these key concepts. Database thesauri are also used for this purpose; and 5) create subject headings that database indexers have assigned to these articles. The identified search terms are summarized in table 1.

• The Boolean operator ‘AND’ will be used for the key concepts and the search terms of each of these concepts will be combined by the Boolean operator ‘OR’. To avoid inappropriate exclusion of relevant studies we will not use The “NOT” operator.

• We will pilot test these strategies for each database and subsequently fine-tune them62. Examples are presented for the search strategy of MEDLINE and Google Scholar in Table 264,70.

• Search strategies for each database will be listed in a table, which presents the number of records retrieved with the specific start and end dates of the searches. This table will facilitate future updating of this systematic review and improve the transparency of the search methods67.

• Fraudulent studies, other retracted publications, errata and comments will be searched in MEDLINE62,71.

95

Additional file 2: Protocol for contacting authors. Protocol for contacting authors During the study selection and data extraction and analysis procedures, authors of primary research studies will be contacted to obtain additional information on eligibility issues or unclear or missing data. For this purpose the following protocol will be applied: (1) an email will be sent to the pertinent authors to ask their willingness to provide such information; (2) a second email will be sent in case of no reply; (3) if necessary, a third email will be sent from a different email address. This decision was made to avoid the risk of an email address being classified as “spam mail” by the internet provider73; (4) if emails are still not answered, at least one of the co-authors of the pertinent research study will also be contacted; (5) questions will be presented as a combination of open-and closed-ended questions. An example of our initial email is presented under here. The information obtained from the contacted investigators and the consequences of this information on e.g. outcomes or risk of bias assessments will be summarized in a table. Because the validity of making author contact is currently unknown, all information obtained through the “contacting authors” procedures will be first analyzed and reported separately from the data extracted from the original articles74. Email example for contacting authors

Dear Professor …………., We are currently conducting a systematic review* on orthodontic mini implants and would like to obtain some additional information on your article: “ …………………………………………………………….”

In the near future we would like to ask you some simple questions on this publication. Answering these questions will probably not require more than 2 minutes of your time.

In this email we are only interested to find out whether you are willing to provide us with this additional information or not. We therefore ask you the following question: Are you in the near future willing to respond to some simple questions pertinent to the indicated article ? - If you are willing to respond to our questions please reply to this email, a "YES" is enough. - If you are not willing to respond to our questions please respond this email with a “NO”. In case, you do not reply, we will send you a reminder email 14 days after the current email.

-In case you respond with a “NO” or do not reply to this email and to the reminder, we will not further contact you, but will contact at least one of the other co-authors of your research study.

Thank you very much for your cooperation. Sincerely, Reint Meursinge Reynders and Nicola Di Girolamo ----------------------------------------------------- *This systematic review originated as an assignment for the Evidence Based Health Care program at the University of Oxford, UK. The aforementioned individuals are subjectively responsible for the information requested.

96

Additional file 3: Data collection forms. Pilot tested collection forms used for this protocol. Data collection form 1. Source and eligibility56,72

Entry Description and character of the information Name of reviewer Report the name of the reviewer. Authors, title, journal of article List authors, title, and journal. Source of article Describe how the article was retrieved, e.g. Medline, grey literature, hand searching

of review articles etc. Language of the article Describe in which language the article was published. Trial register Describe if the trial was registered, under what number and in which register. Eligible/not eligible Confirm eligibility.

If not eligible, explain the reason for exclusion. Purpose Copy the objectives of the paper according to the authors.

Data collection form 2. The target condition and primary outcomes72*

Entry Description and character of the information Target condition

Present the definition of the target condition according to the authors, e.g., no root contact, single or multiple root contact without root penetration (glancing), root contact with root penetration.

Unit of measurement Primary outcome

Describe the unit of measurement of measuring torque of the primary outcome, e.g., Newton centimeter (Ncm).

Time point of outcome assessment Primary outcome

Describe the time point of the primary outcome assessment, e.g. maximum insertion torque during the entire insertion process or measured during the last turn of the OMI.


97

Data collection form 3. Results72* Entry Description and character of the information Study design and selection procedures

e.g. consecutively treated randomized patients.

Setting e.g. private practice, laboratory setting, university clinic. Country and location Indicate the country and the city where the study was conducted. Characteristics of the population and number of participants and implants

e.g. Self-drilling group: n = 50 (70 implants) 20 males and 20 females mean age 22.3 ± 7.9 years Pre-drilling group: n= 50 (70 implants) 20 males and 20 females mean age 23.6 ± 8.1 years If indicated summarize the division of experimental groups, e.g. number of patients or implants per trial arm or subgroup.

Implant type Indicate the type (tapered or cylindrical etc.), the diameter, and length of the implant and the company.

Location of insertion e.g. between maxillary first molar and second bicuspid. Insertion technique Self-drilling and pre-drilling. Index test Present the type and the company of the index test and indicate whether the torque

sensor is mechanical or digital. Torque measurement e.g., Maximum insertion torque (MIT) during the terminal rotation of the screw. Pre-specified torque threshold Present the cut-off point for implant-root contact, e.g. 50% increase in MIT. Reference standard Present the type and company of the reference standard. Definition of the target condition

No root contact, root contact without penetration (glancing), root contact with penetration.

Prevalence of implant-root contact

Present the prevalence of implant-root contact with the percentages for the different target conditions, e.g., with or without root penetration.

Accuracy Present the accuracy of the index test Maximum insertion torque values with or without implant-root contact

Self-drilling 1 point contact: 8.0 ± 3.0 Ncm Self-drilling multiple contacts: 9.5 ± 2.3 Ncm Pre-drilling 1 point contact: 5.4 ± 1.3 Ncm Pre-drilling multiple contacts: 7.4 ± 2.1 Ncm

Adverse effects Present ratios of adverse effects of interventions, e.g. implant fracture ratios (fractured implants/total number of inserted implants). Score adverse effects according to ‘Data collection form 4’.

*Locate where information on each item can be found, e.g., Page 12 column 3. Data collection form 4. Adverse effects72*

Entry Description and character of the information Biologic damage Define identified biologic damage and how it is scored79. Implant fracture Describe for each subgroup the number of implants that fractured during their

insertion and removal. Present the fracture ratio (the number of fractured implants/total number of implants).

Inflammation Define identified inflammation and how it is scored79. Pain and discomfort Define identified pain and discomfort and how they are scored79. Summary critique Summarize the strength and weaknesses of adverse effects in the research study.


98

Data collection form 5. Miscellaneous72* Entry Description and character of the information Funding source Describe the funding source and indicate if it is an implant company. Key conclusions of study authors

Describe the main conclusions by the authors of the research study.

Important comments by the study authors










Contacting authors Indicate whether authors should be contacted and indicated what additional information is necessary.

Discuss with statistician Describe points that need to be discussed with a statistician. Points of interest Describe points of interest, e.g. the description of a new type of index test. Future studies Describe ideas for future studies suggested by the authors or reviewers. Other comments by review authors

Any additional comment that has not been covered.

*Locate where information on each item can be found, e.g., Page 12 column 3. QUADAS-2 Collection form 1. Domain 1. Patient selection “risk of bias”56,72,82*

Entry Description and character of the information Date of the study Describe when the study was started and completed. Study design Describe the type of study design, e.g. randomized controlled trial, case-control

study, cohort, splitmouth etc.? Consecutively treated Describe whether participants were consecutively treated or not. Sequence generation 1) Random sequence generation? Describe the type of randomization, e.g.

computerized. 2) Non-random sequence generation?

Allocation concealment Concealment of allocation. Describe blinding of patients, surgeon, test operators, and personnel.

Case-control design Describe whether patients were enrolled according to a case-control design. Power calculation Describe how power calculation was conducted. Inappropriate exclusions Describe whether single patients or specific groups of patients were inappropriate

excluded during any phase of the patient selection process. Approved by ethical board Describe if the research study was approved by an ethical board and describe the

components of this ethical board. *Locate where information on each item can be found, e.g., Page 12 column 3.

99

QUADAS-2 Collection form 2. Domain 1: Patient selection “Concerns of applicability”56,72,82*

*Locate where information on each item can be found, e.g., Page 12 column 3. QUADAS-2 Collection form 3. Domain 2: Index test-Implants “risk of bias”56,72,82*

Entry Description and character of the information Number of implants Present the total number of implants and if applicable the number of implants per

subgroup. Implant type Present the type of implant (code number) with the name of the company(ies) and if

applicable the number of implants per subgroup. Implant form Describe the form of the screw: tapered/cylindrical/or a combination of forms.

Describe these specifics for each part of the screw. Implant material Describe the chemical make-up of the screw. Diameter (D) Length (L)

Describe the thread diameter and core diameter in mm. Describe the thread length and body length in mm.

Surface finish Describe the type of surface finish. Drilling design Describe if the screw has a self-drilling or a pre-drilling design. Thread characteristics Describe the width and angles of the crest of the screw. Pitch dimensions Describe the dimensions of the pitch of the screw. Flute in tip Describe if the tip of the screw is fluted or not.

Describe the length of the flute. Flute in core Describe if the core of the screw is fluted or not.

Describe the length of the flute. Sterilized Describe if the screws are sterilized by the company or by the operator.


Entry Description and character of the information Number Present the total number of patients and the number of patients per subgroup. Ethnicity Which ethnic group(s)? If applicable, present the number of patients per subgroup. Socio-economic status Describe the socio-economic status of the patient(s)? If applicable, present the

number of patients per subgroup. Sex Present the divisions of the sexes in the research study. If applicable, present the

number of patients per subgroup. Age List age in years and months with standard deviations and/or ranges. If applicable,

present the number of patients per subgroup. Additional patient characteristics Describe additional characteristics, e.g. university or high school students, dental

students etc. If applicable, present the number of patients per subgroup. Country Describe in which country(ies) the research study was conducted. If applicable,

present the number of patients per subgroup. Medical and dental health condition

Describe medical variables, e.g. uncontrolled diabetes, osteoporosis, smoking, pharmacological treatment etc. If applicable, present the number of patients per subgroup. Describe dental variables, e.g. periodontal disease, loss of teeth etc. If applicable, present the number of patients per subgroup.

Setting Describe the setting of the research study, e.g. dental school. If applicable, present the number of patients per subgroup.

Co-existent conditions Describe possible past interventions or co-interventions, previous interventions etc. If applicable, present the number of patients per subgroup.

Previous testing Describe possible previous testing procedures prior to conducting the diagnostic test that could have changed the spectrum of the patients. If applicable, present the number of patients per subgroup.

Severity of the target condition Describe the severity of the target condition. The numbers of root glancing versus root penetration could be skewed. If applicable, present the number of patients per subgroup.

Intended use of the index test Assess whether the index test was used for assessing the target condition or for stability assessments of OMIs or both. If applicable, present the number of patients per subgroup.

Other comorbid conditions

Describe other comorbid conditions, e.g., change in patient’s health? If applicable, present the number of patients per subgroup.

100

QUADAS-2 Collection form 4. Domain 2: Index test-Location “risk of bias”56,72,82* Entry Description and character of the information Implant site Describe the exact insertion site of the implant. Distance to root Describe the distance between the implant and the root of a tooth in mm. Assessment of bone condition Describe if any quality assessment of the bone e.g. radiographically was undertaken

prior to surgery. Bone thickness Describe the thickness of the cortical and trabecular bone in mm. Keratinized/non keratinized mucosa

Describe if the implant was placed in the keratinized or non-keratinized mucosa.

Mucosal thickness Describe the thickness of the mucosa in mm. Exposed/non exposed Describe if the implant was left exposed or non-exposed under the mucosa after

surgery. *Locate where information on each item can be found, e.g., Page 12 column 3. QUADAS-2 Collection form 5. Domain 2: Index test-Surgery “risk of bias”56,72,82*

Entry Description and character of the information Experience of the operator Describe the number of mini-implants inserted by the operator prior to this study. One or multiple operators Describe if all surgeries were conducted by the same operator and if not how many

implants were placed by each operator. Flap or flapless surgery Describe if surgery was flap or flapless. Self-drilling or pre-drilling insertion technique

Describe if surgery was conducted according to the self-drilling or pre-drilling insertion technique.

Pilot holes for the self-drilling technique

Describe if pilot holes were drilled for the self-drilling technique and describe diameter and depth in mm.

Pre-drilling device for the self-drilling pilot holes

Describe the type of device that was used for the self-drilling pilot holes.

Pre-drilling speed for the selfdrilling pilot holes

Describe the pre-drilling speed in rpm (rounds per minute) for the self-drilling pilot holes.

Pilot holes for the pre-drilling technique

Describe if pilot holes were drilled for the predrilling technique and describe the diameter and depth in mm.

Pre-drilling device for the pre-drilling pilot holes

Describe the type of device that was used for the pre-drilling pilot holes.

Pre-drilling speed for the predrilling pilot holes

Describe the pre-drilling speed in rpm (rounds per minute) for the predrilling pilot holes.

Distance between the screws In the case of multiple screws, describe the distances between the screws in mm. Insertion depth Describe the insertion depth in mm. Friction between the head of the screw and the bone

Describe if the screw head touched the bone and had caused friction. Such a contact could have caused stripping of the bone.

The direction of insertion Describe the angulation of insertion of the screw in degrees. Insertion speed Describe the insertion speed in rpm (rounds per minute). Axial insertion force Describe the axial insertion force in kg. Insertion with insertion torqueing device

Describe if the insertion was conducted with a torqueing device.

Type of torqueing device Describe the type of the insertion torqueing device and describe if it was a mechanical or a digital device.

Stripping of the bone Describe if stripping occurred (rotation of the screw without penetration of the bone).

Mono or bicortical anchorage Describe if the anchorage was mono or bicortical. *Locate where information on each item can be found, e.g., Page 12 column 3.

101

QUADAS-2 Collection form 6. Domain 2: Index test-collected items “risk of bias” and “concerns of applicability”56,72,82* Entry Description and character of the information Implant, location, or surgery related factors

Describe implant, location or surgery related factors that could have caused biased outcomes of the index test. (extracted in data collection forms 3-5)

Experience of the operator Describe the experience of the operator, i.e., the operator that inserts the implant with the digital torque sensor.

Foreknowledge of the results of the reference standard

Describe whether operators had foreknowledge of results of the reference standard. This entry is never applicable for our index test, because this test is always conducted prior to the reference standard.

Foreknowledge of intermediate test results

Describe whether operators had foreknowledge of results of intermediate test results, intermediate recordings with the index test, i.e., insertion torque or other intermediate tests, e.g., resonance frequency assessments.

Pre-specified threshold of the index test

Describe what the threshold of the index test is and how and when it was established, i.e., a priori or post hoc.

Timing of outcome measurement

Describe when the outcome (maximum insertion torque) was recorded, i.e. maximum torque during the entire insertion process or maximum torque during the final rotation of the screw.

Quality torque recording Describe the quality of the torque recording, e.g., conducted with a mechanical or a digital device.

Number of outcome assessors Describe the number of outcome assessors. Ideally more than one operator assesses the outcomes in order to avoid inadequate readings or exclusions of readings.

Calibration of the torque sensor

Describe whether the torque sensor was calibrated and how this procedure was conducted.

Standardization of the index test

Describe whether always the same instrument was used for all recordings with the index test.

Conflict of interest Describe whether the authors or any of the stakeholders had a conflict of interest for using a specific index test and whether this could have led to biased outcomes.

Other procedural variables Describe whether other procedural variables could have influenced the outcomes of the index test, e.g., pain, high number of implant fractures, non-continuous implant insertion etc.


102

QUADAS-2 Collection form 7. Domain 3: Reference standard “risk of bias” and “concerns of applicability”56,72,82* Entry Description and character of the information Experience of the operator Describe the experience of the operator, i.e., the operator that records the reference

standard, e.g., the radiology technician Foreknowledge of the results of the index test

Describe whether operators had foreknowledge of results of the index test.

Foreknowledge of intermediate test results

Describe whether operators that conduct the reference standard had foreknowledge of results of intermediate test results, intermediate recordings with the index test, i.e., insertion torque or other intermediate tests, e.g., resonance frequency assessments.

Quality reference standard Describe the quality of the reference standard, e.g., 2D versus 3D radiographs. Assess whether the quality of the reference standard is sufficient to provide an accurate reading. 3D images are the preferred standard

Number of outcome assessors Describe the number of outcome assessors. Ideally more than one operator assesses the outcomes in order to avoid inadequate readings or exclusions of readings.

Calibration of the reference standard

Describe whether the reference standard was calibrated and how this procedure was conducted.

Calibration of the operators

Describe whether operators were calibrated prior to classifying the target condition and how this procedure was conducted

Intra-or inter-operator differences

Present the intra-or inter operator differences. Assess whether these differences are small enough to be ignored.

Standardization of the reference standard

Describe whether always the same instrument was used for all recordings with the reference standard.

Co-interventions Describe whether co-interventions were conducted between the index test and the reference standard. Assess whether these interventions could influence the results of the reference standard.

Defining the target condition Describe how the target condition was defined, e.g., no root contact, root contact without root penetration, root contact with penetration.

Conflict of interest Describe whether the authors or any of the stakeholders had a conflict of interest for using a specific reference standard and whether this could have led to biased outcomes.

Other procedural variables Describe whether other procedural variables could have influenced the outcomes of the reference standard.

*Locate where information on each item can be found, e.g., Page 12 column 3. QUADAS-2 Collection form 8. Domain 4: Flow and timing “risk of bias”56,72,82*

Entry Description and character of the information Time interval index test reference standard

Describe the time interval between the index test(s) and the reference standard. Describe whether this interval was correct or too long.

All patients both tests Describe whether all patients received both the index test and the reference standard. Control in flow diagram.

Inclusion in analysis Describe whether all patients tested were also included in the data analysis Withdrawals Describe whether patients withdrew and present the explanation for this withdrawal.


103

Additional file 4: QUADAS-2 quality assessment tool tailored to the clinical question80,82

Phase 1: State the review question:*

Patients (setting, intended use of index test, presentation, prior testing): Index test(s): Reference standard and target condition:

*The review question and the eligibility criteria are defined in the protocol of this systematic review

Phase 2: Draw a flow diagram for the primary study

104

Phase 3: Risk of bias and applicability judgments QUADAS-2 is structured so that 4 key domains are each rated in terms of the risk of bias and the concern regarding applicability to the research question (as defined above). Each key domain has a set of signaling questions to help reach the judgments regarding bias and applicability.

DOMAIN 1: PATIENT SELECTION A. Risk of Bias Describe methods of patient selection:

Signaling questions regarding the patients selected in the research study:

Was a consecutive or random sample of patients enrolled? Yes/No/Unclear Was a case-control design avoided? Yes/No/Unclear Were the total number of patients adequate to answer the research question? (Was a power calculation conducted?)

Yes/No/Unclear

Did the study avoid inappropriate exclusions? Yes/No/Unclear Could the selection of patients have introduced bias ? RISK: LOW/HIGH/UNCLEAR B. Concerns regarding applicability Describe included patients (previous testing, severity of the target condition, intended use of the index test, demographic features, presence of differential diagnosis or comorbid conditions, setting of the study):

Signaling questions regarding the patients selected in the research study compared to those eligible for the review questions : (These signaling questions are only used to facilitate making a judgment regarding applicability)

Could demographic features, gender, age, ethnicity, knowledge of languages, socio-economic factors, location, disabilities, and mobility up to the moment of starting treatment in the selected patients influence the applicability?

Yes/No/Unclear

Could the medical and dental health conditions up to moment of conducting the index test, e.g., the medical and dental health status in the selected patients influence the applicability?

Yes/No/Unclear

Could the setting of the selected patients influence applicability? Yes/No/Unclear Could co-existent conditions in the selected patients influence applicability? Yes/No/Unclear Could previous testing in the selected patients influence applicability? Could the severity of the target condition in the selected patients influence applicability?

Yes/No/Unclear

Could the intended use of the index in the selected patients influence applicability? Yes/No/Unclear Could other comorbid conditions up to the moment of conducting the index test in the selected patients influence applicability, e.g., change in patient’s health, implant (e.g., implant design and dimensions)?

Yes/No/Unclear

Is there concern that the included patients CONCERN: LOW/HIGH/UNCLEAR do not match those targeted by the review question?

105

DOMAIN 2: INDEX TEST(S) If more than one index test was used, please complete for each test A. Risk of Bias Describe the index test and how it was conducted and interpreted:

Signaling questions regarding the selected index test in the research study:

Were biased outcomes of the index test caused by implant, location or surgery related factors avoided?

Yes/No/Unclear

Was the insertion procedure (which includes the torque recording with the index test) conducted by an experienced operator(s)?

Yes/No/Unclear

Were the index test results interpreted without knowledge of the results of the reference standard?

Yes/No/Unclear

Were the index test results interpreted without knowledge of other intermediate test results?

Yes/No/Unclear

If a threshold was used, was it pre-specified? Yes/No/Unclear In the research study is the outcome, insertion torque, measured during the final rotation of the screw ?

Yes/No/Unclear

Is the quality of the torque recording accurate (assess the quality of the torque sensor, e.g., mechanical versus digital, design, easy to use)?

Yes/No/Unclear

Was the outcome of the index test assessed by more than one outcome assessors? Yes/No/Unclear Was the torque sensor calibrated? Yes/No/Unclear Were always the same instruments used for the index test? Yes/No/Unclear Was there no conflict of interest in using the index test? Yes/No/Unclear Were there no other procedural variables that could have influenced the outcomes of the index test, e.g., implant, location, and surgery related factors?

Yes/No/Unclear

Could the conduct or interpretation of the index test RISK: LOW/HIGH/UNCLEAR have introduced bias? B. Concerns regarding applicability Describe variations in test technology, execution, or interpretation of a test:

Is there concern that the index test, its conduct, CONCERN: LOW/HIGH/UNCLEAR or its interpretation differ from the review question ?

106

DOMAIN 3: REFERENCE STANDARD A. Risk of Bias Describe the reference standard and how it was conducted and interpreted:

Signaling questions regarding the selected reference standard in the research study:

Was the reference standard conducted by an experienced operator(s)? Yes/No/Unclear Were the reference standard results interpreted without knowledge of the results of the index test(s)?

Yes/No/Unclear

Were the reference standard results interpreted without knowledge of other intermediate test results?

Yes/No/Unclear

Is the reference standard likely to correctly classify the target condition (compare 2D versus 3D radiographs)?

Yes/No/Unclear

Did more than 1 operator classify the target condition? Yes/No/Unclear Was the reference standard machinery calibrated? Yes/No/Unclear Were the operators calibrated prior to classifying the target condition ? Yes/No/Unclear Were intra-or inter-operator differences in classifying the target condition sufficiently small that they can be ignored?

Yes/No/Unclear

Were always the same instruments used for the reference standard? Yes/No/Unclear Were no specific interventions conducted between the index test(s) and the reference standard that could influence the diagnosis of the target condition?

Yes/No/Unclear

Did the study present a clear definition of the target condition? Yes/No/Unclear Was there no conflict of interest in using the reference standard? Yes/No/Unclear Were there no other procedural variables that could have influenced the outcomes of the reference standard?

Yes/No/Unclear

Could the reference standard, its conduct, or its RISK: LOW/HIGH/UNCLEAR interpretation have introduced bias? B. Concerns regarding applicability Describe the target condition

Is there concern that the target condition CONCERN: LOW/HIGH/UNCLEAR as defined by the reference standard does not match the review question?

107

DOMAIN 4: FLOW AND TIMING A. Risk of Bias Describe any patients who did not receive the index test(s) and/or reference standard or who were excluded from the 2 x 2 table (refer to flow diagram):

Describe the time interval and any interventions between index test(s) and reference standard:

Signaling questions:

Was there an appropriate interval between index test(s) and reference standard? Yes/No/Unclear Did all patients (implants) receive a reference standard? Yes/No/Unclear Did all patients (implants) receive the same reference standard? Yes/No/Unclear Were all patients (implants) included in the analysis? Yes/No/Unclear Were withdrawals explained? Yes/No/Unclear

Could the patient flow have introduced bias? RISK: LOW/HIGH/UNCLEAR

108

Additional file 5: Tabular presentation for QUADAS-2 results of the selected studies80. Exemplary table for the presentation of the QUADAS-2 scores.

STUDY RISK OF BIAS APPLICABILITY CONCERNS

Patient selection

Index test

Reference standard

Flow and timing

Patient selection

Index test Reference standard

Low High ? Unclear

Additional file 6: Protocols for conducting meta-analyses and assessing statistical heterogeneity. Protocol for conducting meta-analyses In Cochrane diagnostic accuracy reviews 2 strategies for summary statistics are used: (1) estimating a summary sensitivity and specificity at a chosen common threshold . Studies that do not report these values at any of these thresholds will be excluded from the meta-analysis88; (2) estimating the expected summary Receiver Operating Characteristic (ROC) curve across a variety of thresholds. This analysis only includes one threshold per study88. The choice between these 2 strategies will depend on the variation of these thresholds in the eligible studies88. If both strategies could complement each other, SROC curves and average operating points will be both displayed. Revman 5.3 will be used to create these figures for each experimental model and for each specific analysis89. The summary ROC plots will present a single sensitivity-specificity point with its confidence intervals for the outcomes of each individual study88. When pooling of outcomes is possible, summary statistics will be added to these graphs. Paired forest plots will present the sensitivity and specificity together with their 95% confidence intervals88. In these plots, studies will be ordered to facilitate a visual representation of a possible association between sensitivity and specificity. If such an association is lacking, studies will be presented in alphabetical order88. Summary statistics will not be added to these coupled forest plots88. A random effects analysis will be the most suitable model for the meta-analysis, because heterogeneity is expected in studies of diagnostic test accuracy. The fixed effect model will only be used when the variability between studies is small and explicable by chance88. Review manager 5.3 will be used for the meta-analysis, but a statistician will also be consulted for potential external software89. For the secondary research question, we will conduct a fixed-effect meta-analysis when the effects of interventions of all eligible studies are similar in both magnitude and direction. Random effects models are applied when the effect sizes are not identical, but related and follow some distribution90. The mean difference between maximum insertion torque values with and without root contact will be the effect measure for this meta-analysis. Review manager 5.3 will also be used for the meta-analysis for the secondary research question89. Protocol for assessing statistical heterogeneity For the primary research question both the bivariate and hierarchical SROC (HSROC) models will be used to investigate heterogeneity by assessing the effects of covariates on summary sensitivity and

109

specificity and the shape and position of the summary ROC curve88,92,93. When only a small number of studies will be available these statistical models will be either avoided or used with caution88. For the secondary research question a variety of statistics: Q, p, T2, T, and I2 are used to test for heterogeneity90,94-97. In the presence of minimal, moderate, or substantial heterogeneity, i.e., I2 in the range of 0-90%, the dispersion of outcomes will be presented in a forest plot with or without the summary diamond90,91,98. The following strategies will be considered for dealing with (statistical) heterogeneity90: 1) controlling the correctness of the extracted data; 2) not conducting a meta-analysis. In the presence of an I2 higher than 90%, a narrative format will be applied for this systematic review, because conducting a meta-analysis in the presence of excessive heterogeneity could mislead the reader 91,99; 3) further exploring the causes of heterogeneity and its impact on the treatment effects through subgroup analyses and meta-regression90; 4) ignoring heterogeneity by applying a fixed-effect meta-analysis; 5) incorporating heterogeneity by applying a random-effects meta-analysis; 6) changing the effect measure, e.g., recording torque values at different time points during the insertion process; 7) excluding studies90.

110

Protocol 5

Barriers and facilitators to the implementation of orthodontic mini-implants in clinical practice:

a protocol for a systematic review and meta-analysis

This is the non-edited protocol for the manuscript: Meursinge Reynders RA, Ronchi L, Ladu L, Di Girolamo N, De Lange J, Roberts N, Mickan S. Barriers and facilitators to the implementation of orthodontic mini-implants in clinical practice: A systematic review. Status: Submitted This non-edited protocol was published as: Meursinge Reynders RA, Ronchi L, Ladu L, Di Girolamo N, De Lange J, Roberts N, Mickan S. Barriers and facilitators to the implementation of orthodontic mini-implants in clinical practice: A protocol for a systematic review. Syst Rev. 2016 Feb 5;5(1):22.

http://www.ncbi.nlm.nih.gov/pubmed/?term=barriers+and+facilitators+orthodontic+mini-iimplants

111

Abstract Background: Most orthodontic treatment plans need some form of anchorage to control the reciprocal forces of tooth movement. Orthodontic mini implants (OMIs) have been hailed for having revolutionized orthodontics, because they provide anchorage without depending on the collaboration of patients, they have a favorable effectiveness compared with conventional anchorage devices, and they can be used for a wide scale of treatment objectives. However, surveys have shown that many orthodontists never or rarely use them. To understand the rationale behind this knowledge-to-action gap we will conduct a systematic review that will identify and quantify potential barriers and facilitators to the implementation of OMIs in clinical practice for all potential stakeholders, i.e., patients, family members, clinicians, office staff, clinic owners, policy makers etc. The prevalence of clinicians that do not use OMIs will be our secondary outcome. Methods: The Preferred Reporting Items for Systematic review and Meta-Analysis Protocols (PRISMA-P) 2015 Statement was adopted as the framework for reporting this manuscript. We will apply broad-spectrum search strategies and will search MEDLINE and more than 40 other databases. We will conduct searches in the grey literature, screen reference lists, and hand-search 12 journals. All study designs, stakeholders, interventions, settings, and languages will be eligible. We will search studies that report on barriers or facilitators to the implementation of OMIs in clinical practice. Implementation constructs and their prevalence among pertinent stakeholders will be our primary outcomes. All searching and data extraction procedures will be conducted by 3 experienced reviewers. We will also contact authors and investigators to obtain additional information on data items and unidentified studies. Risk of bias will be scored with tools designed for the specific study designs. We will assess heterogeneity, meta-biases, and the robustness of the overall evidence of outcomes. We will present findings in a systematic narrative synthesis and plan meta-analyses when pertinent criteria are met. Discussion: Knowledge creation on this research topic could identify and quantify both expected and unexpected implementation constructs and their stakeholders. Such knowledge can help develop strategies to address implementation issues and redirect future studies on OMIs towards knowledge translation. This could lead to improved patient-health experiences and a reduction in research waste.

Keywords Mini-implant, screw, orthodontics, implementation, knowledge translation, barriers, facilitators, contacting authors, systematic review.

112

Background The introduction of a new technique into a health-care system is a complex process, depends on the successful interaction between a variety of stakeholders, but often fails1-4. This failure is a global problem and has created a knowledge-to-action (KTA) gap, which is the gap between evidence-based knowledge and the use of this information in practice5. This gap also applies to orthodontic mini-implants (OMIs), because the implementation of this health technology into clinical practice is often suboptimal6-9. To address this issue, it is important to identify and quantify barriers and facilitators to the implementation of OMIs. This systematic review will focus on these objectives. Orthodontists use various types of anchorage to control the reciprocal forces of tooth movement. Anchorage is necessary in most orthodontic treatment plans and is usually obtained by applying forces to groups of teeth or through extra-oral sources, for example the neck or cranium10. However, these techniques pose serious limitations such as 1) their restricted area of application 2) they may still cause loss of anchorage, and 3) they depend on the collaboration of the patient10. OMIs are not conditioned by most of these shortcomings, are indicated for a wide variety of treatment mechanics, and can be used in both jaws over long time periods11 (Figure 1). These advantages should be considered in the context of some of limitations of these devices such as: 1) the need to conduct surgery for their placement; 2) risk factors associated with surgery; 3) implant failure and its implications; 4) costs; and 5) numerous implementation issues. Conducting surgical interventions in orthodontic offices is still very uncommon. Implementation can be conditioned by variables such as: 1) the lack of knowledge and skills of clinicians to conduct such interventional procedures; 2) the lack of knowledge-management skills of pertinent stakeholders; 3) lack of an adequate organization; 4) lack of time and resources 5) attitudes towards new knowledge; 6) perceptions of various stakeholders regarding the quality and validity of the evidence on OMIs; 7) resistance within the organization; 8) resistance from the patient5.

Figure 1. Schematic of an orthodontic mini-implant

113

The interventional procedure can be divided in 4 phases: 1) implant insertion; 2) orthodontic loading; 3) implant maintenance; and 4) implant removal. In the first phase, specific OMIs are selected according to the indicated anchorage objectives. Machine surfaced OMIs with a diameter of 1.3-2 mm and a length of 6-10 mm are currently the most frequently used (Figure 1)12. After the administration of a local anaesthetic, OMIs are inserted into either the maxillary or mandibular bone. Immediately after insertion, they are generally loaded with light orthodontic forces of 100 grams (Figure 2). The first appointment of the maintenance phase is programmed 1 week after implant insertion. Patients are subsequently checked every 4 weeks. During these maintenance visits, appliances and orthodontic forces are controlled, implant stability is assessed with an explorer, and implant maintenance related factors are re-enforced. At the completion of all anchorage objectives, OMIs are removed, usually without the need of anaesthesia. Insertion sites generally heal without complications.

Figure 2. An orthodontic mini-implant* immediately after insertion and orthodontic loading *Quattro implants PSM Medical Solutions; Tuttlingen, Germany. OMIs have been hailed for having revolutionized clinical orthodontics and systematic reviews have reported low implant failure rates and have recorded favourable effectiveness of OMI reinforced anchorage compared with anchorage obtained with conventional treatment mechanics11-14. OMIs are among the most frequently presented topics at orthodontic meetings and a 4 day world implant orthodontic conference is organized annually with OMIs as its exclusive topic15. In addition, the number of publications on OMIs has increased exponentially since the introduction of OMIs by Kanomi in November 1997 (Figure 3) and numerous orthodontic implant companies have been founded16,17.

114

Figure 3. Scatter plot of the number of articles on OMIs identified per year since 1997 in Pubmed (MEDLINE) Search strategy in PubMed: ((orthodont* AND (mini implant* OR micro implant* OR microimplant* OR screw* OR mini screw* OR miniscrew* OR micro screw* OR microscrew* OR temporary anchorage device*)) However, surveys in a variety of countries have shown that many orthodontists never or rarely use these devices6-9,18. This KTA gap is a surprise for a technique that was introduced almost 20 years ago and that has the advantageous characteristics described previously11-13,16. This is further surprising, because anchorage systems are indicated in most orthodontic treatment plans10. A recent survey showed that more than 75% of the surveyed doctors would include OMIs in a treatment plan for a common orthodontic patient19. To understand the rationale behind the KTA gap it is necessary to identify and quantify barriers and facilitators to the implementation of OMIs (the implementation constructs) for both demand-side stakeholders, i.e., orthodontic patients and their family members and potential supply-side stakeholders, e.g., clinicians, office staff, clinic owners, researchers, guideline developers, policy makers, implant companies etc. Barriers to implementation could include: the invasiveness of the procedure, learning a new technique, fear of complications, financial obstacles, the large volume of research evidence, the lack of trust in research evidence, the applicability of the new health technology to a local context, etc.6,8,20-22. Facilitators to implementation could include: shortened treatment time, better outcomes, improved esthetics during orthodontic treatment, recommendations by patients who had undergone treatment with OMIs etc.6,23,24. Identification and quantification of these barriers

115

and facilitators are important, because knowledge gaps can be subsequently assessed and strategies can be developed to deal with them. For our secondary objectives we will record the prevalence of clinicians that do not use OMIs in the eligible studies of our primary question. This statistic quantifies the knowledge-to-action gap for these stakeholders. Our scoping searches showed that no earlier systematic review or protocol have addressed the objectives of this systematic review protocol.

Objectives The objectives of this systematic review are as follows: Primary objectives

• To identify and quantify barriers and facilitators to the implementation of OMIs for all potential stakeholders such as patients and their family members, clinicians, office staff, clinic owners, researchers, guideline developers, policy makers, implant companies etc.

Secondary objectives

• To record the prevalence of clinicians that do not use OMIs in the studies that will be selected for the primary objectives.

Methods To develop our methods section, we consulted (1) conceptual models for assessing barriers and facilitators to knowledge use20,25-28; (2) guidelines and handbooks for conducting quantitative and qualitative systematic reviews29-32; (3) checklists for reporting research studies33; (4) systematic reviews that were developed to identify barriers and facilitators on a variety of issues and asked similar research questions as our research study2,21,22,34-37; and (5) our previous systematic reviews and a protocol of a systematic review on OMIs38-41. We adopted The Preferred Reporting Items for Systematic review and Meta-Analysis Protocols (PRISMA-P) 2015 statement as the guideline for reporting this protocol42,43. This manuscript is not registered in PROSPERO, because our research questions are not covered by the inclusion criteria of this register44.

Eligibility criteria Study designs For the primary objectives we defined the following eligibility criteria:

• To maximize the breadth of data collection, studies will not be excluded based on their research design. Studies that present an original collection of data on identified barriers and facilitators to the use of OMIs will be eligible. Such studies generally refer to interviews, focus groups, surveys, and questionnaires with any of the pertinent stakeholders22.

• Studies that identify or quantify barriers and facilitators to knowledge use as a primary or as an additional objective of a larger study will be both eligible. For example, a study that addresses our research questions nested in a larger research model such as mixed methods will also be eligible.

116

For the secondary objectives we defined the following eligibility criteria: • Only quantitative studies, for example surveys, that addressed the primary objectives of this

systematic review will be eligible. Stakeholders (participants) Barriers to implementation can vary for different stakeholders. To avoid inappropriate exclusion of pertinent barriers, we will apply broad-spectrum eligibility criteria that include any possible demand-and supply side stakeholders. The former group refers to orthodontic patients and their family members. The latter stakeholders refer to clinicians, office staff, clinic owners, researchers, guideline developers, policy makers, and implant companies etc. Interventions Interventions that use one or more implants with diameters smaller than 2.5 mm for orthodontic anchorage purposes will be eligible. No restrictions will be applied to implant length or design, the connection to plates, the location of insertion, the type of insertion procedure, the type of orthodontic loading, and the type of implant maintenance. Interventions with OMIs on patients of either sex, and in any age or demographic group will be eligible. Outcomes Any type of barrier or facilitator to the use of OMIs in clinical practice will be our primary outcome. A barrier is defined as any variable that impedes or obstructs their use. A facilitator is defined as any variable that eases and promotes the use of OMIs. Examples of barriers and facilitators were presented in the introduction. Barriers and facilitators are eligible as outcomes when they are described as implementation constructs by the eligible stakeholders37. For example patient’s perceptions of the interventional procedure or assessments of health experiences such as pain and discomfort during implant insertion will not be considered as eligible outcomes when they are not specifically defined as barriers to the use of OMIs by these patients. These eligibility criteria avoid mislabelling of implementation constructs during qualitative analyses as a result of bias or misinterpretation of outcomes by systematic reviewers. Setting and language No setting and language restrictions will be applied.

Information sources Information sources will be searched from January 1 1997, the year of the first publication on OMIs, onwards16. We adopted a variety of information sources from previous systematic reviews on OMIs38-

41. TRIP and SUMSearch databases were also consulted to identify pertinent search engines45,46. Electronic searches

• The following general and subject-specific electronic databases will be searched: Google Scholar Beta, PubMed (MEDLINE), EMBASE (Ovid), Cochrane Central Register of Controlled Trials (CENTRAL), CINAHL, PsycINFO, Sociological Abstracts, and PROSPERO47-52.

• The “Related Articles” feature in PubMed will be consulted. • The following Web of Science Core collection citation indexes will be searched: Science

Citation Index Expanded (SCI-EXPANDED); ARTS & Humanities Citation Index (A&HCI); and Social Sciences Citation Index (SSCI)49,53,54.

• A series of national and regional databases will also be searched: African Index Medicus, African Journals online (AJOL), Informit Health Collection, Index Medicus for the Eastern

117

Mediterranean Region, IndMED, KoreaMed, LILACS, Index Medicus for the South-East Asia Region (IMSEAR), Western Pacific Region Index Medicus (WPRIM)49,53.

Searching other resources Grey literature We will consult the following grey databases49,50,53,55:

• General databases: Open Grey, Google Scholar Beta, The National Technical Information Service (NTIS), and The Health Management Information Consortium (HMIC).

• Dissertations and databases of theses: ProQuest and Dissertations and Theses (Global full text plus UK and Ireland abstracts).

• Conference proceedings and abstracts: Meeting Abstracts, ISI Conference Proceedings, IEEE Conference Proceedings, and Google Scholar.

• Review databases: Database of Abstracts of Reviews of Effects (DARE), Health Technology Assessment database (HTA), and NHS EED all via the Centre for Reviews and Dissemination (CRD), and Turning Research into Practice (TRIP).

• Guidelines will be searched in MEDLINE, EMBASE, and TRIP. In addition, evidence-based guidelines of the following organizations will be consulted: Australian National Health and Medical Research Council, Canadian Medical Association, National Guideline Clearinghouse, National Library of Medicine Guidelines, New Zealand Guidelines Group, and NICE Clinical Guidelines.

• We will search Pubmed and Ovid for Citation alerts. Handsearching

• American Journal of Orthodontics & Dentofacial Orthopedics, Angle Orthodontist, Australian Journal of Orthodontics, European Journal of Orthodontics, International Journal of Adult Orthodontics and Orthognathic surgery, Journal of Clinical Orthodontics, Journal of Orthodontics, Journal of the World Federation of Orthodontics, Orthodontics & Craniofacial Research, Seminars in Orthodontics.

Reference lists

• We will also manually screen the reference lists of all selected articles, reviews, and guidelines for eligible papers that are not identified during the other searching procedures.

Correspondence

• A variety of stakeholders, e.g., subject specialists, authors of selected papers and pertinent systematic reviews, researchers on OMIs and manufacturers of implants, will be contacted to identify ongoing or unpublished research studies53,56.

Search strategy

• Methods to find pertinent subject headings and key words are adopted from our previous systematic reviews on OMIs38-41. For the electronic database searches we will use variations of search terms for the field of interest (orthodontics) and the intervention of interest (mini implants). To avoid the inappropriate exclusion of pertinent studies, we will not include “the outcomes of interest” as a selection criteria and will aim at a broad-spectrum search strategy. Search terms are identified in collaboration with an information specialist (NR) and include: orthodontic(s), orthodontist(s), implant(s), mini implant(s), micro implant(s), microimplant(s), screw(s), mini screw(s), miniscrew(s), micro screw(s), microscrew(s), temporary anchorage device(s).

118

• Search strategies will be developed specifically for each database and will be subsequently pilot tested and fine-tuned53.

• An information specialist (NR) will assist with the development of these search strategies. • To avoid the incorrect exclusion of eligible studies The Boolean ‘NOT’ operator will not be used. • The search strategies of all general and subject-specific electronic databases will be listed in a

table together with the search dates and the number of identified items. Examples of the search strategy of MEDLINE and Google Scholar Beta are presented in Table 149,57.

Table 1. Search strategy for the MEDLINE and Google Scholar Beta databases $

PubMed (Medline) orthodont* AND (implant* OR mini implant* OR micro implant* OR microimplant* OR screw* OR mini screw* OR miniscrew* OR micro screw* OR microscrew* OR temporary anchorage device*)

Google Scholar Beta $$

(orthodontics OR orthodontic OR orthodontist OR orthodontists) (implant OR implants OR “mini implant” OR “mini implants” OR screw OR screws OR “mini screw” OR “mini screws” OR “miniscrew” OR “miniscrews” OR “microscrew” OR “temporary anchorage device”)

$ For each database we will adapt the pertinent characters for the exploration and truncation of the search terms. All search strategies will be copied and pasted directly into the search box of the search engines. This procedure will be applied, because re-typing of search terms could introduce errors in the search strategy49. $$ We will use this shortened search strategy, because the search string of Google Scholar is limited to under 256 characters57.

Study records Data management To reduce inter-examiner disagreements on study eligibility we will adopt the procedures described in the Cochrane Handbook for Systematic Reviews of Interventions and in the PRISMA-P 2015 statement42,43,58. Prior to starting the formal study selection process, we will pilot test our selection procedures on a sample of abstracts. These calibration procedures are conducted to clarify and potentially fine-tune our selection criteria and to apply them consistently. All 3 review authors (RMR, LR, and LL) will participate in these calibration exercises. Selection process

• We consulted the study selection procedures of our previous systematic reviews on OMIs to develop this section38-41. We will select studies that fulfil our eligibility criteria.

• Studies will be selected independently by 3 experienced systematic reviewers (RMR, LR, and LL), who are also topic experts.

• We will screen titles and abstracts for eligible studies. Each selected abstract will be linked to the data source of origin. Full texts of potentially relevant articles will be subsequently reviewed. To reduce the risk of inappropriate exclusion, ambiguous articles will also be assessed for eligibility.

• Unpublished research studies, e.g. those found in grey literature data bases, will also be reviewed for eligibility by the 3 reviewers. Such studies are only considered when sufficient data are reported to permit peer-reviewing.

• Authors will be contacted when potential multiple publications of their research data are identified. We will apply our protocol for contacting authors that we described in a recent protocol for a systematic review40. Characteristics of studies suspect of multiple publications include: (1) studies with a retrospective research design and that use similar methodology; (2) overlapping authors in similar research studies and placed in a different order; (3) the publication of similar research findings within a short time span in different journals; and (4) reference lists of the multiple publication articles tend to exclude the references of the other similar publications40.

119

• Disagreements between authors on eligibility will be resolved through discussions. Persisting disagreements will be addressed through consultations with a fourth author (NDG) or through our protocol for contacting authors56,59. A detailed description of our protocol for contacting authors is described in Additional file 1 and is based on one of our recent systematic review protocols40. The Cochrane Glossary will be consulted to avoid misinterpretation of terminology used in the email correspondence with authors60.

• Study selection procedures will be presented in a PRISMA flow diagram56,61. Excluded studies together with the rationale for their exclusion will be listed in a table.

Data collection process

• Prior to the formal study selection and data extraction process, a list of ‘potential’ barriers and facilitators to the implementation of OMIs will be developed. Items on this list will be extracted from 3 groups of publications: (1) systematic reviews that focussed on the identification of barriers and facilitators to the implementation of health-related issues and technologies2,21,22,34-37 (2) conceptual models for assessing barriers and facilitators to knowledge use20,25-28 and (3) our previous systematic reviews on OMIs38-41.

• These articles will be examined by all three reviewers (RMR, LR, and LL), which are all topic experts. Each of these operators will develop a list of barriers and facilitators based on the findings in these publications that could be pertinent to the implementation of OMIs. These lists will be subsequently discussed between these 3 reviewers and a final summary list of ‘potential’ barriers and facilitators will be created. We will also link specific stakeholders, e.g., patients, clinicians, office staff etc. to each of these variables. In this list ‘potential’ barriers and facilitators are presented as a series of constructs and are classified according to 5 domains: intervention characteristics, outer setting, inner setting, characteristics of the individuals involved, and the process of implementation25. We adopted this classification from the Consolidated Framework For Implementation Research (CFIR)25.

• Our procedures for the identification of ‘potential’ barriers and facilitators will be conducted prior to the study selection and data extraction process and are only used as a calibration exercise for the 3 reviewers and to increase their background knowledge. Our list of ‘potential’ barriers and facilitators will not be used as a reference checklist during the study selection and data collection process, because this could result in the inappropriate exclusion of ‘unexpected’ barriers and facilitators to the implementation of OMIs in clinical practice.

• For the development of our data extraction forms we explored the reporting checklists of pertinent research designs of the Equator network33. We also consulted data collection forms in previous systematic reviews on OMIs38-41 and the 3 groups of publications that were used to develop the list of potential implementation constructs. Pertinent items for the extraction of data for the secondary research question were also explored during this research phase.

• Data extraction forms were first pilot tested on a series of articles by 3 reviewers (RMR, LR, and LL), and subsequently fine- tuned. These procedures were also used to calibrate reviewers. The pilot-tested data extraction forms are presented in Additional file 243.

• Data extraction procedures will be conducted independently by the 3 aforementioned operators, who are experienced systematic reviewers and topic experts. Disagreements on extracted items will be resolved through rereading and discussions and if necessary an arbitrator (NDG) will be consulted to adjudicate these disagreements. All data extraction procedures will be similar for our primary and secondary research questions.

Data items

• We decided to extract a broad spectrum of data items and define them in great detail, because revisiting all selected papers as a result of inaccurate or deficient listings of data items is a waste of ‘reviewers’ time and could introduce mistakes43.

120

• Extracted data items for our primary and secondary objectives include: the source, eligibility, duplicate publication, the study design, selection procedures, stakeholders, the setting, interventions, outcomes, flow and timing, adverse effects, withdrawals and missing outcomes, the funding, and miscellaneous data of the selected studies42,43. Many of these items are further subdivided and all extracted entries are listed in tables in Additional file 2. Descriptions of each item are presented in these tables and entries that could bias the outcomes are also recorded.

• Diagrams will be created to depict the flow of the stakeholders and the timing of the various research phases from the start of the selection procedures to the completion of the recording of outcomes (Additional file 2).

• Data from articles that appear ‘suspect’ of multiple publication of the same research studies will be extracted identically as those from articles that are not suspect. Potential overlap of research data will be subsequently assessed.

• When during the review process new relevant data items will be identified, we will collect them and will report these changes of the protocol and their rationales.

• Authors of selected studies will be contacted to find information on unclear or missing items and to resolve remaining disputes between reviewers (Additional file 1). Persistent disagreements will be reported.

Outcomes and prioritization Primary outcomes

• The primary outcomes will be all barriers and facilitators to the implementation of OMIs in clinical practice identified by all demand-and supply-side stakeholders. We will record the prevalence of these implementation constructs among pertinent stakeholders.

• Pertinent stakeholders are defined in Table 2 and are further subdivided in ‘users’ and ‘non users’. A barrier is defined as any variable that impedes or obstructs the use of OMIs. A facilitator is defined as any variable that eases and promotes their use.

• All primary outcomes will be presented with an explicit description according to the author(s) of the pertinent eligible study. We will report whether these outcomes refer to ’specified’ or ‘non-specified’ interventions (Table 2).

• When studies record our primary outcomes at different time points, we have decided a priori to subdivide them as pre-, immediate post-, and long-term post-intervention recordings (Table 2). We will also present the setting, country, and design of the research study.

• The pre-intervention recording of barriers or facilitators to a ‘specified’ intervention with OMIs for ‘non user’ clinicians or for patients that have not undergone this intervention previously in any type of setting or research design will be our ‘preferred’ primary outcome.

• The prevalence of identified barriers and facilitators among the surveyed or interviewed pertinent stakeholders will be calculated as follows:

Prevalence of an identified barrier or facilitator = The number of stakeholders that scored a particular construct as a barrier or facilitator to the implementation of OMIs in clinical practice/ The total number of stakeholders that scored on the role of this particular construct as a barrier or facilitator to the implementation of OMIs in clinical practice This prevalence will be presented for example as: 30/50

• ‘Characteristics and findings of included studies’ tables are presented in Additional file 3.These tables list barriers and facilitators to the implementation of OMIs (implementation constructs), prevalence statistics, and other pertinent items such as the type of research design, stakeholders, settings, interventions, and outcomes, and time points for recording them. The

121

procedures to extract and categorize primary outcomes and anticipated exemplary tables of categorized implementation constructs are also presented (Additional file 3)43.

Secondary outcomes

• The secondary outcomes will be the prevalence of clinicians that do not use OMIs and represents the knowledge-to-action gap. This statistic will be calculated as follows:

The prevalence of clinicians that do not use OMIs = The number of clinicians that do not use OMIs/ The total number of surveyed clinicians that reported on the use of OMIs in clinical practice Information that could give further insights in the understanding of the knowledge-to-action gap, e.g., the number of implants placed per clinician per year will also be recorded. Table 2. Potential subgroups for which outcomes are recorded

General subgroups Specific subgroups Stakeholders

Demand-side stakeholders: orthodontic patients and their family members Supply-side stakeholders: clinicians, office staff, clinic owners, researchers, guideline developers, policy makers, implant companies etc. Users/non users: both demand-and supply-side stakeholders can be further subdivided in those that have used OMIs previously (users) and those that do not use these devices (non-users)

Interventions Specified interventions: these interventions refer to a specific phase or type of the interventional procedure. Phases of the intervention refer to the: anesthetics, implant insertion, orthodontic treatment with OMIs, implant removal, or the healing phase. Types of interventions refer to the: implant type and dimensions, number of implants, use of plates, the surgical procedure, implant location, timing and forces of orthodontic loading etc.41. ‘Non specified’ interventions: these interventions refer to “any orthodontic treatment with OMIs”. Additional information on the specific phase or type of the interventional procedure is not provided

Time points

Pre-intervention recordings, i.e., recordings of outcomes prior to the interventional procedure Immediate post-intervention recordings, i.e., recordings of outcomes within 2 weeks after the completion of the interventional procedure Long-term post-intervention recordings, i.e., recordings of outcomes more than 2 weeks after the completion of the interventional procedure.

Setting/Country Private practice: stakeholders treated or working in a private practice University setting: stakeholders treated or working in a university clinic Country: stakeholders treated or working in a specific country

Research design Surveys or questionnaires: outcomes obtained through either surveys or questionnaires Interviews: outcomes obtained from interviews with stakeholders Focus groups: outcomes obtained from focus groups with stakeholders

Risk of bias in individual studies The methodological quality of each eligible study will be assessed with critical appraisal tools that are specific for the type of research design used in that study. For qualitative studies such as focus groups and interviews we will use the Joanna Briggs Institute Qualitative Assessment and Review Instrument (JBI QARI)30. For quantitative studies such as surveys and questionnaires we will use The Joanna Briggs Institute critical appraisal tool for studies reporting prevalence and incidence data31,35,62. These instruments are listed in ‘Additional file 4’. The guidelines for scoring these tools will be first examined by the 3 reviewers (RMR, LR, LL)30,31,62,63. A series of studies will be then used to calibrate these reviewers for each appraisal tool. Eligible studies will be subsequently scored independently by these

122

authors. To facilitate the comparisons of appraisal scores, all three authors will record the rationale for each of these scores and the location in the article. Initial critical appraisals are conducted during the data extraction phase of each eligible study and are revisited and fine-tuned after the completion of these procedures for all selected studies. Scoring differences between reviewers are resolved through discussions. A fourth reviewer (NDG) is called upon in the case of disagreement between reviewers. Authors of eligible studies are contacted in the case of persistent disagreements on appraisal scores (Additional file 1). The critical appraisal scores for each selected study will be listed in tables and for each appraisal tool separately (Additional file 4)30,31. We will calculate the prevalence of ‘Yes’ scores (Number of ‘Yes’/Number of articles) for each individual appraisal question35. No attempts will be made to calculate overall appraisal scores. The potential influence of each of the scored answers on the outcomes of each selected study will be weighted during the data synthesis and will be used to assess the overall strength of evidence of the review (See ‘Confidence in cumulative evidence’)43.

Data synthesis Criteria for a quantitative synthesis The prevalence statistics of our primary and secondary outcomes can both be synthesized quantitatively. A random-effects model meta-analysis will be indicated, because between study variance is expected for both outcomes. However, we will only conduct a meta-analysis when: 1) the risk of bias in the eligible studies is low; 2) outcomes are consistent between studies; 3) publication bias is low; 4) a high number of studies is included; and 5) heterogeneity is low64-66. Summary measures for a quantitative synthesis The prevalence data for our primary outcomes will be presented as event rates, e.g. 0.70, which indicates that 70 stakeholders scored a particular construct as a barrier to the implementation of OMIs out of a total sample of 100 stakeholders that scored on the role of this particular construct as a barrier to the implementation of OMIs in clinical practice. Event rates will also be recorded for our secondary outcomes, which represent the number of clinicians that do not use OMIs/the total number of surveyed clinicians that reported on the use of OMIs in clinical practice. These statistics are quantitatively synthesized and the summary event rate will be calculated and presented with the p-value and the 95% confidence interval (CI) for both primary and secondary outcomes. Comprehensive Meta-analysis (CMA) software will be used to conduct all statistical analyses in this systematic review67,68. Forest plots will be used to display these calculations and their dispersion. Unit-of-analysis issues for a quantitative synthesis To deal with unit-of-analysis issues, we will assess at which level randomization was conducted65. We will assess whether all participants underwent the same intervention, multiple interventions, and whether outcomes were recorded at different or multiple time points65. Subgroups will be created to deal with these issues (See section ‘Subgroup analyses and meta-regression). Dealing with missing data for a quantitative synthesis We will apply the protocol of the Cochrane Handbook for Systematic Reviews of Interventions for dealing with missing data69. We will first contact the pertinent authors to obtain such data (Additional file 1). We will then evaluate why data are missing and will assess whether they are missing at random or not69. Our key policy will be to include studies with missing data and assess the consequences of this inclusion in the qualitative synthesis69. For the quantitative synthesis we will weigh the following strategies for dealing with missing data: i.e., (1) analysing available data only; (2) imputing the missing data; (3) perform sensitivity analyses69. We will assess these strategies in the results and discussion

123

section of our systematic review and will discuss the potential implications for excluding missing data from the meta-analysis69. Investigation of heterogeneity We will consider 3 sources of heterogeneity: methodological, clinical, and other sources of heterogeneity65,70. These sources are selected a priori based on information from previous systematic reviews on this research topic and through discussions between the reviewers, who include both methodologists and topic experts71,38-40. These potential sources of heterogeneity are listed in additional file 541,65,66,70,72-74. The type of stakeholders, i.e. patients, clinicians, office staff etc. is excluded as a source of diversity, because outcomes are analyzed separately for each type of stakeholder. We will report when ‘post hoc’ defined sources of heterogeneity will be investigated. Statistical heterogeneity is the consequence of one or more of the sources of diversity. The presence of statistical heterogeneity is investigated by calculating Cochran’s Q, the degrees of freedom based on the number of eligible studies, and the pertinent p-value75-77. We will also calculate the following statistics: Kendall’s Tau2, Tau, and I2 68,75,77-80. These calculations, their use, and strategies for dealing with heterogeneity are explained in additional file 565.

Subgroup analyses and meta-regression Subgroup analyses are either used to explore heterogeneity or to address questions about specific stakeholders, interventions, or study designs65. For the former objective we will use the subgroups defined under ‘methodological, clinical, and ‘other’’ sources of heterogeneity in Additional file 5. For the latter objective we will assess the following ‘a priori’ subgroups, which could be used for either subsets of participants or studies65:

• Research design, i.e., surveys or questionnaires, interviews, or focus groups • Conduct and analysis of the study, i.e., high versus low risk of selection, performance,

detection, attrition, or reporting bias66. • Stakeholders, i.e., ethnicity, sex, age, and previous experience with OMIs. • Type of interventions, i.e., ‘specified’ or ‘non specified’ interventions (additional file 5). • Outcomes, i.e., pre-, immediate post-, or long-term post-intervention recordings of outcomes

(additional file 5). • Setting, i.e., private practice or university setting.

Rationales to conduct analyses with additional subgroups will be reported in the systematic review and these subgroups will be defined as ‘post hoc’. Random effects meta-regression will be used to assess differences in subgroups, but will be only conducted when there are more than ten eligible studies in the meta-analyis65. We will interpret subgroup analyses and meta-regressions with caution, because these analyses are not based on randomized comparisons and are therefore strictly observational65,81. Additional caveats of these analyses as defined by Oxman and Guyatt will also be considered81.

Sensitivity analysis We will conduct sensitivity analyses to ass the impact on outcomes of certain decisions that were made during the systematic review process. We will plan such analyses for the decisions to include: 1) small studies (compared to the other eligible studies); 2) surveys and questionnaires; 3) grey literature; 4)

124

studies at high risk of bias; and 5) studies with follow-ups beyond 6 months65. We will also conduct sensitivity analyses to assess the effects of imputing missing data65,69.

Qualitative synthesis We will provide a systematic narrative synthesis even when quantitative analyses are possible43. Our narrative synthesis will be conducted systematically and transparently to reduce the potential for bias82. We will refrain from vote counting, i.e., counting those studies that yielded a significant result and those that did not83,84. As suggested by the PRISMA-P 2015 statement43, we will adopt the ‘established methods’ for conducting systematic narrative syntheses according to the guidance of the Centre for Reviews and Dissemination (CRD)82. The CRD framework for conducting such a synthesis consists of 4 phases: (1) developing a theory why and how each barrier or facilitator could affect the implementation of OMIs for each linked stakeholder; (2) developing an initial synthesis of the findings of the eligible studies; (3) exploring relationships within and between studies; and (4) assessing the robustness of the synthesized evidence82. These steps do not have to be conducted exactly according to the order of this framework and will be conducted iteratively by the 3 topic experts (RMR, LR, LL)82. Disagreements will be resolved through discussions and persistent disagreements will be resolved through the arbitrage of a fourth author (NDG) or through contacting pertinent authors. Each phase of this synthesis will be presented in the results section of the systematic review82. Each step of this 4 phase framework for a systematic narrative synthesis is presented under here: Phase 1. Developing a theory why and how each barrier or facilitator could affect the implementation of OMIs for each linked stakeholder Theories of how barriers or facilitators could affect the implementation of OMIs for various stakeholders will be initially developed during the discussions on ‘Potential’ barriers and facilitators to the implementation of OMIs in clinical practice prior to the selection and data collection process (See section ‘Data collection process’). These theories will be fine-tuned after the data extraction procedures. We expect that most identified barriers and facilitators are obvious implementation constructs for the pertinent stakeholders and do not require much additional theory to why and how they could affect the implementation of OMIs. However, such causal links could be more difficult to explain for certain ‘unexpected’ barriers and facilitators. In those circumstances we will present ‘hypothetical’ associations descriptively or in a diagram or both82. Phase 2. Developing an initial synthesis of the findings of the eligible studies Our tables with the characteristics and findings of included studies will be consulted for this preliminary synthesis. We will also revisit the data collection tables and search for additional items that might have been overlooked and could influence this synthesis (Additional files 2 and 3). To facilitate the synthesizing process, we will reorganize data items, fine-tune textual descriptions of each included study, assess the validity of outcome measures, and create new tables with implementation constructs and pertinent stakeholders to improve visualization of the data82. We will assess whether barriers and facilitator can be linked to one of the five domains defined by25, i.e., intervention characteristics, outer setting, inner setting, characteristics of the individuals involved, and the process of implementation. The prevalence statistics of our primary outcomes will be subsequently consulted and theories about the direction and magnitude of the effects of specific implementation constructs on pertinent stakeholders will be developed. Phase 3. Exploring relationships within and between studies In this phase we will assess whether and how variables within and between studies could have influenced outcomes. For this purpose we will: (1) explore the characteristics and findings of each selected study and compare them with each other; (2) consult our tables with ‘a priori’ defined sources

125

of heterogeneity (additional file 5) and will consider sources of diversity that are identified ‘post hoc’65,70. Various tools will be used to further explore relationships within and between studies: (1) investigating statistical heterogeneity (additional file 5); (2) visualizing of statistical heterogeneity in forest plots68; (3) conducting subgroup analyses and assessing moderator variables (See section ‘Subgroup analyses and meta-regression’)82. Based on these assessments, we will modify and fine-tune our initial synthesis and will explain the rationale for these changes. Phase 4. Assessing the robustness of the synthesized evidence The robustness of the synthesized evidence will depend on: (1) the number and size of the eligible studies; (2) within and between study diversity; (3) risk of bias assessments (magnitude and direction); (4) the consistency of the outcomes between studies; (5) the magnitude of the outcomes; and (6) the presence of publication bias. To assess the robustness of this evidence we will: (1) weigh the role of these variables; (2) revisit the data collection forms and the critical appraisal tools to assess whether items have been overlooked; and (3) contact authors to obtain additional information.

Meta-biases and confidence in the cumulative evidence Meta-bias refers to the biased selection of research data and covers both reporting bias (selective outcome reporting) and publication bias43. We will apply the following strategies for dealing with reporting bias:

1) We will assess whether the protocol of a pertinent research study was published prior to the recruitment of patients43. Literature searches and screening of the Clinical Trial Register at the International Clinical Trials Registry Platform of the World Health Organization will be conducted for this purpose85.

2) When protocols will be identified, we will assess discrepancies between the outcomes planned in the protocol and those reported in the final manuscript. When no protocol will be found, we will scrutinize the eligible article for unclear or missing outcomes.

3) If necessary, research protocols or additional information on the outcomes will be requested from the authors of the research study. We will also compare overall outcomes with and without including the data obtained from contacted authors.

4) Subsequent strategies for ‘dealing with missing data for a quantitative synthesis’ were described under this heading.

We will contain the risk of publication bias through (1) our broad spectrum search strategy and extensive searches of the literature that also include the grey literature; (2) our protocol for assessing multiple publications (3) searching studies and collecting data by 3 experienced operators (4) contacting pertinent stakeholders that could provide information on unpublished or ongoing studies. We will apply a combination of 4 methods in our quantitative synthesis to assess the presence and the possible impact of publication bias86. These methods are presented under here and are all based on the assumption that publication bias increases as study sample sizes decrease86:

1) Plot the studies according to their sample size in a forest plot and assess a possible relationship between the sample size and the effect size.

2) Conduct subgroup analyses in which articles published in peer-reviewed journals are compared with those published in the grey literature and assess whether these latter publications tend to have smaller effect sizes.

3) Compare the summary outcomes of the random versus the fixed effects model to assess the role of small sample sizes on these effect measures.

4) Explore the relationship between study size and effect size by displaying these values in funnel plots and assessing their symmetry50. We will also assess the best estimate of the unbiased

126

effect size using the Duval and Tweedie’s Trim and Fill procedure87. However, these graphs will only be designed when ten or more studies are eligible50,72.

For the assessment of the strength of the body of evidence we will consult the guidelines described by the GRADE approach88 and will weigh the variables described under the subheading ‘Assessing the robustness of the synthesized evidence’ in the section ‘qualitative synthesis’. We will not score the ‘levels of evidence’ according to the GRADE approach. Our research questions do not qualify for this approach, because they do not address questions about interventions, management strategies, or policies89.

Differences between the protocol and the review All preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 and the pertinent pages are listed in additional file 642,43. We will fully report all changes in the methods during the conduct of this research study compared with those planned in the protocol. We will describe the type, timing, and the rationale of each of these modifications. We will also report the consequences of these changes on the direction, the magnitude, and the validity of the outcomes90.

Discussion Orthodontic research focuses primarily on the effectiveness of new health technologies and little information is published on their subsequent implementation in clinical practice. Scoping searches showed that this is the first protocol for a systematic review that assesses barriers and facilitators to the implementation of a new orthodontic technique. The strengths of this systematic review include: 1) it will focus on a broad spectrum of stakeholders and interventional procedures. This is important, because our findings could demonstrate that not patient-related variables, but doctor-or office staff-related constructs are the key barriers to the implementation of OMIs. Broad eligibility criteria could also reveal unexpected barriers or facilitators; 2) research procedures are reported in great detail in this protocol, which will improve transparency and will facilitate the update of this systematic review by future reviewers; 3) extensive literature searches with broad spectrum search strategies will be undertaken, which will accept low precision and will aim at high sensitivity; 4) literature searches and data extraction will be conducted independently by 3 topic experts; and 5) this research study will be conducted by experienced reviewers and methodologists, who have published various systematic reviews on OMIs38-41; A weakness of this systematic review could be that some answers to our research questions will not be addressed or remain suboptimal, because studies on implementation issues are of low quality or are deficient in the orthodontic literature. However, the identification of such knowledge gaps could be a stimulus for a new wave of research studies, that will assess why certain stakeholders refrain from using specific orthodontic therapies and which strategies can be applied to deal with these issues. Patients, clinicians, staff members, researchers, guideline developers, policy makers, and orthodontic companies will all be beneficiaries of the outcomes of this systematic review. Its findings could also apply to a variety of interventional procedures in dentistry and oral surgery. Redirecting research on OMIs towards studies that address implementation issues could ultimately reduce research waste and improve patient health experiences91-93.

127

Competing interests The authors declare that they have no financial or non-financial competing interest in relation to this manuscript.

Authors’s contributions RMR conceived and designed the study and wrote the protocol and is the guarantor. LL and LR helped with the pilot testing of the data extraction forms. NR assisted with the development of the search strategy and NDG, JDL, and SM have edited the draft protocol. All authors read and approved the final protocol.

Acknowledgements The authors wish to extend special thanks to Hisham Kalhunek for preparing the figure of the orthodontic mini-implant (Figure 1).

Funding All expenses for conducting this systematic review were paid evenly by each reviewer.

Additional files Additional file 1. Format: docx Title and description: Protocol for contacting authors Additional file 2. Format: docx Title and description: Data collection forms Additional file 3. Format: docx Title and description: Characteristics and findings of included studies Additional file 4. Format: docx Title and description: Critical appraisal tools Additional file 5. Format: docx Title and description: Heterogeneity. Sources of heterogeneity and strategies for dealing with heterogeneity Additional file 6. PRISMA-P 2015 Checklist Format: doc Title and description: PRISMA-P 2015 Checklist

128

References

1. Balas EA, Boren SA. Managing clinical knowledge for health care improvement. In: Bemmel J, McCray AT, editors. Yearbook of Medical Informatics 2000: patient-centered systems. Stuttgart, (Germany: Schattauer Verlagsgesellschaft mbH; 2000. p. 65–70.

2. Chaudoir SR, Dugan AG, Barr CH. Measuring factors affecting implementation of health innovations: a systematic review of structural, organizational, provider, patient, and innovation level measures. Implement Sci. 2013;8:22.

3. Madon T, Hofman KJ, Kupfer L, Glass RI. Public health. Implementation science. Science. 2007;318(5857):1728–9. 4. Merino GA, Lema LV. Chapter 7: Needs and demands of policy-makers. In: Garrido MV, Kristensen FB, Nielsen CP,

Busse R (editors). Health technology assessment and health policy-making in Europe: Current status, challenges, and potential. Developed by the European network for Health Technology Assessment (Eunethta). WHO Regional Office for Europe, Scherfigsvej 8, DK-2100 Copenhagen, Denmark: World Health Organization on behalf of the European Observatory on Health Systems and Policies; 2008.

5. Straus SE, Tetroe J, Graham ID. Chapter 1.1: Knowledge to action: what it is and what it isn’t. In: Straus SE, Tetroe J, Graham ID, editors. Knowledge translation in health care: moving from evidence to practice. Chichester (UK): Wiley-Blackwell; 2009.

6. Bock NC, Ruf S. Skeletal anchorage for everybody? A questionnaire study on frequency of use and clinical indications in daily practice. J Orofac Orthop. 2015;76(2):113–28.

7. Keim RG, Gottlieb EL, Vogels 3rd DS, Vogels PB. JCO study of orthodontic diagnosis and treatment procedures, Part 1: results and trends. J Clin Orthod. 2014;48(10):607–30.

8. Meeran NA, Venkatesh KG, Parveen MFJ. Current trends in miniscrew utilization among Indian orthodontists. J Orthod Science. 2012;1(2):46–50.

9. Patil A, Revankar AV. Evaluation of microimplant use in orthodontic practices in India: an opinion-based survey. J Ind Orthod Soc. 2012;46:269–72.

10. Skeggs RM, Benson PE, Dyer F. Reinforcement of anchorage during orthodontic brace treatment with implants or other surgical methods. Cochrane Database Syst Rev. 2007;3:CD005098.

11. Papadopoulos MA, Papageorgiou SN, Zogakis IP. Clinical effectiveness of orthodontic miniscrew implants: a meta-analysis. J Dent Res. 2011;90(8):969–76.

12. Papageorgiou SN, Zogakis IP, Papadopoulos MA. Failure rates and associated risk factors of orthodontic miniscrew implants: a meta-analysis. Am J Orthod Dentofacial Orthop. 2012;142(5):577–595.e7.

13. Jambi S, Walsh T, Sandler J, Benson PE, Skeggs RM, O'Brien KD. Reinforcement of anchorage during orthodontic brace treatment with implants or other surgical methods. Cochrane Database Syst Rev. 2014;19:8.

14. Nanda R, Upadhyay M. Sheldon Friel Lecture 2011 Skeletal and dental considerations in orthodontic treatment mechanics: a contemporary view. Eur J Orthod. 2013;35:634–43.

15. World Implant Orthodontic Conference (WIOC). [online] Available from: http://www.wioc2015.com/ (accessed September 25th 2015).

16. Kanomi R. Mini-implant for orthodontic anchorage. J Clin Orthod. 1997;31(11):763–7. 17. Prabhu J, Cousley RR. Current products and practice: bone anchorage devices in orthodontics. J Orthod.

2006;33(4):288–307. 18. Banks P, Elton V, Jones Y, Rice P, Denwent S, Odondi L. The use of fixed appliances in the UK: a survey of specialist

orthodontics. J Orthod. 2010;37(1):43–55. 19. Markic G, Katsaros C, Pandis N, Eliades T. Temporary anchorage device usage: a survey among Swiss

orthodontists. Prog Orthod. 2014;15(1):29. 20. Cabana MD, Rand CS, Powe NR, Wu AW, Wilson MH, Abboud PA, et al. Why don't physicians follow clinical

practice guidelines? A framework for improvement. JAMA. 1999;282(15):1458–65. 21. Gravel K, Légaré F, Graham ID. Barriers and facilitators to implementing shared decision-making in clinical

practice: a systematic review of health professionals' perceptions. Implement Sci. 2006;1:16. 22. Wallace J, Nwosu B, Clarke M. Barriers to the uptake of evidence from systematic reviews and meta-analyses: a

systematic review of decision makers' perceptions. BMJ Open. 2012;1:2(5). 23. Lee TC, McGrath CP, Wong RW, Rabie AB. Patients' perceptions regarding microimplant as anchorage in

orthodontics. Angle Orthod. 2008;78(2):228–33. 24. Sandhu JS, Sandhu SV, Bector K, Sandhu SS. Patients’ perception and postoperative discomfort with mini-

implants. J Ind Orthod Soc. 2013;47(4):199–201. 25. Damschroder L, Aron D, Keith R, Kirsh S, Alexander J, Lowery J. Fostering implementation of health services

research findings into practice: a consolidated framework for advancing implementation science. Implem Sci. 2009;4:50.

26. Grimshaw JM, Eccles MP, Lavis JN, Hill SJ, Squires JE. Knowledge translation of research findings. Implement Sci. 2012;7:50.

27. Grol R, Grimshaw J. From best evidence to best practice: effective implementation of change in patients' care. Lancet. 2003;362(9391):1225–30.

129

28. Legaré F. Chapter 3.4: assessing barriers and facilitators to knowledge use. In: Straus SE, Tetroe J, Graham ID, editors. Knowledge translation in health care: moving from evidence to practice. Chichester (UK): Wiley-Blackwell; 2009.

29. Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 [updated March 2011]. The Cochrane Collaboration, 2011. [online] Available from: www.cochrane-handbook.org. (accessed September 25th 2015).

30. The Joanna Briggs Institute Reviewers’Manual 2014 Edition. [online] Available from: http://joannabriggs.org/sumari.html (accessed September 25th 2015).

31. The Joanna Briggs Institute Reviewers’Manual 2014. The Systematic review of Prevalence and Incidence Data. [online] Available from: http://joannabriggs.org/sumari.html (accessed September 25th 2015).

32. Noyes J, Booth A, Hannes K, Harden A, Harris J, Lewin S, Lockwood C. Supplementary Guidance for Inclusion of Qualitative Research in Cochrane Systematic Reviews of Interventions. Version 1 (updated August 2011). Cochrane Collaboration Qualitative Methods Group, 2011. [online] Available from: http://cqrmg.cochrane.org/supplemental-handbook-guidance (accessed September 25th 2015).

33. Enhancing the Quality and Transparency Of health Research (Equator) network. [online] Available from: http://www.equator-network.org/reportingguidelines/ (accessed September 25th 2015).

34. Colvin CJ, Konopka S, Chalker JC, Jonas E, Albertini J, Amzel A, et al. A systematic review of health system barriers and enablers for antiretroviral therapy (ART) for HIV-infected pregnant and postpartum women. PLoS One. 2014;9(10):e108150.

35. Gibson O, Lisy K, Davy C, Aromataris E, Kite E, Lockwood C, et al. Enablers and barriers to the implementation of primary health care interventions for Indigenous people with chronic diseases: a systematic review. Implement Sci. 2015;10:71.

36. Kyei-Nimakoh M, Carolan-Olah M, McCann TV. Barriers to obstetric care at health facilities in sub-Saharan Africa—a systematic review protocol. Syst Rev. 2015;4(1):54.

37. Oliver K, Innvar S, Lorenc T, Woodman J, Thomas J. A systematic review of barriers to and facilitators of the use of evidence by policymakers. BMC Health Serv Res. 2014;14:2.

38. Meursinge Reynders RA, Ronchi L, Ladu L, van Etten-Jamaludin F, Bipat S. Insertion torque and success of orthodontic mini-implants: a systematic review. Am J Orthod Dentofacial Orthop. 2012;142(5):596–614.e5.

39. Meursinge Reynders R, Ronchi L, Ladu L, Van Etten-Jamaludin F, Bipat S. Insertion torque and orthodontic mini-implants: a systematic review of the artificial bone literature. Proc Inst Mech Eng H. 2013;227(11):1181–202.

40. Meursinge Reynders R, Ladu L, Ronchi L, Di Girolamo N, de Lange J, Roberts N, et al. Insertion torque recordings for the diagnosis of contact between orthodontic mini-implants and dental roots: protocol for a systematic review. Syst Rev. 2015;4(1)):39.

41. Reynders R, Ronchi L, Bipat S. Mini-implants in orthodontics: a systematic review of the literature. Am J Orthod Dentofacial Orthop. 2009;135:564–e1-19.

42. Moher D, Shamseer L, Clarke M, Ghersi D, Liberati A, Petticrew M, et al. Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols (PRISMA-P) 2015 statement. Syst Rev. 2015;4(1):1.

43. Shamseer L, Moher D, Clarke M, Ghersi D, Deceased LA, Petticrew M, et al. Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015: elaboration and explanation. BMJ. 2015;349:g7647.

44. PROSPERO: Centre for Reviews and Dissemination. University of York, UK. [online] Available from: [http://www.crd.york.ac.uk/PROSPERO/] (accessed September 25th 2015).

45. SUMSearch2. [online] Available from: www.sumsearch.org (accessed September 25th 2015). 46. Turning Research Into Practice (TRIP). [online] Available from: www.tripdatabase.com (accessed September 25th

2015). 47. The Cochrane Qualitative & Implementation Methods Group (CQIMG). [online] Available from:

http://cqim.cochrane.org/ (accessed September 25th 2015). 48. Harris J. Chapter 2: Using qualitative research to develop robust effectiveness questions and protocols for

Cochrane systematic reviews. In: Noyes J, Booth A, Hannes K, Harden A, Harris J, Lewin S, Lockwood C (editors), Supplementary Guidance for Inclusion of Qualitative Research in Cochrane Systematic Reviews of Interventions. Version 1 (updated August 2011). Cochrane Collaboration Qualitative Methods Group, 2011. [online] Available from: http://cqrmg.cochrane.org/supplemental-handbookguidance (accessed September 25th 2015).

49. Lefebvre C, Manheimer E, Glanville J. Chapter 6: Searching for studies. In: Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011. [online] Available from: www.cochrane-handbook.org. (accessed September 25th 2015).

50. Sterne JAC, Egger M, Moher D: Chapter 10: Addressing reporting biases. In: Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011. [online] Available from: www.cochrane-handbook.org. (accessed September 25th 2015).

51. Tong A, Flemming K, McInnes E, Oliver S, Craig J. Enhancing transparency in reporting the synthesis of qualitative research: ENTREQ. BMC Med Res Methodol. 2012;12:181.

52. Sociological Abstracts. [online] Available from: www.proquest.com/productsservices/socioabs-set-c.html (accessed September 25th 2015).


http://cqrmg.cochrane.org/supplemental-handbook-guidance

http://www.equator-network.org/reportingguidelines/

http://www.sumsearch.org/

http://cqrmg.cochrane.org/supplemental-handbookguidance


130

53. De Vet HCW, Eisinga A, Riphagen II, Aertgeerts B, Pewsner D. Chapter 7: Searching for Studies. In: Cochrane Handbook for Systematic Reviews of Diagnostic Test Accuracy Version 0.4 [updated September 2008]. The Cochrane Collaboration, 2008. [online] Available from: www.srdta.cochrane.org/handbook-dta-reviews (accessed September 25th 2015).

54. Web of Science Core Collection: Science Citation Index Expanded (SCIEXPANDED), Social Sciences Citation Index (SSCI), Arts & Humanities Citation Index (A&HCI). [online] Available from: http://thomsonreuters.com/en/products-services/scholarly-scientific-research/scholarly-search-and-discovery/web-of-science-core-collection.html (accessed September 25th 2015).

55. Hopewell S, McDonald S, Clarke M, Egger M. Grey literature in metaanalyses of randomized trials of health care interventions. Cochrane Database Syst Rev. 2007;18(2):MR000010.

56. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gøtzsche PC, Ioannidis JP, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. PLoS Med. 2009;6(7):e1000100. Epub 2009 Jul 21.

57. Bramer WM, Giustini D, Kramer BMR, Anderson PF. The comparative recall of Google Scholar versus PubMed in identical searches for biomedical systematic reviews: a review of searches used in systematic reviews. Syst Rev. 2013;2:115.

58. Higgins JPT, Deeks JJ. Chapter 7: Selecting studies and collecting data. In: Higgins JPT, Green S (editors), Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011. [online] Available from: www.cochrane-handbook.org. (accessed September 25th 2015).

59. Atherton H. Is there value in making author contact when determining risk of bias in a Cochrane Systematic Review. [online] Available from: http://community.cochrane.org/news/blog/there-value-making-author-contactwhen-determining-risk-biascochrane-systematic-review (accessed Setember 25th 2015).

60. Glossary of Terms in The Cochrane Collaboration. Version 4.2.5. Updated May 2005. [online] Available from: http://community.cochrane.org/sites/default/files/uploads/glossary.pdf (accessed September 25th 2015).

61. Moher D, Liberati A, Tetzlaff J, Altman DG, PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med. 2009;6(7):e1000097.

62. Munn Z, Moola S, Riitano D, Lisy K. The development of a critical appraisal tool for use in systematic reviews addressing questions of prevalence. Int J Health Policy Manag. 2014;3(3):123–8.

63. Harder T. Some notes on critical appraisal of prevalence studies: comment on: "The development of a critical appraisal tool for use in systematic reviews addressing questions of prevalence". Int J Health Policy Manag. 2014;3(5):289–90.

64. Borenstein M, Hedges LV, Higgins JPT, Rothstein HR. Chapter 40: when does it make sense to perform a meta-analysis? In: Borenstein M, Hedges LV, Higgins JPT, Rothstein HR, editors. Introduction to meta-analysis. Chichester (UK): Wiley; 2009.

65. Deeks JJ, Higgins JPT, Altman DG. Chapter 9: Analysing data and undertaking meta-analyses. In: Higgins JPT, Green S (editors), Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011. [online] Available from: www.cochrane-handbook.org. (accessed September 25th 2015).

66. Higgins JPT, Altman DG, Sterne JAC. Chapter 8: Assessing risk of bias in included studies. In: Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011. [online] Available from: www.cochranehandbook.org. (accessed September 25th 2015).

67. Bax L, Yu LM, Ikeda N, Moons KG. A systematic comparison of software dedicated to meta-analysis of causal studies. BMC Med Res Methodol. 2007;10:40.

68. Borenstein M, Hedges LV, Higgins JPT, Rothstein HR. Comprehensive Meta-Analysis (CMA) software. Version 3.0 Englewood, NJ, Biostat; 2014. [online] Available from: http://www.meta-analysis.com (accessed September 25th 2015).

69. Higgins JPT, Deeks JJ, Altman DG (editors). Chapter 16: Special topics in statistics. In: Higgins JPT, Green S (editors), Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011. [online] Available from: www.cochranehandbook.org. (accessed September 25th 2015).

70. West SL, Gartlehner G, Mansfield AJ, Poole C, Tant E, et al. Comparative Effectiveness Review Methods: Clinical Heterogeneity. Rockville, USA: Agency for Healthcare Research and Quality; 2010. Publication No. 10-EHC070-EF.

71. Agency for Healthcare Research and Quality (AHRQ). Methods Guide for Effectiveness and Comparative Effectiveness Reviews. [online] Available from: http://effectivehealthcare.ahrq.gov/ehc/products/60/318/CERMethods-Guide-140109.pdf (accessed September 25th 2015).

72. Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ. 1997;315(7109):629–34.

73. Egger M, Jüni P, Bartlett C, Holenstein F, Sterne J. How important are comprehensive literature searches and the assessment of trial quality in systematic reviews? Empirical study. Health Technol Assess. 2003;7:1.

74. Ioannidis JPA. Why most published research findings are false. PLoS Med. 2005;2(8):e124.


http://community.cochrane.org/news/blog/there-value-making-author-contactwhen-determining-risk-biascochrane-systematic-review

http://community.cochrane.org/news/blog/there-value-making-author-contactwhen-determining-risk-biascochrane-systematic-review

131

75. Borenstein M, Hedges LV, Higgins JPT, Rothstein HR. Chapter 16: identifying and quantifying heterogeneity. In: Borenstein M, Hedges LV, Higgins JPT, Rothstein HR, editors. Introduction to meta-analysis. Wiley: Chichester (UK); 2009.

76. Cochran WG. The combination of estimates from different experiments. Biometrics. 1954;10:101–29. 77. Higgins JPT, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ. 2003;327:557–

60. 78. DerSimonian R, Laird N. Meta-analysis in clinical trials. Controlled Clinical Trials. 1986;7:177–88. 79. Higgins JP. Commentary: heterogeneity in meta-analysis should be expected and appropriately quantified. Int J

Epidemiol. 2008;37(5):1158–60. 80. Kendall M. A new measure of rank correlation. Biometrika. 1938;30(1–2):81–9. 81. Oxman AD, Guyatt GH. A consumer's guide to subgroup analyses. Ann Intern Med. 1992;116(1):78–84. 82. Centre for Reviews and Dissemination (CRD). Systematic Reviews. CRD’s guidance for undertaking reviews in

health care. [online] Available from: https://www.york.ac.uk/crd/ (accessed September 25th 2015). 83. Borenstein M, Hedges LV, Higgins JPT, Rothstein HR. Chapter 28: vote counting—a new name for an old problem.

In: Borenstein M, Hedges LV, Higgins JPT, Rothstein HR, editors. Introduction to meta-analysis. Wiley: Chichester (UK); 2009.

84. Hedges L, Olkin I. Vote-counting methods in research synthesis. Psychological Bulletin. 1980;88:359–69. 85. Clinical Trial Register at the International Clinical Trials Registry Platform of the World Health Organization.

[online] Available from: http://apps.who.int/trialsearch/ (accessed September 25th 2015). 86. Borenstein M, Hedges LV, Higgins JPT, Rothstein HR. Chapter 30: Publication bias. In: Borenstein M, Hedges LV,

Higgins JPT, Rothstein HR, editors. Introduction to meta-analysis. Wiley: Chichester (UK); 2009. 87. Duval S, Tweedie R. Trim and fill: a simple funnel-plot-based method of testing and adjusting for publication bias

in meta-analysis. Biometrics. 2000;56(2):455–63. 88. The Grading of Recommendations Assessment, Development and Evaluation (GRADE) Working Group. [online]

Available from: www.gradeworkinggroup.org (accessed September 25th 2015). 89. Guyatt G, Oxman AD, Akl EA, Kunz R, Vist G, Brozek J, et al. GRADE guidelines: 1. Introduction-GRADE evidence

profiles and summary of findings tables. J Clin Epidemiol. 2011;64(4):383–94. 90. Higgins JPT, Green S: Chapter 4: Guide to the contents of a Cochrane protocol and review. In Cochrane Handbook

for Systematic Reviews of Interventions Version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011. [online] Available from: www.cochrane-handbook.org. (accessed September 25th 2015).

91. Chalmers I, Glasziou P. Avoidable waste in the production and reporting of research evidence. Lancet. 2009;374(9683):86–9.

92. Methodology Committee of the Patient-Centered Outcomes Research Institute (PCORI). Methodological standards and patient-centeredness in comparative effectiveness research: the PCORI perspective. JAMA. 2012;307(15):1636–40.

93. Selby JV, Beal AC, Frank L. The Patient-Centered Outcomes Research Institute (PCORI) national priorities for research and initial research agenda. JAMA. 2012;307(15):1583–4.


132

Additional file 1. Protocol for contacting authors Our protocol for contacting authors is based on protocols that we used in previous systematic reviews on OMIs38-41. The objectives of this protocol are to obtain additional information on: 1) the eligibility of specific research studies; and 2) unclear or missing data in primary research studies. Our protocol is presented as a step by step process. All emails sent to the contacted authors will be accompanied by the research protocol of our systematic review and all contacting attempts are conducted at 2 week intervals. Step 1. In our initial email we will ask reference authors whether they are willing to provide pertinent additional information on their research studies. Authors are only asked to answer this question with a ‘Yes’ or a ‘No’. We decided to use this simple ‘Willingness to reply’ format, because: 1) it can be used to confirm the validity of the email address of a reference author; 2) researchers might initially be hesitant to reply when confronted with large numbers of questions on their research study. An example of our initial email is presented in exemplary email 1. Step 2. When no response is received after 2 weeks, a reminder email will be send to the reference author. Step 3. When after another 2 weeks the reference author has still not responded, the initial email will be send to at least one of the co-authors. Step 4. In the case of no response, a reminder of this email will be sent after 2 weeks. Step 5. When after another 2 weeks still no response has been obtained, steps 1-4 will be repeated from another email address. This method was chosen, because our email address could be classified as a ‘spam’ email address by the internet provider of the contacted authors and therefore become invisible to these authors38. Step 6. Two weeks after the last email, we will send an email (Exemplary email 2) with pertinent questions to all authors that responded our ‘Willingness to reply’ question with a ‘Yes’. Questions will be presented as a combination of open-and closed-ended questions. For each specific research design we will consult the pertinent checklists of the Equator Network to fine-tune these questions33. We aim at not exceeding a total of 10 questions. Step 7. We will summarize the proceedings of our methods for contacting of authors in separate exemplary tables (Tables 1 and 2) of this additional file. In the final systematic review we will present a table that summarizes these proceedings and also lists the consequences of this information on e.g. eligibility, primary and secondary outcomes, and the assessment of risk of bias. We will report this information separately from the data extracted from the original research studies and do not incorporate these data in our analysis, because the validity of data obtained through contacted authors is currently unknown59.

133

Exemplary email 1. Assessment of author’s willingness to reply to questions of systematic reviewers40 Dear Professor …………., We are currently conducting a systematic review* on orthodontic mini implants and would like to obtain some additional information on your article: “ …………………………………………………………….”

In the near future we would like to ask you some simple questions on this publication. Answering these questions will probably not require more than 2 minutes of your time.

In this email we are only interested to find out whether you are willing to provide us with this additional information or not. We therefore ask you the following question:

Are you in the near future willing to respond to some simple questions pertinent to the indicated article ?

- If you are willing to respond to our questions please reply to this email, a "YES" is enough. - If you are not willing to respond to our questions please respond this email with a “NO”. In case, you do not reply, we will send you a reminder email 14 days after the current email.

-In case you respond with a “NO” or do not reply to this email and to the reminder, we will not further contact you, but will contact at least one of the other co-authors of your research study.

Thank you very much for your cooperation.

Sincerely, Reint Meursinge Reynders and Nicola Di Girolamo ----------------------------------------------------- *This systematic review originated as an assignment for the Evidence Based Health Care program at the University of Oxford, UK. The aforementioned individuals are subjectively responsible for the information requested.

134

Exemplary email 2. Research questions for authors to obtain additional data on their research studies*40 Dear Professor ………..,

In an earlier mail we asked you to answer some simple questions on your research article listed under here. Your answers could be a great contribution to improve on the quality of our systematic review**. The published protocol of our systematic review can be found in the attachment.

As we wrote in our previous email we would like to ask you some (9 in total) short questions regarding your article:

“ …………………………………………………………….”

All questions are combined closed-and open-ended questions. If you answer a question with a YES, we would like you to give a short description. Definitions of specific items are presented under the Question and Answer table.

Questions Answer Was random sampling1 of a particular subset of the population applied?

If you answer with a YES, could you please describe how this procedure was conducted

If you answer with a NO, it is not necessary to provide any additional information

Were particular groups of persons excluded?

If you answer with a YES, could you please describe which participants were excluded and why


Were consecutively treated2 participants enrolled?

If you answer with a YES, could you please describe how this procedure was conducted


Was a power calculation conducted to determine the adequate sample size?

If you answer with a YES, could you please describe how this procedure was conducted and the outcome of this calculation


Did all stakeholders refer to the same interventional procedure?

If you answer with a YES, could you please describe the interventional procedure


Was the setting3 the same for all stakeholders ?

If you answer with a YES, could you please describe the setting

135

If you answer with a NO, it is not necessary to provide any additional information Were all stakeholders interviewed by the same interviewer?

If you answer with a YES, could you please describe who interviewed all stakeholders and his/her experience with interviewing


Was an interview guide used for conducting the interview?

If you answer with a YES, could you please describe whether questions were open-ended, closed-ended or a mix of these questions


Did you record the response rate of the surveyed population?

If you answer with a YES, could you please provide the response rate


Definitions of the various items:

1Random sample The “Glossary of terms in the Cochrane Collaboration”60 defines a random sample as ‘A group of people selected for a study that is representative of the population of interest. This means that everyone in the population has an equal chance of being approached to participate in the survey, and the process is meant to ensure that a sample is as representative of the population as possible’. 2Consecutively treated participants Refers to the inclusion of all participants treated with the interventional procedure over a period of time. 3Setting The setting refers to the location where the research study was conducted, e.g., private practice, university setting, hospital etc. -------------------------------------------------------------------------------------------------------------------------------- We thank you again for your help and time. We will send you a copy of our systematic review as soon as it will be accepted for publication.

In the case that you have changed your mind and are not willing to reply to these questions, could you please explain the rationale for this decision.

Thank you again very much for your cooperation.

Sincerely, Reint Meursinge Reynders and Nicola Di Girolamo **This systematic review originated as an assignment for the Evidence Based Health Care program at the University of Oxford, UK. The aforementioned individuals are subjectively responsible for the information requested.

136

Table 1. Exemplary outcomes of (2 fictive) authors’s willingness to reply to questions of systematic reviewers 1Article and year

2Contacted author(s)

3Number of contacting attempts and 4number of days to get a reply

5Reminder mails?

6Co-authors contacted?

7Willingness to reply?

Author A (2014)

Author A 1 attempt Response same day

No No Yes

Author B (2015)

Author B and co-author C

4 attempts Response after 80 days

Yes Yes No

1Article and year: Presentation of the first author and the year of publication 2Contacted author(s): Presentation of each author that was contacted, i.e., the reference author and co-authors 3Number of contacting attempts: The number of contacting attempts refers to:

• The total number of attempts by email to get a response from a contacted author with either a ‘Yes’ or a ‘No’ on his/her willingness to reply to questions on his/her research studies. This number includes the total number of attempts of contacting both the reference author and potential co-authors.

• The initial email or the subsequent reminder email are each counted as 1 attempt. Emails that are reported as ‘failed’ by the internet provider, because they did not arrive at the desired email address are not counted as attempts.

• Our emails could also be identified as ‘spam mail’ by the receiving internet provider and we will therefore also send emails from our secondary email address (See step 5). When authors reply to this secondary email address, we will not count the earlier email attempts from our primary email address.

• As soon as authors have replied with either a ‘Yes’ or a ‘No’, we will not consider any successive email correspondence as part of the attempt count.

• Ideally only 1 contacting attempt is made by the systematic reviewers. 4Number of days to get a reply: The number of days to get a reply refers to:

• The total number of days necessary to get a response from a contacted author with either a ‘Yes’ or a ‘No’ on his/her willingness to reply to questions on his/her research studies. This number of days includes the total number of days of contacting both the reference author and potential co-authors to get such a reply.

• The day of sending the initial email to the reference author is the starting point for recording the ‘Number of days to get a reply’. The day of receiving a ‘Yes’ or a ‘No’ to our question from either the reference author or a co-author is the final recording of this outcome.

• We do not count days when emails are reported as ‘failed’ by the internet provider, because these letters did not arrive at the desired email address. When authors do not respond to our primary email address, but only to our secondary email address, we will only record the days necessary to get a reply to our email from the secondary address.

• As soon as authors have replied with either a ‘Yes’ or a ‘No’, we will not count the days of any successive email correspondence.

• Ideally authors respond the same day to the initial ‘Willingness to reply’ email. 5Reminder mails: Score with a ‘Yes’ or a ‘No’ 6Co-authors contacted: Score with a ‘Yes’ or a ‘No’ 7Willingness to reply: Score with a ‘Yes’ or a ‘No’

137

Table 2. Exemplary outcomes of answering by (2 fictive) contacted authors to research questions of systematic reviewers 1Article and year

2Contacted author(s)

3Number of contacting attempts and 4number of days to get a reply to our research questions

5Reminder mails?

6Prevalence of research questions answered

Author A (2014)

Author A 1 attempt

No 9 of 9 questions

Author B (2015)

Author B and co-author C

5 attempts Response after 80 days

Yes 2 of 6 questions

1Article and year: Presentation of the first author and the year of publication 2Contacted author(s): Presentation of each author that was contacted, i.e., the reference author and co-authors 3Number of contacting attempts: The number of contacting attempts refers to:

• The total number of attempts to get a reply to our email with research questions. These emails will only be sent to authors that answered our ‘Willingness to reply email’ with a ‘Yes’.

• Eligible replies vary from ‘unwillingness to answer’ our research questions to answering some or all of them. Subsequent emails to get explanations or to obtain data on unanswered questions are not counted as additional attempts.

• Ideally only 1 contacting attempt is made to get a reply to our email with research questions. 4Number of days to get a reply: The number of days to get a reply refers to:

• The total number of days necessary to get an eligible reply from a contacted author. • The day of sending the email with research questions to a pertinent author is the starting point for

recording the ‘Number of days to get a reply’. The day of receiving an eligible outcome is the final recording of this outcome

• Ideally authors respond to all questions on the same day that they receive our email with research questions.

5Reminder mails: Score with a ‘Yes’ or a ‘No’ 6Prevalence of research questions answered: Score the prevalence, i.e. the number of research questions answered out of total number of questions asked.

138

Additional file 2. Data collection forms The pilot-tested data collection forms with a description of each extracted item are presented in the tables under here. Pertinent items to facilitate the assessment of risk of bias are also included. Data collection form 1. Source and eligibility

Entry Description and character of the information Name of reviewer Report the name of the reviewer. Authors, title, journal of article List authors, title, and journal. Source of article Describe how the article was retrieved and from which database, e.g. Medline, grey

literature, hand searching of review articles etc. Language of the article Describe in which language the article was published. Registration of study Describe if study was registered, under what number, and in which register. Duplicate publication Describe whether the study as a whole or parts of the study are suspect of having

been published more than once. If suspicion of duplicate publication has been identified, the reviewer should also present the reference of pertinent duplicate studies and which parts of the study overlap.

Eligible/not eligible Confirm eligibility. If not eligible, explain the reason for exclusion.

Purpose Copy the objectives of the paper according to the authors. Data collection form 2. Miscellaneous *

Entry Description and character of the information Funding source Describe the funding source and indicate if it is an implant company. Key conclusions of study authors Describe the main conclusions by the authors of the research study. Important comments by the study authors










Contacting authors Indicate whether authors should be contacted and indicate what additional information is necessary.

Discuss with statistician Describe points that need to be discussed with a statistician. Points of interest Describe points of interest, e.g. an unexpected barrier or facilitator to the

implementation of OMIs Future studies Describe ideas for future studies suggested by the authors or reviewers. Other comments by review authors

Any additional comment that has not been covered.


139

Data collection form 3. ‘Study design and selection procedures’ *


Entry Description and character of the information Date of the study Describe when the study was started and completed. Study design Was the study design reported, e.g., survey, questionnaire, interview, focus group

etc.? If answered with ‘Yes’, describe the study design. If answered with ‘No’, describe the study design.

Selection procedures general Were the selection procedures of the stakeholders, e.g., random sampling, consecutively treated, case control design etc. reported? If answered with ‘Yes’, describe the selection procedures.

Consecutively treated Did the authors report on consecutive treatment of stakeholders? If answered with ‘Yes’, describe this entry.

Sequence generation Did the authors report on sequence generation, e.g., random or non random? If answered with ‘Yes’, describe the sequence generation. 1) Random sequence generation? Describe the type of randomization, e.g. computerized. 2) Non-random sequence generation?

Allocation concealment Did the authors report on concealment of allocation? If answered with ‘Yes’, describe blinding of pertinent operators and stakeholders, e.g., patients, clinicians, test operators, and personnel during the allocation procedures.

Case-control design Was a case-control design used? If answered with ‘Yes’, describe how stakeholders were enrolled according to a case-control design.

Power calculation Did the authors conduct a power calculation? If answered with ‘Yes’, describe how the power calculation was conducted.

Inappropriate exclusions Were single stakeholders or specific groups of stakeholders inappropriately excluded during any phase of the selection process? If answered with ‘Yes’, describe these procedures

Approved by ethical board Was the research study approved by an ethical board? If answered with ‘Yes’, describe this approval and the components of the ethical board.

Risk of bias Could any of the entries described in this table have caused biased study outcomes? If answered with ‘Yes’, present the rationale for this judgment. Present the magnitude of the bias (High, Low, Unclear) for each entry that was scored with a Yes. Also indicate the direction of this bias on the outcomes, i.e. increase or decrease of the magnitude of outcomes

140

Data collection form 4. ‘Stakeholders’ * Entry Description and character of the information Type of stakeholder Was the type of stakeholder, i.e., patients, clinicians, office staff, clinic owners,

policy makers etc. clearly reported? If answered with ‘Yes’, describe the type of stakeholder

Type of ‘patient’ stakeholders Describe this type of stakeholder, e.g., patients that finished treatment with palatal OMIs less than 3 months previously etc.

Type of ‘clinician’ stakeholders Describe this type of stakeholder, e.g. members of a dental association etc. Type of ‘clinic staff’ Describe this type of stakeholder, e.g. secretaries, assistants, laboratory staff etc. Type of ‘policy makers’ Describe this type of stakeholder, e.g., Governmental body, organization of

professionals etc. Other types of stakeholders Describe other pertinent stakeholders, e.g. family members etc. Number, sex, and age of stakeholders (including subgroups)

e.g. 20 males and 20 females mean age 23.6 ± 8.1 years Indicate particular subgroups if indicated, e.g., experienced versus non-experienced surgeons or patients that had undergone the intervention and those that have not, differences in interventions etc.

Ethnicity Was the ethnicity of the stakeholders reported? If answered with ‘Yes’, describe the ethnic group(s) and if applicable, present the number of stakeholders per ethnic subgroup.

Socio-economic status Was the socio-economic status of the stakeholder(s) described? If answered with ‘Yes’, describe the socio-economic status and if applicable, present the number of stakeholder(s) per socio-economic subgroup.

Experience with the interventional procedure

Was the experience of the stakeholders with the interventional procedure reported? If answered with ‘Yes’, describe the experience of stakeholders with this procedure, e.g., for patients (had they undergone the procedure or not), for clinicians (had they conducted the interventional procedure or not), etc. Include some additional variables that could be pertinent to this domain and if applicable, present the number of stakeholders per subgroup.

Co-existent conditions Were co-existent conditions described? If answered with ‘Yes’, describe co-existent conditions for example:

1) For patients this could be possible past interventions or co-interventions etc. Assess whether these interventions could have influenced the outcomes.

2) For clinicians and policy makers, prejudiced views on the interventional procedures etc.

If applicable, present the number of stakeholders per subgroup. Medical and dental health condition

Were medical variables, e.g. uncontrolled diabetes, osteoporosis, smoking, pharmacological treatment etc. reported? If answered with ‘Yes’, present the type of medical variables and the number of stakeholders per subgroup. Were dental variables, e.g. periodontal disease, loss of teeth etc. reported? If answered with ‘Yes’, present the type of dental variables and if applicable, present the number of stakeholders per subgroup.

Other stakeholder-related variables

Were other stakeholder-related variables described that could have influenced outcomes? If answered with ‘Yes’, describe these ‘other’ stakeholder-related variables and how they could have influenced outcomes.



141

Data collection form 5. ‘Setting, country and city’ * Setting Was the setting, e.g. private practice, university clinic etc., reported?

If answered with ‘Yes’, describe the setting of the research study and if applicable, present the number of stakeholders per subgroup.

Country/city Was the country(ies) and city(ies) or other geographical contexts (outside city) where the research study was conducted described? If answered with ‘Yes’, describe in which country(ies), city (or not) the research study was conducted and if applicable, present the number of stakeholders per country or city subgroup.

Other setting-related variables Were other setting-related variables described that could have influenced outcomes? If answered with ‘Yes’, describe these ‘other’ setting-related variables and how they could have influenced outcomes.



142

Data collection form 6. ‘Interventions’ * Entry Description and character of the information Definition of the interventional procedure

Did the authors define the interventional procedure(s) for which barriers and facilitators were identified? If answered with ‘Yes’ copy the definition of the interventional procedure according to the authors.

Specified or non specified interventions

Did the authors refer to a ‘specified’ or ‘non specified intervention’? Specified interventions: these interventions refer to a specific phase or type of the interventional procedure. Phases of the intervention refer to the: anesthetics, implant insertion, orthodontic treatment with OMIs, implant removal, or the healing phase. Types of interventions refer to the: implant type and dimensions, number of implants, use of plates, the surgical procedure, implant location, timing and forces of orthodontic loading etc.41. ‘Non specified’ interventions: these interventions refer to “any orthodontic treatment with OMIs”. Additional information on the specific phase or type of the interventional procedure is not provided If answered with ‘Yes’ describe the type of interventional procedure.

Plates ? Were plates connected to the OMIs? Describe if answered with ‘Yes’.

Type and number of OMIs Were the type and number of OMIs presented? Describe if answered with ‘Yes’, e.g., 2 Quattro*implants diameter 1.5 mm and length 9 mm (*PSM Medical Solutions; Tuttlingen, Germany)

Location ? Was the location of the insertion site described? Describe if answered with ‘Yes’, e.g., palate

Duration of the intervention Was the duration of the intervention described? Describe the duration of the intervention (in weeks) if answered with ‘Yes’.

Additional specification of the interventional procedure

Did the authors present any additional information on the interventional procedure? If Yes, specify. Examples of such specifications are presented under here:

1) With or without plates 2) Number of implants per patient 3) Implant type (code name and number) and the name of the company 4) Implant form, i.e., tapered, cylindrical or a combination of these forms 5) Implant dimensions, i.e., diameter and length 6) Drilling design, i.e., self-drilling or a pre-drilling design 7) Insertion technique, i.e., self-drilling or pre-drilling technique 8) Flap or flapless surgical intervention 9) Location of implant insertion, i.e., vestibular or palatal in addition to the

location between specific teeth 10) Location with respect to type of mucosa, i.e., in the keratinized or non-

keratinized mucosa 11) Exposure of implant, i.e., exposed or non-exposed under the mucosa 12) Orthodontic loading, i.e., immediately after implant insertion or after a

certain loading period 13) Implant removal with or without anesthetics etc. 14) Adverse effects, i.e., pain and discomfort, inflammation of oral tissues,

and biologic damage41 Interventional subgroups Were subgroups for interventional procedures presented?

Define subgroups if answered with ‘Yes’. Other intervention related variables

Were other interventional related variables described that could have influenced outcomes? If answered with ‘Yes’, describe these ‘other’ intervention related variables and how they could have influenced outcomes.



143

Data collection form 7. ‘Outcomes’ * Entry Description and character of the information Type of barrier or facilitator to the implementation of OMIs

List barriers or facilitators to the implementation of OMIs that were identified in the research study.

Definitions of barriers or facilitators to the implementation of OMIs

Did the author(s) present definition(s) of the barrier or facilitator to the implementation of OMIs? If answered with ‘Yes’, copy the definition of these items according to those presented by the author(s) of the selected study.

Clarity of definitions of barrier or facilitator to the implementation of OMIs

Was the definition of the barrier or facilitator to the implementation of OMIs sufficiently clear, in the sense that there was no doubt to which item the authors were referring? If not explain.

Time point(s) of scoring outcomes

Were time point(s) of identifying barriers and facilitators to the implementation of OMIs scored? Score time point(s) (in days) if answered with ‘Yes’, i.e., (1) pre-intervention recordings, i.e., recordings prior to the interventional procedure (2) immediate post-intervention recordings, i.e., recordings within 2 weeks after the completion of the interventional procedure; and (3) long-term post-intervention recordings, i.e., recordings after more than 2 weeks after the completion of the interventional procedure ?

Multiple time points for recording outcomes

Were outcomes recorded at more than 1 time point, e.g., pre-and post intervention outcomes? Describe the multiple time points for measuring outcomes and their time intervals (in days) if answered with ‘Yes’. Describe whether this interval was considered correct, too short or too long. Explain the rationale for this judgment.

Response rate of stakeholders that were contacted to participate in surveys

Was the response rate of the stakeholders reported? If answered with ‘Yes’, present the prevalence of the contacted stakeholders that actually answered the survey, i.e., The number of stakeholders that answered the survey questions/The total number of stakeholders that was contacted to answer these questions Present the response rate for example as: 200/1000

Inclusion in analysis (Lost to follow-up)

Were all surveyed/interviewed stakeholders also included in the data analysis. If answered with ‘No’, present total numbers of lost to follow-up. If answered with ‘No’, Present the number of stakeholders that were lost to follow-up/The total number of stakeholders that answered the survey or participated in the interviews. Present this statistic as the lost to follow-up rate for example as 200/1000 Present explanations for not being included in the analysis.

Prevalence of pertinent barriers and facilitators to the implementation of OMIs

Was the prevalence of identified barriers and facilitators among the surveyed or interviewed pertinent stakeholders reported? If answered with ‘Yes’, this statistic is calculated as: Prevalence of an identified barrier or facilitator = The number of stakeholders that scored a particular construct as a barrier or facilitator to the implementation of OMIs in clinical practice/ The total number of stakeholders that scored on the role of this particular construct as a barrier or facilitator to the implementation of OMIs in clinical practice This prevalence will be presented for example as: 30/50

Quality of the outcome assessors Was the quality, experience, and competence of the outcome assessor(s) described? If answered with ‘Yes’, describe these characteristics

Number of outcome assessors Was the number of outcome assessors presented? If answered with ‘Yes’, present the number of outcome assessors. Ideally more than one operator assesses the outcomes in order to avoid inadequate outcome assessments or exclusions of outcomes.

Calibration of outcome assesors Were outcome assessors calibrated? If answered with ‘Yes’, describe how outcome assessors were calibrated.

Intra-or inter-operator differences

Were intra-or inter operator differences of outcome assessors presented? If answered with ‘Yes’, describe these differences. Assess whether these differences are small enough to be ignored.

Other procedural variables Could other procedural variables have influenced outcomes? If answered with ‘Yes’, describe these procedural variables and how they could have influenced outcomes.

144


Secondary outcome Prevalence of clinicians that do not use OMIs

Was the prevalence of clinicians that do not use OMIs reported? If answered with ‘Yes’, this statistic is calculated as: The prevalence of clinicians that do not use OMIs = The number of clinicians that do not use OMIs/ The total number of surveyed clinicians that reported on the use of OMIs in clinical practice Present the prevalence for example as: 200/1000

Other pertinent outcomes on the use of OMIs

Were other pertinent outcomes on the use of OMIs by clinicians reported? If answered with ‘Yes’ present such outcomes, e.g., the number of implants placed per clinician per year.

*Locate where information on each item can be found, e.g., Page 10 column 2. Collection form 8. ‘Flow and timing’ *

Entry Description and character of the information Flow diagram Design a diagram of the flow of the stakeholders and the timing of the various

research phases from the start of the selection procedures to the completion of the recording of outcomes. This flow diagram should be divided in 5 research phases: 1) the population from which stakeholders were selected; 2) selection procedures; 3) conducting the research and data collection; 4) data analysis; and 5) reporting data. During each of these phases the number of stakeholders will be presented as well as the duration and division (include time intervals) of individual research phases. When possible a rationale for withdrawals and missing data should be given.

Other flow and timing related variables

Could other flow and timing related variables have influenced outcomes? If answered with ‘Yes’, describe these ‘other’ flow and timing related variables and how they could have influenced outcomes.


*Locate where information on each item can be found, e.g., Page 12 column 3. Data collection form 9. ‘Funding’*

Entry Description and character of the information Funding sources Was the funding of the research study reported?

If answered with ‘Yes’, describe the funding sources of the research study. Conflict of interest Could the authors or any of the stakeholders have a conflict of interest?

If answered with ‘Yes’, describe the type of conflict of interest and the pertinent stakeholders.



145

Additional file 3. Characteristics and findings of included studies The procedures to extract and categorize primary outcomes consist of 3 steps: Step 1: Barriers and facilitators to the implementation of OMIs (implementation constructs) and other pertinent items extracted from the eligible studies will be listed in ‘Characteristics and findings of included studies’ tables (Tables 1-4). Step 2: The implementation constructs will be subdivided in either barriers or facilitators and will be categorized according to the type of stakeholder. If specified, we will also subdivide these constructs according to the type of setting, interventions, the time points for recording these outcomes, and the research design of the study from which these outcomes were extracted. Step 3: We will list our secondary outcomes in a separate table (Table 5). Five anticipated exemplary tables present the characteristics and findings of the included studies (Tables 1-5). Pertinent items are defined for each of these tables. Table 1. Research design and response rate*

Study and year Research design Selection procedures Power calculation? Response rate

Study and year: The name of the first author and the year of the publication Research design: Qualitative study: interview, focus group etc. or a quantitative study: survey, questionnaire etc. Selection procedures? Describe selection procedures, e.g., randomized selection of doctors from an orthodontic society, consecutively treated patients with OMIs etc. Power calculation? Was a power calculation conducted? Yes/Not reported/Unclear Response rate: This statistic is calculated as: The number of stakeholders that answered the survey questions/The total number of stakeholders that was contacted to answer these questions. Present the response rate for example as: 200/1000 *Score as ‘Not reported’: When data on this entry are not reported in the research study *Score as ‘Not applicable’: When the entry is not applicable for the research study Table 2. Stakeholders (interviewed or surveyed population) and settings *

Study and year Type, number, and age of the stakeholders

Ethnicity Experience of stakeholders

Settings/country/ city

Study and year: The name of the first author and the year of the publication Type, number, and age of stakeholders: Characteristics of the participants from which research data were obtained, e.g. Indian doctors, or Japanese patients, or subgroups of stakeholders etc. Include the number and age (years and months) of the stakeholders Ethnicity: Present the ethnicity of the stakeholders, e.g., Indian, Japanese etc. Experience of stakeholders: Stakeholders with or without experience with the interventional procedure, e.g., patients who had already undergone treatment with OMIs or clinicians who had experience with using OMIs as anchorage devices Setting/country/city: Present the setting, the country, city (or outside) of the research study, e.g., private practice or university clinic in Milan, Italy etc. *Score as ‘Not reported’: When data on this entry are not reported in the research study *Score as ‘Not applicable’: When the entry is not applicable for the research study

146

Table 3. Interventions* Study and year

Definition interventions?

‘Specified’ or ‘Non specified’ intervention

Plates? Type and number of OMIs

Location? Duration (weeks)

Study and year: The name of the first author and the year of the publication Definition interventions: Did the author(s) present a definition of the interventional procedure? Yes/Not reported/Unclear ‘Specified’ or ‘Non specified’ interventions: Describe the interventional procedures. ‘Specified’ interventions: these interventions refer to a specific phase or type of the interventional procedure. Phases of the intervention refer to the: anesthetics, implant insertion, orthodontic treatment with OMIs, implant removal, or the healing phase. Types of interventions refer to the: implant type and dimensions, number of implants, use of plates, the surgical procedure, implant location, timing and forces of orthodontic loading etc.41. ‘Non specified’ interventions: these interventions refer to “any orthodontic treatment with OMIs”. Additional information on the specific phase or type of the interventional procedure is not provided Plates: Were plates connected to the OMIs? Yes/Not reported/Unclear Type and number of OMIs: Describe the type and number of the OMIs, e.g., 2 Quattro**implants diameter 1.5 mm and length 9 mm (**PSM Medical Solutions; Tuttlingen, Germany) Location: Describe the implant insertion site, e.g., palate Duration: Duration of total treatment time with OMIs in weeks *Score as ‘Not reported’: When data on this entry are not reported in the research study *Score as ‘Not applicable’: When the entry is not applicable for the research study Table 4. Outcomes*

Study and year

Barriers and stakeholders

Facilitators and stakeholders

Definition of the barriers and facilitators

Prevalence of the barrier or facilitator

Time point of measuring outcomes

Study and year: The name of the first author and the year of the publication Barriers and stakeholders: List all identified barriers to the implementation of OMIs in clinical practice and list the pertinent stakeholders to which these barriers apply, e.g. patients, or clinicians etc. Facilitators and stakeholders: List all identified facilitators to the implementation of OMIs in clinical practice and list the pertinent stakeholders to which these facilitators apply, e.g. patients, or clinicians etc. Definition of the barriers and facilitators: Were barriers and facilitators defined by the author(s) of the eligible research study? Yes/Not reported/Unclear Prevalence of the barrier or facilitator: This statistic is calculated as: The number of stakeholders that scored a particular construct as a barrier or facilitator to the implementation of OMIs in clinical practice/The total number of stakeholders that scored on the role of this particular construct as a barrier or facilitator to the implementation of OMIs in clinical practice This prevalence will be presented for example as: 30/50. Time point of measuring outcomes: Outcomes are either scored prior to the intervention, immediately (within 2 weeks) or long-term (more than 2 weeks) after the interventional procedure. *Score as ‘Not reported’: When data on this entry are not reported in the research study *Score as ‘Not applicable’: When the entry is not applicable for the research study Table 5. Use of OMIs by clinicians*

Study and year

Research design

Response rate

Prevalence of clinicians that do not use OMIs

Additional information on the use by clinicians of OMIs

Study and year: The name of the first author and the year of the publication Research design: Qualitative study: interview, focus group etc. or a quantitative study: survey, questionnaire etc. Response rate: This statistic is calculated as: The number of stakeholders that answered the survey questions/The total number of stakeholders that was contacted to answer these questions. Present the response rate for example as: 200/1000 Prevalence of clinicians that do not use OMIs: This statistic is calculated as: The number of clinicians that do not use OMIs/The total number of surveyed clinicians that reported on the use of OMIs in clinical practice Additional information on the use of OMIs by clinicians: Information that could give further insights in the understanding of the knowledge-to-action gap, e.g., the number of implants placed per clinician per year. *Score as ‘Not reported’: When data on this entry are not reported in the research study *Score as ‘Not applicable’: When the entry is not applicable for the research study

147

Additional file 4. Critical appraisal tools Table 1. Joanna Briggs Qualitative Assessment and Review Instrument (JBI QARI)30* **

Appraisal questions Yes No Unclear Not applicable Comments 1)Is there congruity between the stated philosophical perspective and the research methodology?

2)Is there congruity between the research methodology and the research question or objectives?

3)Is there congruity between the research methodology and the methods used to collect data?

4)Is there congruity between the research methodology and the representation and analysis of data?

5)Is there congruity between the research methodology and the interpretation of results?

6)Is there a statement locating the researcher culturally or theoretically?

7)Is the influence of the researcher on the research, and vice-versa, addressed?

8)Are participants, and their voices, adequately represented?

9)Is the research ethical according to current criteria or, for recent studies, and is there evidence of ethical approval by an appropriate body?

10)Do the conclusions drawn in the research report flow from the analysis, or interpretation, of the data?

* Critical appraisal scores: Yes No ? Unclear NA Not Applicable ** Locate where information on each item can be found, e.g., Page 12 column 3. Table 2. Tabular presentation of the scores of the Joanna Briggs Institute Qualitative Assessment and Review Instrument (JBI QARI)30*

Study Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Author(s) ref

* Critical appraisal scores: Yes No ? Unclear NA Not Applicable

148

Table 3. The Joanna Briggs Institute critical appraisal tool of prevalence and incidence data31,62* ** Appraisal questions Yes No Unclear Not applicable Comments 1)Was the sample representative of the target population?

2)Were study participants recruited in an appropriate way?

3)Was the sample size adequate? 4)Were the study subjects and the setting described in detail?

5)Was the data analysis conducted with sufficient coverage of the identified sample?

6)Were objective, standard criteria used for the measurement of the condition?

7)Was the condition measured reliably? 8)Was there appropriate statistical analysis? 9)Are all important confounding factors/subgroups/differences identified and accounted for?

10)Were subpopulations identified using objective criteria?

* Critical appraisal scores: Yes No ? Unclear NA Not Applicable ** Locate where information on each item can be found, e.g., Page 12 column 3. Table 4. Tabular presentation of the scores of the Joanna Briggs Institute critical appraisal tool of prevalence and incidence data31,62*

Study Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Author(s) ref

* Critical appraisal scores: Yes No ? Unclear NA Not Applicable

149

Additional file 5. Heterogeneity

Sources of heterogeneity Potential sources of heterogeneity in the eligible research studies that could influence outcomes are divided in 3 main categories: methodological, clinical, and other sources of heterogeneity65,70. A series of these sources are defined a priori under here. We will consult these variables to explore heterogeneity in outcomes. Methodological heterogeneity Methodological heterogeneity refers to differences in outcomes within and between studies as a result of the design, conduct and analysis, and size of the study65,66,70. We will consider the following potential sources of methodological heterogeneity (Table 1). Table 1. Potential sources of methodological heterogeneity

Potential sources of methodological heterogeneity

Description of pertinent subgroups for each domain

Research design Surveys or questionnaires, interviews, and focus groups etc. Conduct and analysis of the study The risk of selection, performance, detection, attrition, and reporting

bias66. Study size ‘Study size’ will be explored, because smaller studies often lack power and

their methodological rigour tends to be suboptimal72,73,74. Clinical heterogeneity Clinical heterogeneity refers to differences in outcomes within and between studies as a result of specific characteristics of stakeholders, interventions, outcomes, and setting65,70. We will consider the following potential sources of clinical heterogeneity (Table 2). Table 2. Potential sources of clinical heterogeneity

Potential sources of clinical heterogeneity


Characteristics of stakeholders Ethnicity, sex, age, previous experience with OMIs etc. Type of interventions Specified interventions: this type of intervention refers to a specific phase or

type of the interventional procedure. Phases of the intervention refer to the: anesthetics, implant insertion, orthodontic treatment with OMIs, implant removal, or the healing phase. Types of interventions refer to the: implant type and dimensions, number of implants, use of plates, the surgical procedure, implant location, timing and forces of orthodontic loading41. ‘Non specified’ interventions: this type of intervention refers to “any orthodontic treatment with OMIs”. Additional information on the specific phase or type of the interventional procedure is not provided by the authors.

Outcomes Pre-intervention recordings, i.e., recordings of outcomes prior to the interventional procedure. Immediate post-intervention recordings, i.e., recordings of outcomes within 2 weeks after the completion of the interventional procedure. Long-term post-intervention recordings, i.e., recordings of outcomes after more than 2 weeks after the completion of the interventional procedure.

Setting Private practice and university setting

150

Other sources of heterogeneity Other sources of heterogeneity refers to differences in outcomes within and between studies as a result of scientific misconduct, funding of the research study, cost of the interventional procedure, and conflicts of interest70. We will consider the following potential ‘other’ sources of heterogeneity (table 3). Table 3. Potential ‘other’ sources of heterogeneity

Potential ‘other’ sources of heterogeneity


‘Other’ sources of heterogeneity Scientific misconduct, funding of the research study, cost of the interventional procedure, conflicts of interest etc.

Strategies for dealing with heterogeneity We will create forest plots to visualize statistical heterogeneity68. We will test the presence of statistical heterogeneity by obtaining a p-value from the comparison of Cochran’s Q with a Chi2 distribution with k−1 degrees of freedom (k = the number of eligible studies)75-77. We will take in consideration that the power of this test is low when the number of eligible studies is low and is excessive when the numbers of studies is high and particularly when study sizes are large75,77. We will calculate Kendall’s Tau2 for estimating the variance of true effect sizes between studies and Tau as the standard deviation of this estimate68,75,77-80. These estimates will be used to represent the magnitude of the heterogeneity75,79. I2 will also be calculated and reflects the proportion of the between study variance to the total variance across the observed effect estimates 75,77,79. The interpretation of this measure of inconsistency will be conducted according to the guidelines described in the Cochrane Handbook for Systematic Reviews of Interventions, i.e., I2: 0% to 40%: might not be important; I2: 30% to 60%: may represent moderate heterogeneity; I2: 50% to 90%: may represent substantial heterogeneity; and I2: 75% to 100%: considerable heterogeneity65. If clarification on issues of heterogeneity is indicated, we will contact authors according to our protocol (Additional file 1). Besides assessing the role of ‘a priori’ defined sources of heterogeneity we will also explore whether additional (‘post hoc’) sources of heterogeneity could be identified in the eligible studies. We will weigh the various strategies for dealing with statistical heterogeneity according to the Cochrane Handbook for Systematic Reviews of Interventions65.

1) Double-check whether the data extraction from the eligible studies and the data entry into the Comprehensive Meta-Analysis (CMA) program were done correctly68.

2) Assess whether pooling of outcomes is indicated based on our criteria for conducting a quantitative synthesis.

3) Investigate potential causes of heterogeneity through subgroup analyses and meta-regression. 4) Ignore heterogeneity by applying a fixed-effect meta-analysis. 5) Investigate heterogeneity that cannot be explained through a random-effects meta-analysis. 6) Change the measure of effect size. 7) Exclude studies, but consider the consequences of introducing bias.

151

Additional file 6. PRISMA-P 2015 Checklist PRISMA-P (Preferred Reporting Items for Systematic review and Meta-Analysis Protocols) 2015 checklist: recommended items to address in a systematic review protocol*

Section and topic Item No Checklist item

ADMINISTRATIVE INFORMATION Title:

Identification 1a Identify the report as a protocol of a systematic review (Title page 110) Update 1b If the protocol is for an update of a previous systematic review, identify as such

(Is not an update) Registration 2 If registered, provide the name of the registry (such as PROSPERO) and registration number

(Not registered. Explanation in Methods on Page 115) Authors:

Contact 3a Provide name, institutional affiliation, e-mail address of all protocol authors; provide physical mailing address of corresponding author (Author details and email addresses are available in the published version of this manuscript)

Contributions 3b Describe contributions of protocol authors and identify the guarantor of the review (Authors’s contributions Page 127)

Amendments 4 If the protocol represents an amendment of a previously completed or published protocol, identify as such and list changes; otherwise, state plan for documenting important protocol amendments (Not applicable)

Support: Sources 5a Indicate sources of financial or other support for the review

(All expenses for conducting this systematic review were paid evenly by each reviewer Page 127)

Sponsor 5b Provide name for the review funder and/or sponsor (Not applicable)

Role of sponsor or funder

5c Describe roles of funder(s), sponsor(s), and/or institution(s), if any, in developing the protocol (Not applicable)

INTRODUCTION Rationale 6 Describe the rationale for the review in the context of what is already known

(Background Pages 112 -115) Objectives 7 Provide an explicit statement of the question(s) the review will address with reference to

participants, interventions, comparators, and outcomes (PICO) (Objectives Page 115)

METHODS Eligibility criteria 8 Specify the study characteristics (such as PICO, study design, setting, time frame) and report

characteristics (such as years considered, language, publication status) to be used as criteria for eligibility for the review (Eligibility criteria Pages 115-116)

Information sources 9 Describe all intended information sources (such as electronic databases, contact with study authors, trial registers or other grey literature sources) with planned dates of coverage (Information sources Pages 116-117)

Search strategy 10 Present draft of search strategy to be used for at least one electronic database, including planned limits, such that it could be repeated (Search strategy Pages 117-118)

Study records: Data management 11a Describe the mechanism(s) that will be used to manage records and data throughout the review

(Data management Pages 118-120) Selection process 11b State the process that will be used for selecting studies (such as two independent reviewers)

through each phase of the review (that is, screening, eligibility and inclusion in meta-analysis) (Selection process Pages 118-120)

Data collection process

11c Describe planned method of extracting data from reports (such as piloting forms, done independently, in duplicate), any processes for obtaining and confirming data from investigators (Data collection process Page 119)

152

Data items 12 List and define all variables for which data will be sought (such as PICO items, funding sources), any pre-planned data assumptions and simplifications (Data items Pages 119-120)

Outcomes and prioritization 13 List and define all outcomes for which data will be sought, including prioritization of main and additional outcomes, with rationale (Outcomes and prioritization Pages 120-121)

Risk of bias in individual studies

14 Describe anticipated methods for assessing risk of bias of individual studies, including whether this will be done at the outcome or study level, or both; state how this information will be used in data synthesis (Risk of bias in individual studies Pages 121-122)

Data synthesis 15a Describe criteria under which study data will be quantitatively synthesized (Criteria for a quantitative synthesis Pages 122-123)

15b If data are appropriate for quantitative synthesis, describe planned summary measures, methods of handling data and methods of combining data from studies, including any planned exploration of consistency (such as I2, Kendall’s τ) (Page 123)

15c Describe any proposed additional analyses (such as sensitivity or subgroup analyses, meta-regression) (Subgroup analyses and meta-regression and sensitivity analysis Pages 123)

15d If quantitative synthesis is not appropriate, describe the type of summary planned (Qualitative synthesis Pages 124-125)

Meta-bias(es) 16 Specify any planned assessment of meta-bias(es) (such as publication bias across studies, selective reporting within studies) (Meta biases Pages 125-126)

Confidence in cumulative evidence

17 Describe how the strength of the body of evidence will be assessed (such as GRADE) (Confidence in cumulative evidence Pages 125-126)

* It is strongly recommended that this checklist be read in conjunction with the PRISMA-P Explanation and Elaboration (cite when available) for important clarification on the items. Amendments to a review protocol should be tracked and dated. The copyright for PRISMA-P (including checklist) is held by the PRISMA-P Group and is distributed under a Creative Commons Attribution Licence 4.0. From: Shamseer L, Moher D, Clarke M, Ghersi D, Liberati A, Petticrew M, Shekelle P, Stewart L, PRISMA-P Group. Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015: elaboration and explanation. BMJ. 2015 Jan 2;349(jan02 1):g7647.

UvA-DARE (Digital Academic Repository) Evidence-based ... › ws › files › 2811566 › 177075_PhD... · Evidence-based knowledge creation on orthodontic mini-implants: ‘Why

Documents