Systematic review Systematic review of antibiotic treatment for acute calculous cholecystitis A. H. van Dijk 1 , P. R. de Reuver 3 , T. N. Tasma 1 , S. van Dieren 1,2 , T. J. Hugh 4 and M. A. Boermeester 1 1 Department of Surgery and 2 Clinical Research Unit, Academic Medical Centre, Amsterdam, and 3 Department of Surgery, Radboud University Medical Centre, Nijmegen, The Netherlands, and 4 Upper Gastrointestinal Surgery Unit, Royal North Shore Hospital, University of Sydney, Sydney, New South Wales, Australia Correspondence to: Ms A. H. van Dijk, Academic Medical Centre, Meibergdreef 9, 1105 AZ Amsterdam, The Netherlands (e-mail: [email protected]) Background: Intravenous antibiotics are frequently used in the initial management of acute calculous cholecystitis (ACC), although supportive care alone preceding delayed elective cholecystectomy may be sufficient. This systematic review assessed the success rate of antibiotics in the treatment of ACC. Methods: A systematic search of MEDLINE, Embase and Cochrane Library databases was performed. Primary outcomes were the need for emergency intervention and recurrence of ACC after initial non-operative management of ACC. Risk of bias was assessed. Pooled event rates were calculated using a random-effects model. Results: Twelve randomized trials, four prospective and ten retrospective studies were included. Only one trial including 84 patients compared treatment with antibiotics to that with no antibiotics; there was no significant difference between the two groups in terms of length of hospital stay and morbidity. Some 5830 patients with ACC were included, of whom 2997 had early cholecystectomy, 2791 received initial antibiotic treatment, and 42 were treated conservatively. Risk of bias was high in most studies, and all but three studies had a low level of evidence. For randomized studies, pooled event rates were 15 (95 per cent c.i. 10 to 22) per cent for the need for emergency intervention and 10 (5 to 20) per cent for recurrence of ACC. The pooled event rate for both outcomes combined was 20 (13 to 30) per cent. Conclusion: Antibiotics are not indicated for the conservative management of ACC or in patients scheduled for cholecystectomy. Paper accepted 5 February 2016 Published online 30 March 2016 in Wiley Online Library (www.bjs.co.uk). DOI: 10.1002/bjs.10146 Introduction Acute calculous cholecystitis (ACC) is diagnosed in 3–10 per cent of patients with acute abdominal pain 1 and accounts for approximately one-third of all surgical emergency admissions 2 . There is controversy regard- ing the timing of cholecystectomy in ACC. Both early cholecystectomy (operation within 7 days of onset of complaints) and initial conservative therapy followed by delayed elective cholecystectomy are reasonable treatment options 3 . A recent meta-analysis 4 showed that morbidity and con- version to an open procedure is the same for early and delayed cholecystectomy. Early cholecystectomy was asso- ciated with a significantly shorter total hospital stay, and is the treatment of choice according to recent guidelines 5,6 . Early cholecystectomy is still not common practice in many countries. Non-operative management, including bowel rest, intravenous hydration, analgesia and intra- venous antibiotics, are used in the initial management of ACC followed by delayed cholecystectomy 7 – 9 . Although evidence for the use of antibiotics in the con- servative management of ACC is limited, guidelines rec- ommend their use 10,11 . The most frequently cited reason for commencing antibiotics is to prevent complications, either in the period before elective cholecystectomy or perioperatively 12,13 . The current view, however, is that the natural course of ACC is often uncomplicated, as ACC is thought to be a primarily inflammatory process. Consequently, supportive care without antibiotics may be sufficient before elective cholecystectomy. The aim of this systematic review was to assess the value of antibiotics in terms of their ability to decrease the need for emergency intervention, to prevent recurrence © 2016 BJS Society Ltd BJS 2016; 103: 797–811 Published by John Wiley & Sons Ltd
Systematic review of antibiotic treatment for acute calculous cholecystitis
Sep 03, 2022
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Systematic review of antibiotic treatment for acute calculous cholecystitisSystematic review of antibiotic treatment for acute calculous cholecystitis
A. H. van Dijk1, P. R. de Reuver3, T. N. Tasma1, S. van Dieren1,2, T. J. Hugh4 and M. A. Boermeester1
1Department of Surgery and 2Clinical Research Unit, Academic Medical Centre, Amsterdam, and 3Department of Surgery, Radboud University Medical Centre, Nijmegen, The Netherlands, and 4Upper Gastrointestinal Surgery Unit, Royal North Shore Hospital, University of Sydney, Sydney, New South Wales, Australia Correspondence to: Ms A. H. van Dijk, Academic Medical Centre, Meibergdreef 9, 1105 AZ Amsterdam, The Netherlands (e-mail: [email protected])
Background: Intravenous antibiotics are frequently used in the initial management of acute calculous cholecystitis (ACC), although supportive care alone preceding delayed elective cholecystectomy may be sufficient. This systematic review assessed the success rate of antibiotics in the treatment of ACC. Methods: A systematic search of MEDLINE, Embase and Cochrane Library databases was performed. Primary outcomes were the need for emergency intervention and recurrence of ACC after initial non-operative management of ACC. Risk of bias was assessed. Pooled event rates were calculated using a random-effects model. Results: Twelve randomized trials, four prospective and ten retrospective studies were included. Only one trial including 84 patients compared treatment with antibiotics to that with no antibiotics; there was no significant difference between the two groups in terms of length of hospital stay and morbidity. Some 5830 patients with ACC were included, of whom 2997 had early cholecystectomy, 2791 received initial antibiotic treatment, and 42 were treated conservatively. Risk of bias was high in most studies, and all but three studies had a low level of evidence. For randomized studies, pooled event rates were 15 (95 per cent c.i. 10 to 22) per cent for the need for emergency intervention and 10 (5 to 20) per cent for recurrence of ACC. The pooled event rate for both outcomes combined was 20 (13 to 30) per cent. Conclusion: Antibiotics are not indicated for the conservative management of ACC or in patients scheduled for cholecystectomy.
Paper accepted 5 February 2016 Published online 30 March 2016 in Wiley Online Library (www.bjs.co.uk). DOI: 10.1002/bjs.10146
Introduction
Acute calculous cholecystitis (ACC) is diagnosed in 3–10 per cent of patients with acute abdominal pain1
and accounts for approximately one-third of all surgical emergency admissions2. There is controversy regard- ing the timing of cholecystectomy in ACC. Both early cholecystectomy (operation within 7 days of onset of complaints) and initial conservative therapy followed by delayed elective cholecystectomy are reasonable treatment options3.
A recent meta-analysis4 showed that morbidity and con- version to an open procedure is the same for early and delayed cholecystectomy. Early cholecystectomy was asso- ciated with a significantly shorter total hospital stay, and is the treatment of choice according to recent guidelines5,6. Early cholecystectomy is still not common practice in
many countries. Non-operative management, including bowel rest, intravenous hydration, analgesia and intra- venous antibiotics, are used in the initial management of ACC followed by delayed cholecystectomy7–9.
Although evidence for the use of antibiotics in the con- servative management of ACC is limited, guidelines rec- ommend their use10,11. The most frequently cited reason for commencing antibiotics is to prevent complications, either in the period before elective cholecystectomy or perioperatively12,13. The current view, however, is that the natural course of ACC is often uncomplicated, as ACC is thought to be a primarily inflammatory process. Consequently, supportive care without antibiotics may be sufficient before elective cholecystectomy.
The aim of this systematic review was to assess the value of antibiotics in terms of their ability to decrease the need for emergency intervention, to prevent recurrence
© 2016 BJS Society Ltd BJS 2016; 103: 797–811 Published by John Wiley & Sons Ltd
798 A. H. van Dijk, P. R. de Reuver, T. N. Tasma, S. van Dieren, T. J. Hugh and M. A. Boermeester
References identified through
Cochrane Library n = 162
Duplicates n = 8390
Full-text articles assessed for eligibility
n = 89
Articles included
n = 26
Not available n = 4
Articles included in meta-analysis
Fig. 1 PRISMA flow diagram for the review
of ACC, to minimize complications, or to avoid elective cholecystectomy after initial conservative treatment.
Methods
The study protocol for this systematic review followed the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) checklist14.
Eligibility criteria
Articles were eligible for inclusion when: the study population consisted of adults with ultrasound-proven ACC with clinical signs; ACC was defined in the arti- cle; the article reported original data and the full text was available; and the article described one group of patients with ACC who were initially treated with antibiotics.
Only studies published after the introduction of laparoscopic cholecystectomy in 1985 were included, to represent current clinical practice. When the full
text of an article was not available, the original author was contacted to request access to the data. No limits regarding language or study design were defined. Arti- cles describing a specific patient population, such as malignancies or patients in intensive care, were excluded. Studies describing prophylactic antibiotics before laparo- scopic cholecystectomy were outside the scope of this review.
Search
The MEDLINE, Embase and Cochrane Library databases were searched for articles (October 2015). The follow- ing keywords and synonyms were used for MEDLINE and Embase: cholecystitis, cholecystectomy, antibacte- rial agents, antibiotics and conservative management; Appendix S1 (supporting information) outlines the complete search strategy. For every keyword (medical subject heading (MeSH) for MEDLINE and Emtree for EMBASE), free text words were also searched. The Cochrane Library was checked with the same key words. Cross-references were hand-searched.
© 2016 BJS Society Ltd www.bjs.co.uk BJS 2016; 103: 797–811 Published by John Wiley & Sons Ltd
Antibiotic treatment for acute calculous cholecystitis 799
Table 1 Characteristics of included studies, ranked by year of publication
Reference Country Study design Type of intervention Length of follow-up Primary goal
Level of evidence17
Kirkil et al.41 Turkey Retrospective ELC versus interval cholecystectomy and DLC
n.r. Comparison between three treatment strategies
4
Agrawal et al.18 India RCT ELC versus DLC in ACC 8 weeks Success of DLC 3
Hasbahçeci et al.40 Turkey Retrospective ELC versus conservative treatment or DLC in ACC
6 months LOS, gallstone-related complications and mortality
4
Wang et al.34 Taiwan Retrospective Antibiotic treatment in all patients with ACC
Median 308⋅5 days Recurrence rate of ACC 4
Oymaci et al.42 Turkey Retrospective ELC versus DLC in ACC n.r. Comparison of ELC versus DLC 4
Barcelo et al.35 Spain Retrospective ELC (within 48 h of admission) versus conservative treatment and DLC
≥ 6 months Therapeutic approach and failure of chosen therapy
4
Al-Qahtani36 Saudi Arabia Retrospective ELC (within 7 days of onset of symptoms) versus conservative treatment with antibiotics followed by DLC
n.r. Success of ELC and conservative treatment
4
Gul et al.19 India RCT ELC (within 72 h of admission) versus conservative treatment with antibiotics and DLC
1 year To evaluate safety and feasibility of LC for ACC and to compare ELC with DLC
3
Gutt et al.20 Germany RCT ELC (within 24 h of admission) versus DLC
75 days To compare 75-day morbidity between ELC and DLC
2
Degrate et al.43 Italy Retrospective ELC (within initial admission) versus DLC
n.r. To compare ELC and DLC 4
Al-Faouri et al.39 Jordan Retrospective ELC versus DLC 2 months to 1 year To compare ELC and DLC in ACC 4
Mazeh et al.21 Israel RCT Supportive care+ antibiotics versus supportive care only
Mean 17⋅1 (range 4–28) months
LOS at index admission and combined LOS for all admissions
2
1 month Security and efficacy of home treatment of ACC
3
n.r. Failure of conservative treatment 4
Barak et al.31 Israel Prospective All patients treated conservatively; when treatment failure, then PC
n.r. Failure of conservative management and predictive factors for failure
3
Casillas et al.38 USA Retrospective ELC (within 24 h of admission) versus antibiotics
n.r. Duration of symptoms, LOS and complications
4
Paran et al.32 Israel Prospective Conservative treatment with antibiotics. If failure of treatment, then emergency LC in patients with symptoms for≤72 h. If symptoms for>72 h, then PC
n.r. Failure of conservative therapy 3
Kolla et al.22 India RCT ELC (within 24 h of randomization) versus conservative treatment with antibiotics and DLC
n.r. Operating time, conversions, LOS, complications
3
Serralta et al.33 USA Prospective ELC (within 72 h of onset of symptoms) versus conservative treatment with antibiotics and DLC
n.r. LOS and postoperative morbidity 3
Johansson et al.26,27 Sweden RCT ELC (within 48 h after randomization, but no later than 7 days after onset of symptoms) versus conservative treatment with antibiotics and DLC
6 months Failure of therapy, perioperative and postoperative morbidity, quality of life
3
Norway RCT Observation with antibiotics versus LC
14 years Cumulative risk of having LC and gallstone-related complications
2
Chandler et al.23 USA RCT ELC (as soon as schedule allowed) versus DLC
n.r. Safety and cost-effectiveness of ELC versus DLC
3
Lai et al.24 Hong Kong RCT ELC (within 24 h of randomization) versus conservative treatment with antibiotics and DLC
≥ 8 weeks Outcome per treatment group, conversion, duration of surgery, LOS, complications
3
Lo et al.25 Hong Kong RCT ELC (within 72 h of admission) versus treatment with antibiotics and DLC
≥12 weeks Safety and cost efficacy of ELC and DLC
3
ELC, early laparoscopic cholecystectomy; DLC, delayed laparoscopic cholecystectomy; n.r., not reported; RCT, randomized clinical trial; ACC, acute calculous cholecystitis; LOS, length of hospital stay; LC, laparoscopic cholecystectomy; PC, percutaneous cholecystostomy.
© 2016 BJS Society Ltd www.bjs.co.uk BJS 2016; 103: 797–811 Published by John Wiley & Sons Ltd
800 A. H. van Dijk, P. R. de Reuver, T. N. Tasma, S. van Dieren, T. J. Hugh and M. A. Boermeester
Table 2 Risk of bias in included non-randomized studies
MINORS score15
Loss to follow-up<5%
of study size Total score
Kirkil et al.41 1 0 0 1 1 0 0 0 3 Hasbahçeci et al.40 2 2 2 1 2 0 0 0 9 Wang et al.34 2 0 1 2 0 2 0 0 7 Oymaci et al.42 2 0 2 2 0 0 0 0 6 Barcelo et al.35 2 0 1 2 2 2 0 0 9 Degrate et al.43 2 2 1 2 0 0 0 0 7 Al-Qahtani36 2 0 1 2 2 0 1 0 8 Al-Faouri et al.39 2 0 1 0 0 2 0 0 5 Kontopodis et al.37 2 2 0 1 1 1 0 0 7 Casillas et al.38 0 1 1 1 2 0 0 0 5 Rodríquez-Cerrillo et al.30 2 0 2 2 0 1 2 0 9 Barak et al.31 2 2 2 1 2 2 0 0 11 Paran et al.32 1 2 2 1 1 0 2 0 9 Serralta et al.33 2 2 2 1 2 0 0 0 9
Items are scored as follows: 0, not reported; 1, reported but inadequate; or 2, reported and adequate. MINORS, methodological index for non-randomized studies.
Table 3 Risk of bias of included randomized studies
Reference
(performance bias)
Selective reporting
(reporting bias)
Agrawal et al.18 ? ? − ? + + Chandler et al.23 ? ? − − ? ? Gul et al.19 + ? − − ? ? Gutt et al.20 + + − − ? ? Johansson et al.27 ? ? + ? + + Kolla et al.22 + ? − − + + Lai et al.24 + ? − ? + + Lo et al.25 + + − ? + + Mazeh et al.21 + + − − + + Vetrhus et al.28 ? + ? ? + +
–, High risk of bias; +, low risk of bias; ?, unclear risk of bias.
The search was also limited to include only articles published between January 1985 and October 2015, to represent current clinical practice.
Selection of eligible articles
Two reviewers independently assessed the available articles for inclusion by screening the title and abstract. All duplicates were removed. Full-text articles of possible relevant studies were reviewed for inclusion by using the inclusion and exclusion criteria by both reviewers.
Data collection
Data on authors, year and country of publication, study design, study population, definition of ACC, details of the
intervention and control group, type of antibiotic used, follow-up period and outcomes were extracted indepen- dently from the full-text articles. Both reviewers assessed the risk of bias using the MINORS (methodological index for non-randomized studies) score15 for non-randomized surgical studies, with an ideal total score of 16, and the risk-of-bias tool for randomized studies, available from the Cochrane Collaboration16.
Grading of evidence
Both reviewers assessed the level of evidence using the Oxford Centre for Evidence-Based Medicine system17. Level 1 represents the best evidence, and level 4 is the weakest level.
© 2016 BJS Society Ltd www.bjs.co.uk BJS 2016; 103: 797–811 Published by John Wiley & Sons Ltd
Antibiotic treatment for acute calculous cholecystitis 801
Patients with ACC
n = 2
Elective surgery
n = 26
No surgery
n = 14
Fig. 2 Flow chart showing details of treatment for all included patients. ACC, acute calculous cholecystitis; ELC, early laparoscopic cholecystectomy
Table 4 Outcomes of study on antibiotics versus no antibiotics
Mazeh et al.21
Need for emergency intervention
5 required PC 2 required PC
Recurrent ACC 4 2 Other biliary complications 2 biliary colic 1 biliary colic, 1
pancreatitis Postoperative complications 0 2 Patients not undergoing LC 6 16
PC, percutaneous cholecystostomy; ACC, acute calculous cholecystitis; LC, laparoscopic cholecystectomy.
Outcomes
The primary outcomes were the need for an emergency intervention (including cholecystectomy or percutaneous drainage) and recurrence of ACC, both preceding delayed elective cholecystectomy and despite antibiotic therapy. Secondary outcomes were morbidity before elective chole- cystectomy and the number of patients forgoing elective surgery after antibiotic treatment. To assess the effect in clinical practice of the use of antibiotics in ACC, the two primary endpoints (need for emergency intervention and recurrence of ACC) were combined and also analysed.
Statistical analysis
Primary outcomes were calculated using the proportion of events and the 95 per cent confidence interval. If only one arm of a randomized study was relevant for the outcome, only the data for that single arm were used for the meta-analysis. The estimated pooled event
rate was calculated with a random-effects model using R – biomedical statistics, version 3.1.1 (R Foundation for Statistical Computing, Vienna, Austria). Randomized trials, prospective and retrospective studies were pooled separately to minimize the effect of bias. A subgroup anal- ysis for both primary outcomes in subgroups of patients with a different severity of ACC was planned.
Forest plots were created for the two primary outcomes and the secondary outcomes. Each forest plot shows the effect size of the individual studies and an overall pooled event rate with a confidence interval.
Statistical heterogeneity between studies was determined by visual assessment and the I2 test, using R – biomedical statistics and reported per study design (randomized, prospective and retrospective). An I2 value of 40 per cent or less was considered as evidence of low heterogeneity, that of 40–70 per cent as moderate, and above 70 per cent as strong heterogeneity. All pooled event rates were shown in the forest plots despite the level of heterogeneity, owing to the diminished importance of statistical heterogeneity in single-arm meta-analysis. As several study designs and different comparisons were included, it was considered prudent to report only the pooled event rates with evi- dence of low and moderate heterogeneity. Funnel plots were created to assess publication bias.
Results
Study selection
A total of 17 990 references were identified through the electronic search on 9 October 2015. One relevant reference was found by handsearching. After reading the
© 2016 BJS Society Ltd www.bjs.co.uk BJS 2016; 103: 797–811 Published by John Wiley & Sons Ltd
802 A. H. van Dijk, P. R. de Reuver, T. N. Tasma, S. van Dieren, T. J. Hugh and M. A. Boermeester
Table 5 Outcomes of included studies on antibiotics versus other intervention or no intervention
Reference No. of patients
treatment
Kirkil et al.41 1557 total n.r. n.r. n.r. 8 of 14 (57) 297 (50⋅4)
589 initially antibiotics
Agrawal et al.18 50 total 0 (0) 0 (0) n.r. 2 (8) LC obligatory in study protocol
25 DLC
Hasbahçeci et al.40 118 total 10 (10⋅0) PC 12 (12⋅0) 15 (15⋅0) acute pancreatitis n.r. 34 (34⋅0)
100 conservative 17 (17⋅0) CBDS
Wang et al.34 779 total n.r. 31 (13⋅7) n.r. n.r. 146 (64⋅6)
226 antibiotics
Oymaci et al.42 165 total n.r. n.r. n.r. 15 (18) n.r.
82 DLC
Barcelo et al.35 105 total 21 (28) required EC 7 (9) 1 death from cholecystitis 5 of 55 (9) 19 (26)
74 DLC
Al-Qahtani36 168 total 3 (5) required EC 0 (0) 3 (5) biliary colic 0 (0) 12 (21)
56 DLC 2 (4) acute pancreatitis
Gul et al.19 60 total n.r. n.r. n.r. 4 (13) LC obligatory in study protocol
30 DLC
Gutt et al.20 618 total n.r. n.r. n.r. 86 (27⋅4) 23 (7⋅3)
314 DLC
Degrate et al.43 316 total n.r. 20 (37) n.r. 8 (15) LC obligatory in study protocol
54 DLC
Al-Faouri et al.39 317 total 11 (5⋅9) required EC 11 (5⋅9) 31 (16⋅7) biliary colic 22 (11⋅8) LC obligatory in study protocol
186 DLC 3 (1⋅6) acute pancreatitis
4 (2⋅2) ERCP owing to CBDS
Rodríquez-Cerrillo et al.30 136 total 2 (8) required EC n.r. n.r. n.r. n.r.
25 home treatment
Kontopodis et al.37 315 total, all antibiotics 16 (5⋅1) required EC n.r. 24 of 299 (8⋅0) ERCP due to CBDS
2 of 299 (0⋅6) LC obligatory in study protocol
Barak et al.31 103 total, all antibiotics 27 (26⋅2) required PC n.r. n.r. n.r. n.r.
Casillas et al.38 173 total 45 (44⋅1), 26 EC, 19 PC n.r. n.r. 1 of 46 (2) 11 of 57 (19)
102 antibiotics
Paran et al.32 224 total, all antibiotics 58 (25⋅9) required PC (in 3 after PC then EC)
n.r. n.r. n.r. n.r.
Kolla et al.22 40 total 0 (0) 0 (0) n.r. 3 (15) LC obligatory in study protocol
20 DLC
Serralta et al.33 169 total 8 (9) required EC n.r. 1 (1) CBDS 13 (15) LC obligatory in study protocol
87 DLC
Johansson et al.26,27 145 total 14 (20) required EC 4 (6) n.r. 7 (10) LC obligatory in study protocol
71 DLC
Vetrhus et al.28, 64 total n.r. 9 (27) 4 (12) biliary pain 2 of 11 (18) 22 (67)
Schmidt et al.29 33 antibiotics 2 of 23 (9) CBDS
Chandler et al.23 43 total n.r. 7 (32) n.r. 2 LC obligatory in study protocol
22 DLC
Lai et al.24 104 total 7 (14) required EC 1 (2) n.r. 3 of 38 (8) 5 (10)
51 DLC
Lo et al.25 86 total 8 (20) required EC 4 (10) 1 (2) biliary colic 12 (29) 5 of 50 (10)
2 (5) acute cholangitis41 DLC
Values in parentheses are percentages. ACC, acute calculous cholecystitis; LC, laparoscopic cholecystectomy; n.r., not reported; DLC, delayed cholecystectomy; PC, percutaneous cholecystostomy; CBDS, common bile duct stones; EC, emergency cholecystectomy; ERCP, endoscopic retrograde cholangiopancreatography.
© 2016 BJS Society Ltd www.bjs.co.uk BJS 2016; 103: 797–811 Published by John Wiley & Sons Ltd
Antibiotic treatment for acute calculous cholecystitis 803
Agrawal et al.18 0 of 25
5 of 42
0 of 20
14 of 71
7 of 51
8 of 41
2 of 25
27 of 103
58 of 224
8 of 87
10 of 100
21 of 74
3 of 56
11 of 186
16 of 315
45 of 102
0·15 (0·10, 0·22) 22·0
0·08 (0·01, 0·26)
0·26 (0·18, 0·36)
0·26 (0·20, 0·32)
0·09 (0·04, 0·17)
0·18 (0·11, 0·29) 76·8
0·10 (0·05, 0·18)
0·28 (0·19, 0·40)
0·05 (0·01, 0·15)
0·06 (0·03, 0·10)
0·05 (0·03, 0·08)
0·44 (0·34, 0·54)
0·12 (0·05, 0·29)
0·14 (0·09, 0·20)
95·1
88·3
Randomized studies
Estimated pooled…
A. H. van Dijk1, P. R. de Reuver3, T. N. Tasma1, S. van Dieren1,2, T. J. Hugh4 and M. A. Boermeester1
1Department of Surgery and 2Clinical Research Unit, Academic Medical Centre, Amsterdam, and 3Department of Surgery, Radboud University Medical Centre, Nijmegen, The Netherlands, and 4Upper Gastrointestinal Surgery Unit, Royal North Shore Hospital, University of Sydney, Sydney, New South Wales, Australia Correspondence to: Ms A. H. van Dijk, Academic Medical Centre, Meibergdreef 9, 1105 AZ Amsterdam, The Netherlands (e-mail: [email protected])
Background: Intravenous antibiotics are frequently used in the initial management of acute calculous cholecystitis (ACC), although supportive care alone preceding delayed elective cholecystectomy may be sufficient. This systematic review assessed the success rate of antibiotics in the treatment of ACC. Methods: A systematic search of MEDLINE, Embase and Cochrane Library databases was performed. Primary outcomes were the need for emergency intervention and recurrence of ACC after initial non-operative management of ACC. Risk of bias was assessed. Pooled event rates were calculated using a random-effects model. Results: Twelve randomized trials, four prospective and ten retrospective studies were included. Only one trial including 84 patients compared treatment with antibiotics to that with no antibiotics; there was no significant difference between the two groups in terms of length of hospital stay and morbidity. Some 5830 patients with ACC were included, of whom 2997 had early cholecystectomy, 2791 received initial antibiotic treatment, and 42 were treated conservatively. Risk of bias was high in most studies, and all but three studies had a low level of evidence. For randomized studies, pooled event rates were 15 (95 per cent c.i. 10 to 22) per cent for the need for emergency intervention and 10 (5 to 20) per cent for recurrence of ACC. The pooled event rate for both outcomes combined was 20 (13 to 30) per cent. Conclusion: Antibiotics are not indicated for the conservative management of ACC or in patients scheduled for cholecystectomy.
Paper accepted 5 February 2016 Published online 30 March 2016 in Wiley Online Library (www.bjs.co.uk). DOI: 10.1002/bjs.10146
Introduction
Acute calculous cholecystitis (ACC) is diagnosed in 3–10 per cent of patients with acute abdominal pain1
and accounts for approximately one-third of all surgical emergency admissions2. There is controversy regard- ing the timing of cholecystectomy in ACC. Both early cholecystectomy (operation within 7 days of onset of complaints) and initial conservative therapy followed by delayed elective cholecystectomy are reasonable treatment options3.
A recent meta-analysis4 showed that morbidity and con- version to an open procedure is the same for early and delayed cholecystectomy. Early cholecystectomy was asso- ciated with a significantly shorter total hospital stay, and is the treatment of choice according to recent guidelines5,6. Early cholecystectomy is still not common practice in
many countries. Non-operative management, including bowel rest, intravenous hydration, analgesia and intra- venous antibiotics, are used in the initial management of ACC followed by delayed cholecystectomy7–9.
Although evidence for the use of antibiotics in the con- servative management of ACC is limited, guidelines rec- ommend their use10,11. The most frequently cited reason for commencing antibiotics is to prevent complications, either in the period before elective cholecystectomy or perioperatively12,13. The current view, however, is that the natural course of ACC is often uncomplicated, as ACC is thought to be a primarily inflammatory process. Consequently, supportive care without antibiotics may be sufficient before elective cholecystectomy.
The aim of this systematic review was to assess the value of antibiotics in terms of their ability to decrease the need for emergency intervention, to prevent recurrence
© 2016 BJS Society Ltd BJS 2016; 103: 797–811 Published by John Wiley & Sons Ltd
798 A. H. van Dijk, P. R. de Reuver, T. N. Tasma, S. van Dieren, T. J. Hugh and M. A. Boermeester
References identified through
Cochrane Library n = 162
Duplicates n = 8390
Full-text articles assessed for eligibility
n = 89
Articles included
n = 26
Not available n = 4
Articles included in meta-analysis
Fig. 1 PRISMA flow diagram for the review
of ACC, to minimize complications, or to avoid elective cholecystectomy after initial conservative treatment.
Methods
The study protocol for this systematic review followed the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) checklist14.
Eligibility criteria
Articles were eligible for inclusion when: the study population consisted of adults with ultrasound-proven ACC with clinical signs; ACC was defined in the arti- cle; the article reported original data and the full text was available; and the article described one group of patients with ACC who were initially treated with antibiotics.
Only studies published after the introduction of laparoscopic cholecystectomy in 1985 were included, to represent current clinical practice. When the full
text of an article was not available, the original author was contacted to request access to the data. No limits regarding language or study design were defined. Arti- cles describing a specific patient population, such as malignancies or patients in intensive care, were excluded. Studies describing prophylactic antibiotics before laparo- scopic cholecystectomy were outside the scope of this review.
Search
The MEDLINE, Embase and Cochrane Library databases were searched for articles (October 2015). The follow- ing keywords and synonyms were used for MEDLINE and Embase: cholecystitis, cholecystectomy, antibacte- rial agents, antibiotics and conservative management; Appendix S1 (supporting information) outlines the complete search strategy. For every keyword (medical subject heading (MeSH) for MEDLINE and Emtree for EMBASE), free text words were also searched. The Cochrane Library was checked with the same key words. Cross-references were hand-searched.
© 2016 BJS Society Ltd www.bjs.co.uk BJS 2016; 103: 797–811 Published by John Wiley & Sons Ltd
Antibiotic treatment for acute calculous cholecystitis 799
Table 1 Characteristics of included studies, ranked by year of publication
Reference Country Study design Type of intervention Length of follow-up Primary goal
Level of evidence17
Kirkil et al.41 Turkey Retrospective ELC versus interval cholecystectomy and DLC
n.r. Comparison between three treatment strategies
4
Agrawal et al.18 India RCT ELC versus DLC in ACC 8 weeks Success of DLC 3
Hasbahçeci et al.40 Turkey Retrospective ELC versus conservative treatment or DLC in ACC
6 months LOS, gallstone-related complications and mortality
4
Wang et al.34 Taiwan Retrospective Antibiotic treatment in all patients with ACC
Median 308⋅5 days Recurrence rate of ACC 4
Oymaci et al.42 Turkey Retrospective ELC versus DLC in ACC n.r. Comparison of ELC versus DLC 4
Barcelo et al.35 Spain Retrospective ELC (within 48 h of admission) versus conservative treatment and DLC
≥ 6 months Therapeutic approach and failure of chosen therapy
4
Al-Qahtani36 Saudi Arabia Retrospective ELC (within 7 days of onset of symptoms) versus conservative treatment with antibiotics followed by DLC
n.r. Success of ELC and conservative treatment
4
Gul et al.19 India RCT ELC (within 72 h of admission) versus conservative treatment with antibiotics and DLC
1 year To evaluate safety and feasibility of LC for ACC and to compare ELC with DLC
3
Gutt et al.20 Germany RCT ELC (within 24 h of admission) versus DLC
75 days To compare 75-day morbidity between ELC and DLC
2
Degrate et al.43 Italy Retrospective ELC (within initial admission) versus DLC
n.r. To compare ELC and DLC 4
Al-Faouri et al.39 Jordan Retrospective ELC versus DLC 2 months to 1 year To compare ELC and DLC in ACC 4
Mazeh et al.21 Israel RCT Supportive care+ antibiotics versus supportive care only
Mean 17⋅1 (range 4–28) months
LOS at index admission and combined LOS for all admissions
2
1 month Security and efficacy of home treatment of ACC
3
n.r. Failure of conservative treatment 4
Barak et al.31 Israel Prospective All patients treated conservatively; when treatment failure, then PC
n.r. Failure of conservative management and predictive factors for failure
3
Casillas et al.38 USA Retrospective ELC (within 24 h of admission) versus antibiotics
n.r. Duration of symptoms, LOS and complications
4
Paran et al.32 Israel Prospective Conservative treatment with antibiotics. If failure of treatment, then emergency LC in patients with symptoms for≤72 h. If symptoms for>72 h, then PC
n.r. Failure of conservative therapy 3
Kolla et al.22 India RCT ELC (within 24 h of randomization) versus conservative treatment with antibiotics and DLC
n.r. Operating time, conversions, LOS, complications
3
Serralta et al.33 USA Prospective ELC (within 72 h of onset of symptoms) versus conservative treatment with antibiotics and DLC
n.r. LOS and postoperative morbidity 3
Johansson et al.26,27 Sweden RCT ELC (within 48 h after randomization, but no later than 7 days after onset of symptoms) versus conservative treatment with antibiotics and DLC
6 months Failure of therapy, perioperative and postoperative morbidity, quality of life
3
Norway RCT Observation with antibiotics versus LC
14 years Cumulative risk of having LC and gallstone-related complications
2
Chandler et al.23 USA RCT ELC (as soon as schedule allowed) versus DLC
n.r. Safety and cost-effectiveness of ELC versus DLC
3
Lai et al.24 Hong Kong RCT ELC (within 24 h of randomization) versus conservative treatment with antibiotics and DLC
≥ 8 weeks Outcome per treatment group, conversion, duration of surgery, LOS, complications
3
Lo et al.25 Hong Kong RCT ELC (within 72 h of admission) versus treatment with antibiotics and DLC
≥12 weeks Safety and cost efficacy of ELC and DLC
3
ELC, early laparoscopic cholecystectomy; DLC, delayed laparoscopic cholecystectomy; n.r., not reported; RCT, randomized clinical trial; ACC, acute calculous cholecystitis; LOS, length of hospital stay; LC, laparoscopic cholecystectomy; PC, percutaneous cholecystostomy.
© 2016 BJS Society Ltd www.bjs.co.uk BJS 2016; 103: 797–811 Published by John Wiley & Sons Ltd
800 A. H. van Dijk, P. R. de Reuver, T. N. Tasma, S. van Dieren, T. J. Hugh and M. A. Boermeester
Table 2 Risk of bias in included non-randomized studies
MINORS score15
Loss to follow-up<5%
of study size Total score
Kirkil et al.41 1 0 0 1 1 0 0 0 3 Hasbahçeci et al.40 2 2 2 1 2 0 0 0 9 Wang et al.34 2 0 1 2 0 2 0 0 7 Oymaci et al.42 2 0 2 2 0 0 0 0 6 Barcelo et al.35 2 0 1 2 2 2 0 0 9 Degrate et al.43 2 2 1 2 0 0 0 0 7 Al-Qahtani36 2 0 1 2 2 0 1 0 8 Al-Faouri et al.39 2 0 1 0 0 2 0 0 5 Kontopodis et al.37 2 2 0 1 1 1 0 0 7 Casillas et al.38 0 1 1 1 2 0 0 0 5 Rodríquez-Cerrillo et al.30 2 0 2 2 0 1 2 0 9 Barak et al.31 2 2 2 1 2 2 0 0 11 Paran et al.32 1 2 2 1 1 0 2 0 9 Serralta et al.33 2 2 2 1 2 0 0 0 9
Items are scored as follows: 0, not reported; 1, reported but inadequate; or 2, reported and adequate. MINORS, methodological index for non-randomized studies.
Table 3 Risk of bias of included randomized studies
Reference
(performance bias)
Selective reporting
(reporting bias)
Agrawal et al.18 ? ? − ? + + Chandler et al.23 ? ? − − ? ? Gul et al.19 + ? − − ? ? Gutt et al.20 + + − − ? ? Johansson et al.27 ? ? + ? + + Kolla et al.22 + ? − − + + Lai et al.24 + ? − ? + + Lo et al.25 + + − ? + + Mazeh et al.21 + + − − + + Vetrhus et al.28 ? + ? ? + +
–, High risk of bias; +, low risk of bias; ?, unclear risk of bias.
The search was also limited to include only articles published between January 1985 and October 2015, to represent current clinical practice.
Selection of eligible articles
Two reviewers independently assessed the available articles for inclusion by screening the title and abstract. All duplicates were removed. Full-text articles of possible relevant studies were reviewed for inclusion by using the inclusion and exclusion criteria by both reviewers.
Data collection
Data on authors, year and country of publication, study design, study population, definition of ACC, details of the
intervention and control group, type of antibiotic used, follow-up period and outcomes were extracted indepen- dently from the full-text articles. Both reviewers assessed the risk of bias using the MINORS (methodological index for non-randomized studies) score15 for non-randomized surgical studies, with an ideal total score of 16, and the risk-of-bias tool for randomized studies, available from the Cochrane Collaboration16.
Grading of evidence
Both reviewers assessed the level of evidence using the Oxford Centre for Evidence-Based Medicine system17. Level 1 represents the best evidence, and level 4 is the weakest level.
© 2016 BJS Society Ltd www.bjs.co.uk BJS 2016; 103: 797–811 Published by John Wiley & Sons Ltd
Antibiotic treatment for acute calculous cholecystitis 801
Patients with ACC
n = 2
Elective surgery
n = 26
No surgery
n = 14
Fig. 2 Flow chart showing details of treatment for all included patients. ACC, acute calculous cholecystitis; ELC, early laparoscopic cholecystectomy
Table 4 Outcomes of study on antibiotics versus no antibiotics
Mazeh et al.21
Need for emergency intervention
5 required PC 2 required PC
Recurrent ACC 4 2 Other biliary complications 2 biliary colic 1 biliary colic, 1
pancreatitis Postoperative complications 0 2 Patients not undergoing LC 6 16
PC, percutaneous cholecystostomy; ACC, acute calculous cholecystitis; LC, laparoscopic cholecystectomy.
Outcomes
The primary outcomes were the need for an emergency intervention (including cholecystectomy or percutaneous drainage) and recurrence of ACC, both preceding delayed elective cholecystectomy and despite antibiotic therapy. Secondary outcomes were morbidity before elective chole- cystectomy and the number of patients forgoing elective surgery after antibiotic treatment. To assess the effect in clinical practice of the use of antibiotics in ACC, the two primary endpoints (need for emergency intervention and recurrence of ACC) were combined and also analysed.
Statistical analysis
Primary outcomes were calculated using the proportion of events and the 95 per cent confidence interval. If only one arm of a randomized study was relevant for the outcome, only the data for that single arm were used for the meta-analysis. The estimated pooled event
rate was calculated with a random-effects model using R – biomedical statistics, version 3.1.1 (R Foundation for Statistical Computing, Vienna, Austria). Randomized trials, prospective and retrospective studies were pooled separately to minimize the effect of bias. A subgroup anal- ysis for both primary outcomes in subgroups of patients with a different severity of ACC was planned.
Forest plots were created for the two primary outcomes and the secondary outcomes. Each forest plot shows the effect size of the individual studies and an overall pooled event rate with a confidence interval.
Statistical heterogeneity between studies was determined by visual assessment and the I2 test, using R – biomedical statistics and reported per study design (randomized, prospective and retrospective). An I2 value of 40 per cent or less was considered as evidence of low heterogeneity, that of 40–70 per cent as moderate, and above 70 per cent as strong heterogeneity. All pooled event rates were shown in the forest plots despite the level of heterogeneity, owing to the diminished importance of statistical heterogeneity in single-arm meta-analysis. As several study designs and different comparisons were included, it was considered prudent to report only the pooled event rates with evi- dence of low and moderate heterogeneity. Funnel plots were created to assess publication bias.
Results
Study selection
A total of 17 990 references were identified through the electronic search on 9 October 2015. One relevant reference was found by handsearching. After reading the
© 2016 BJS Society Ltd www.bjs.co.uk BJS 2016; 103: 797–811 Published by John Wiley & Sons Ltd
802 A. H. van Dijk, P. R. de Reuver, T. N. Tasma, S. van Dieren, T. J. Hugh and M. A. Boermeester
Table 5 Outcomes of included studies on antibiotics versus other intervention or no intervention
Reference No. of patients
treatment
Kirkil et al.41 1557 total n.r. n.r. n.r. 8 of 14 (57) 297 (50⋅4)
589 initially antibiotics
Agrawal et al.18 50 total 0 (0) 0 (0) n.r. 2 (8) LC obligatory in study protocol
25 DLC
Hasbahçeci et al.40 118 total 10 (10⋅0) PC 12 (12⋅0) 15 (15⋅0) acute pancreatitis n.r. 34 (34⋅0)
100 conservative 17 (17⋅0) CBDS
Wang et al.34 779 total n.r. 31 (13⋅7) n.r. n.r. 146 (64⋅6)
226 antibiotics
Oymaci et al.42 165 total n.r. n.r. n.r. 15 (18) n.r.
82 DLC
Barcelo et al.35 105 total 21 (28) required EC 7 (9) 1 death from cholecystitis 5 of 55 (9) 19 (26)
74 DLC
Al-Qahtani36 168 total 3 (5) required EC 0 (0) 3 (5) biliary colic 0 (0) 12 (21)
56 DLC 2 (4) acute pancreatitis
Gul et al.19 60 total n.r. n.r. n.r. 4 (13) LC obligatory in study protocol
30 DLC
Gutt et al.20 618 total n.r. n.r. n.r. 86 (27⋅4) 23 (7⋅3)
314 DLC
Degrate et al.43 316 total n.r. 20 (37) n.r. 8 (15) LC obligatory in study protocol
54 DLC
Al-Faouri et al.39 317 total 11 (5⋅9) required EC 11 (5⋅9) 31 (16⋅7) biliary colic 22 (11⋅8) LC obligatory in study protocol
186 DLC 3 (1⋅6) acute pancreatitis
4 (2⋅2) ERCP owing to CBDS
Rodríquez-Cerrillo et al.30 136 total 2 (8) required EC n.r. n.r. n.r. n.r.
25 home treatment
Kontopodis et al.37 315 total, all antibiotics 16 (5⋅1) required EC n.r. 24 of 299 (8⋅0) ERCP due to CBDS
2 of 299 (0⋅6) LC obligatory in study protocol
Barak et al.31 103 total, all antibiotics 27 (26⋅2) required PC n.r. n.r. n.r. n.r.
Casillas et al.38 173 total 45 (44⋅1), 26 EC, 19 PC n.r. n.r. 1 of 46 (2) 11 of 57 (19)
102 antibiotics
Paran et al.32 224 total, all antibiotics 58 (25⋅9) required PC (in 3 after PC then EC)
n.r. n.r. n.r. n.r.
Kolla et al.22 40 total 0 (0) 0 (0) n.r. 3 (15) LC obligatory in study protocol
20 DLC
Serralta et al.33 169 total 8 (9) required EC n.r. 1 (1) CBDS 13 (15) LC obligatory in study protocol
87 DLC
Johansson et al.26,27 145 total 14 (20) required EC 4 (6) n.r. 7 (10) LC obligatory in study protocol
71 DLC
Vetrhus et al.28, 64 total n.r. 9 (27) 4 (12) biliary pain 2 of 11 (18) 22 (67)
Schmidt et al.29 33 antibiotics 2 of 23 (9) CBDS
Chandler et al.23 43 total n.r. 7 (32) n.r. 2 LC obligatory in study protocol
22 DLC
Lai et al.24 104 total 7 (14) required EC 1 (2) n.r. 3 of 38 (8) 5 (10)
51 DLC
Lo et al.25 86 total 8 (20) required EC 4 (10) 1 (2) biliary colic 12 (29) 5 of 50 (10)
2 (5) acute cholangitis41 DLC
Values in parentheses are percentages. ACC, acute calculous cholecystitis; LC, laparoscopic cholecystectomy; n.r., not reported; DLC, delayed cholecystectomy; PC, percutaneous cholecystostomy; CBDS, common bile duct stones; EC, emergency cholecystectomy; ERCP, endoscopic retrograde cholangiopancreatography.
© 2016 BJS Society Ltd www.bjs.co.uk BJS 2016; 103: 797–811 Published by John Wiley & Sons Ltd
Antibiotic treatment for acute calculous cholecystitis 803
Agrawal et al.18 0 of 25
5 of 42
0 of 20
14 of 71
7 of 51
8 of 41
2 of 25
27 of 103
58 of 224
8 of 87
10 of 100
21 of 74
3 of 56
11 of 186
16 of 315
45 of 102
0·15 (0·10, 0·22) 22·0
0·08 (0·01, 0·26)
0·26 (0·18, 0·36)
0·26 (0·20, 0·32)
0·09 (0·04, 0·17)
0·18 (0·11, 0·29) 76·8
0·10 (0·05, 0·18)
0·28 (0·19, 0·40)
0·05 (0·01, 0·15)
0·06 (0·03, 0·10)
0·05 (0·03, 0·08)
0·44 (0·34, 0·54)
0·12 (0·05, 0·29)
0·14 (0·09, 0·20)
95·1
88·3
Randomized studies
Estimated pooled…
Related Documents
