Efficacy of interventions to improve adherence to inhaled corticosteroids in adult asthmatics: Impact of using components of the chronic care model

Page 1

Respiratory Medicine (2012) 106, 1211e1225

Available online at www.sciencedirect.com

journal homepage: www.elsevier .com/locate/rmed

REVIEW

Efficacy of interventions to improve adherence toinhaled corticosteroids in adult asthmatics: Impactof using components of the chronic care model

Gregory Moullec a,b,c,g, Gabrielle Gour-Provencal c,g, Simon L. Bacon a,b,c,d,Tavis S. Campbell a,e, Kim L. Lavoie a,b,d,f,*

aMontreal Behavioural Medicine Centre, Montreal, Quebec, CanadabResearch Centre, Hopital du Sacre-Coeur de Montreal e A University of Montreal affiliated hospital, Montreal,Quebec, CanadacDepartment of Exercise Science, Concordia University, Montreal, Quebec, CanadadResearch Centre, Montreal Heart Institute e A University of Montreal affiliated hospital, Montreal, Quebec, CanadaeDepartment of Psychology, University of Calgary, Calgary, Alberta, CanadafDepartment of Psychology, University of Quebec at Montreal (UQAM), Montreal, Quebec, Canada

Received 26 February 2012; accepted 1 June 2012Available online 5 July 2012

KEYWORDSMedicationadherence;Asthma;Inhaledcorticosteroids;Disease management;Behavioral medicine;Meta-analysis

* Corresponding author. Montreal BeBox 8888, Succursale Center-Ville, Mo

E-mail address: [email protected] Both authors contributed equally t

0954-6111/$ - see front matter ª 201http://dx.doi.org/10.1016/j.rmed.201

Summary

Background: Adherence to inhaled corticosteroids (ICS) remains poor among asthmatics, yetlittle is known about the efficacy of interventions to improve adherence. Implementing theChronic Care Model (CCM) components among patients with respiratory disorders has beenassociated with an improvement in outcomes, yet little is known about its effects on ICS adher-ence in asthmatics.Objective: We conducted a systematic review to assess the efficacy of interventions toimprove ICS adherence among adult-asthmatics, and whether the use of CCM components(i.e., teaching self-management skills, providing decision support, delivery system design,and clinical information systems) resulted in greater ICS adherence.Methods: All English language articles testing the efficacy of an intervention including ICSmedication on outcome from MEDLINE and PsychINFO databases through Aug-2010 were re-viewed. Interventions were categorized based on the inclusion of CCM components. We stan-dardized treatment effects to obtain effect-size’s (ES’s) and we combined the ES’s of studiesaccording to the number of CCM components included in their interventions.

havioural Medicine Centre, Department of Psychology, University of Quebec at Montreal (UQAM), P.O.ntreal, Quebec H3C 3P8, Canada. Tel.: þ1 514 987 3000x3835; fax: þ1 514 987 7953.s.qc.ca (K.L. Lavoie).o first authorship of the manuscript.

2 Elsevier Ltd. All rights reserved.2.06.001

Page 2

1212 G. Moullec et al.

Results: Eighteen studies met inclusion criteria. Inclusion of a greater number of CCM compo-nents within interventions was associated with stronger effects on ICS adherence outcomes,with interventions featuring one, two, and four CCM components having medium(ES Z 0.29; 95%CI, 0.16e0.42), large (0.53; 0.40e0.66), and very-large (0.83; 0.69e0.98)effects respectively.Conclusions: Findings provide support for using the CCM as a framework for the design and im-plementation of interventions to improve adherence among adult-asthmatics.ª 2012 Elsevier Ltd. All rights reserved.

Contents

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1212Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1213

Literature search . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1213Study selection and data extraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1213Data synthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1213

Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1214

Study designs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1214Patient characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1214Intervention characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1214Adherence measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1214Efficacy of interventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1215

Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1215

Main results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1215Importance of using a theoretical framework for intervention design . . . . . . . . . . . . . . . . . . . . . . . . . . 1215Importance of using an appropriate intervention development process . . . . . . . . . . . . . . . . . . . . . . . . . 1215

Recommendations for future studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1221

Choice of adherence measure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1221Choice of comparison group . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1221Focus on key intervention ingredients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1223Limitations and strengths of the present review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1223

Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1223Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1224Conflict of interest statement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1224References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1224

Introduction

Asthma is a chronic disorder characterized by reversibleand intermittent airway obstruction, airway inflammation,and hyper-reactivity of the airways in response to a varietyof stimuli (e.g., pollen, dust, pets, smoke, exercise, andemotions).1 Although important advances in pharmacolog-ical treatment and diagnosis have occurred over the pastyears, 300 million people worldwide suffer from asthma,making this condition one of the most frequent chronicdiseases in the world.2,3 Poorly controlled asthma patientsare more likely to experience further exacerbations,a poorer quality of life, and require greater medicalattention compared to well-controlled patients, thereforeplacing a higher burden on the health care system.4e6

According to recent Canadian statistics, less than 42% ofasthma patients have adequately controlled asthma.7

There are several options to optimize asthma controlincluding daily peak expiratory flow (PEF) monitoring,

action plans in the event of symptom exacerbation, mini-mizing exposure to known asthma triggers (e.g., indoorallergens, pollution, smoke), and daily adherence tocontroller medication (i.e., inhaled corticosteroids,ICS).5,6,8 Among these strategies, daily adherence to ICS isconsidered to be one of the most important strategies toimprove control of asthma symptoms.6,8e16 However, anestimated 60%e80% of patients do not take their ICS regi-mens as prescribed (i.e., patients take less than 75% oftheir ICS prescriptions more than 80% of the time),9,11,17e20

including patients suffering from severe asthma19,21 whowould most benefit from this treatment.8,22

Over the past 20 years, researchers have focused ondesigning behavioral interventions to enhance controller(i.e., ICS) adherence among adult asthmatics.23 However,the vast majority did not provide any theoretical rationalefor the development of their intervention.24 One recom-mended method25,26 for exploring mechanisms for positivechange associated with behavioral interventions, is to

Page 3

Improve adherence to ICS in adult asthmatics 1213

evaluate intervention components quantitatively e withina theoretical framework. Wagner’s Chronic Care Model(CCM) is a well-known framework aimed at improving thecare of patients with chronic diseases. The CCM integratesa number of elements into a model designed to encouragemore productive interactions between healthcare teamsand patients.27 Previous systematic reviews have shownthat implementing the CCM components in patients withasthma28,29 and other respiratory diseases such as chronicobstructive pulmonary disease (COPD)29,30 has been asso-ciated with significantly improved outcomes (e.g., qualityof life, exercise tolerance, control of symptoms). However,little is known about the specific effects of implementingthis model on ICS adherence in asthmatic patients.

We conducted a systematic review and meta-analysis tobetter understand the strengths and limitations of testedinterventions for improving ICS use among adult asthmapatients, with the goal of identifying those that are morelikely to be efficacious. We examined whether interven-tions with a greater number of CCM components predictedbetter adherence outcomes among adult asthmaticscompared to interventions with fewer CCM components.We report their relative impact using a standardizedmeasure of effect size, and discuss the clinical impacts andimplications for future research.

Methods

Literature search

This review included all English-language peer-reviewedarticles from the following databases (via the Ovide portal):MEDLINE (1948-August 2010, week 4) and PsychINFO (1967-August 2010, week 4). Search terms were: “asthma orasthmatics or adult asthma”, “intervention or trial orprogram”, “medication adherence or treatment adherenceor treatment compliance”, “medication compliance orcorticosteroids adherence or corticosteroids compliance”.We combined these terms by using the Boolean ‘AND’operator. We further complemented this search by

Table 1 Interventions categorized into the components of the

CCM components Interventions

Self-management (SM) - Education (promote pteach specific prevent

- Behavioral support (pself-monitoring, e.g.,

- Motivational (linking sDecision support (DS) - Use of evidence-based

- Integrate specialty ex- Use of theory-driven e

Delivery system design (DSD) - Use of case managers- Multidisciplinary teampreventative measure

- Scheduling of plannedClinical information system (CIS) - Clinical registries (dat

- Provide timely remind- Share information wit

reviewing the reference sections of the identified articlesand reviews, and by consulting experts in the field.

Study selection and data extraction

Two reviewers (G.G.P. and K.L.L.) independently selectedstudies for inclusion with the following criteria: (1) it had totest the efficacy or effectiveness of an intervention thatincluded asthma controller medication (i.e., ICS) as one ofits outcomes; (2) it had to focus on adult populations(studies focusing on pediatric or adolescent populationswere excluded as adherence behavior in children is oftendependent on their parent’s attitudes and behavior)31,32;and (3) in studies that included combined disease pop-ulations (e.g. asthma and COPD), results had to have beenpresented separately for asthmatics. If an article wasselected by either reviewer, it was included in full-textreview and evaluated by both reviewers. Two reviewers(G.G.P. and G.M.) independently extracted data on a stan-dardized abstraction form created for this study. Inter-ventions were categorized based on the number of CCMcomponents they included. These components includedteaching self-management skills, providing decisionsupport, delivery system design, and clinical informationsystems (see descriptions in Table 1). The remaining CCMcomponents, i.e., ‘healthcare organization’ and ‘commu-nity resources’ (see www.improvingchroniccare.org), werenot included in the studies selected in the current reviewand as such were not included in this study. The quality ofselected studies was assessed using the Downs and Black’schecklist.33 Studies were assigned a quality ranking basedon their final score: high, �16; moderate, 8e16; low, <8.

Data synthesis

Adherence measures were expressed across studies withdifferent units; we therefore standardized the resultingtreatmenteffects (i.e., change inadherence level) toobtainaneffect size (i.e., ES e standard mean difference). So, whenpossible, the ES for the intervention with its 95% confidenceinterval (CI) was calculated from the information provided in

CCM.

atients’ understanding of their respiratory condition andion and treatment strategies)roviding tools to modify behaviours and optimizepeak flow)pecific goals for behavioural changes to clinical information)guidelines to help decision-making

pertise (e.g., referrals for management of co-morbidities)ducation program, structured telephone follow-ups (increased involvement of pharmacists) to coordinates for chronic careasthma follow-up visitsabases available for all providers)ers for providers and patientsh patients and providers

Page 4

Figure 1 Selection process for including studies in system-atic review and meta-analyses. Notes: Reasons: 1 Z didn’t testthe efficacy or effectiveness of an intervention that includedasthma controller medication; 2 Z pediatric or adolescentpopulations; 3 Z combined disease populations; 4 Z non-English.

1214 G. Moullec et al.

the articlee according to establishedmethods.34 Quantitativeanalyses were performed, as recommended by Wilson andLipsey,34 by pooling the ES’s of studies according to the numberof CCM components included in their interventions. For studiesthat provided ranges instead of standard deviations, we con-verted ranges to standard deviations according to the methoddescribed by Walter et al. (2007).35 If authors did not specifythe number of patients included in the final analyses, we usedthe baseline sample size of each group to calculate the ES. Theinconsistency of effects between study findings was measuredby using the I2 statistic as proposed by Higgins et al.36 Theabsolute value of each statistically significant ES was catego-rized as small (<0.20), medium (0.20 to<0.50), large (0.50 to<0.80), or very large (�0.80).37 Beyond the standard inter-pretation (very large in size), anESequal to0.80would indicatethat the intervention groupmean is 0.8 of a standard deviationabove the control group mean.

Results

The literature search yielded a total of 188 references, thetitles and abstracts of which were reviewed for studycontent and relevance and all duplicates and reviews wereremoved. A final sample of 18 studies38e52 met inclusioncriteria and were included in this review (see Fig. 1 as perthe PRISMA statement).53

Study designs

All 18 studies included in this review employed a random-ized control trial (RCT) design. Table 2 summarizes studypopulation characteristics, intervention characteristics,and type of adherence measures used. Among the includedstudies, 1038e47 involved interventions that included oneCCM component (with different combinations of sub-components under self-management support, i.e., educa-tion, behavior therapy, motivational interviewing); sevenstudies38,40,48e52 included two CCM components, and twostudies54,55 included four CCM components.

Patient characteristics

All but one study (94%) was conducted among moderate tosevere asthma patients.38e41,43e52 One was conductedamong both moderate-severe asthmatics and COPD outpa-tients.42 The average sample size across studies was 167patients (range Z 25e612). Among studies that reportedpatient age (14/18, 78%), participants ranged in age from35 to 50 years. Women were somewhat over-represented inmost studies, with participation rates ranging from52%47e73%.50 Participant retention rates over the follow-upperiods ranged from 74%51e98%.41

Intervention characteristics

A summary of intervention characteristics can be found inTable 2. Overall, interventions were quite heterogeneous.We categorized the interventions based on the number ofCCM components they included. All studies included a self-management component (i.e., education, and/or behavior

therapy and/or motivational interviewing). Most studies(i.e., 14/18, 78%) conducted individual inter-ventions,38,39,41,43e47,49e51 and four studies conducted bothindividual and group interventions.40,42,48,52 The duration ofthe intervention varied greatly across studies and rangedfrom a single 30 min session56 to more than 12 h41

(mean Z 3 h; median Z 2 h). In addition, follow-uplength ranged from 1 week41 to 2 years40,55 (mean Z 10months; median Z 9 months). In order to increase thehomogeneity between studies included in our quantitativeanalyses, for studies with follow-ups longer than 12 months,we analysed 12 month data.

Adherence measures

Measuresofadherencealso variedgreatly across studies.Moststudies (i.e, 7/18, 39%) measured adherence using self-reports. Five studies (28%) measured adherence accordingto pharmacy refill rates, with patients being considered

Page 5

Improve adherence to ICS in adult asthmatics 1215

adherent if they filled at least 75% of their ICS prescriptions atleast 80%of the time.38,42,49,51,56 Fivestudies (28%)40,43,47,48,52

measured adherence according to scores on validated self-report questionnaires such as the Morisky Questionnaire andthe Adherence Scale, which have good psychometric proper-ties.57,58 One study45 measured adherence by conducting anaudit of general practitioner records,48 three used an elec-tronic monitoring device,41,44,54 three weighed inhalercanisters,18,39,55 and one used a dose counter.50 Most studies(i.e., 11/18, 61%) included a single measure ofadherence,18,38e40,42,44e46,49,50,52 and seven included twomeasures.41,43,47,48,51,55,56

Efficacy of interventions

Among the 14 studies (78%)38,40e48,50e52,54e56 for whicha standardized ES for adherence outcomes could be calcu-lated (see Fig. 2), the efficacy of the interventions accordingto the number of CCM components they included wasassessed. Interestingly, of the thirteen studies that evalu-ated the efficacy of one CCM component (teaching self-management skills38,40e47,56 or decision support40) onlythree (23%)41,42,47 were associated with significant ES’s(0.52e0.57). Of the five studies incorporating two CCMcomponents, i.e., teaching self-management skills anddelivery system design48,50,52/or decision support,38,51 two(40%) had a large ES’s (0.60 and 0.62)48,52 and the remainderhad medium ES’s (0.43 for both)50,51 though one had noeffect.38 Finally, two54,55 of the three (66%)55 studies whereusual care was compared with a program including four CCMcomponents were associated with the highest (i.e. verylarge) ES’s (0.91 and 1.12). When we combined the ES’s ofstudies according to the number of CCM componentsincluded in their interventions (i.e., one, two, four), weobserved a pooled ES of 0.29 (95%CI, 0.16e0.42), 0.53(0.40e0.66) and 0.83 (0.69e0.98) respectively.

Discussion

Main results

The purpose of this reviewwas to evaluatewhether the use ofCCM-based intervention components resulted in greater ICSadherence outcomes among adult asthma patients partici-pating in RCT’s designed to directly or indirectly improve ICSadherence. As expected, the inclusion of a greater number ofCCM components within interventions was associated withstronger effects on ICS adherence outcomes. Interventionsfeaturingone, twoand fourCCMcomponentshad respectivelymedium, large, and very large effect sizes. These findingsprovide support for the use of the CCM as a framework for thedesign and implementation of interventions that improveadherence outcomes among adult asthmatics.

Importance of using a theoretical framework forintervention design

Despite the fact that heterogeneity exists in the contentand methodological features (e.g., adherence measureused; length of follow-up) of the interventions, we notedthat trials incorporating a greater number of CCM

components had a greater effect on ICS adherence. Thisraises questions about what components of the CCM areresponsible for the efficacy seen in asthmatic patients. It’sdifficult to respond with certainty insofar as few data existhighlighting the causal processes that underlie theprograms’ efficacy. However, we could speculate on howand why these findings were observed. To increase thereadibility of the paragraph, it may be useful to createa subsection here. A recent Dutch conceptual frameworkderived from Program Theory59 offers an interesting modelfor understanding how complicated interventions withinthe context of chronic disease management affectoutcomes of care. The authors delineate mechanisms ofchange according to the target level or focus of the inter-vention, i.e., whether it is patient-centred, health-care-centred, and/or organisational-centred. According to thismodel,60 programs targeting healthcare providers is crucialfor changing the partnership dynamic between patients andproviders, and ultimately, patient behavior change. Accessto decision support should affect provider behavioralintention, which in turn, influences provider behavior andmay lead to improved patient health outcomes. Wheninformed, patients take an active role in managing theirhealth, and providers feel prepared and supported whenthey have access to appropriate resources. This may rendertheir mutual interactions more productive. Therefore,when both patients and providers are targeted, collabora-tive management is more likely, leading to a sustainedworking relationship and mutual understanding of roles andresponsibilities. However, this synergy is possible only if it issupported by a third element, the organisational design ofthe healthcare system (i.e., delivery system design,healthcare organization and community). Thus, accordingto this model, three simultaneous causal pathways are allrequired for a behavioral intervention to succeed, as theyall appear to be essential for changing patient healthbehaviours (such as ICS adherence). The results of thisreview appear, in part, to support this framework. Furtherresearch is needed to test this kind of model and refine ourunderstanding of the mechanisms explaining the synergisticeffects of such three-tiered interventions.

Importance of using an appropriate interventiondevelopment process

Our ability to improve health-related behaviors, such as ICSadherence, is dependent on increasing our understandingof the fundamental bases of human behaviour and trans-lating that knowledge into well-defined, effective inter-ventions.26,61 In drug development, research on thedevelopment, testing and refining of interventions islabelled “bench to bedside” or Translation I research.Translational I research begins with basic biologicalresearch to identify mechanisms and intervention targets,which then proceeds to small-scale human trials to assesssafety and optimal dosing, and often includes pilot studiesto assess feasibility and estimates of effect size. Theseearly phases of drug development are followed by larger-scale, Phase III clinical trials that test the efficacy andeffectiveness of the treatment on the health outcomes ofinterest in clinical practice and community settings. This

Page 6

Table 2 Summary of reviews.

Study Location Study

design

Study

quality

Mean

age,

y (I/C)

% Male

(I/C)

Interventions Duration

individual /

Group

Length of

follow-up

Adherence

measures

ES (95%CI) for

adherence

No. Enrolled or

randomized

No.

Completed

I C I C

Self-Management (Education or Behavioural Support)Cote et al,200138

Canada RCTModerate

35/38 33/46 I: SM (Behaviouralsupport)C: UC

NA Individual 12 mo Prescription forsteroid inhaler (self-reported)

None 0.25(�0.24e0.74)

NA 45 30 35

Schafferand Tian,200456

UnitedStates

RCTModerate

NA NA I: SM [Patient education:(a)audiotape(b) or booklet(c) or

audiotape þ booklet]

C: UC

30 min�1 hIndividual

6 mo 1) % of Prescribeddoses (refillaudit)

2) % of Misseddoses (self-reported)

1) (a) None, 0.16(�0.66e0.99)(b) None, 0.81(�0.01e1.62)(c) None, 0.80(�0.03e1.64)

2) NA

101211

13 NA NA

Self-Management (Education D Behavioural Support [or Motivational])Berg et al,199741

UnitedStates

RCTModerate

47/52 32/38 I: SM [Patienteducation þ Behaviouralsupport (action plan basedon peak flow monitoring)]C: UC

12 h Individual 1 wk (1) % of prescribedICS doses taken(electronicmonitoring)

(2) Frequency ofinhaler use(diary e self-reported)

(1) Large 0.57(0.03e1.12)

(2) NA

31 24 30 24

Couturaudet al,200243

France RCTModerate

38/38 31/33 I: SM [Patienteducation þ Behaviouralsupport (action plan basedon peak flow andsymptoms monitoring)]C: UC

2 h 30e5 hIndividual

12 mo (1) % of days oralcorticosteroiduse (self-report)

(2) Self-reportedadherence toinhaler (all 6items eMoriskyscale)

(1) None 0.15(�0.39e0.68)

(2) NA

36 36 26 28

Cote et al,199739

Canada RCTModerate

37/36 44/30 I: SM [Patient education(1 h one-to-one session,book) þ Behaviouralsupport (action plan basedon peak flow monitoring)]C: UC

z3 h Individual 12 mo > 60% of prescribeddoses taken (weight

of used canisters)

NA* NA NA 50 54

Cote et al,199739

Canada RCTModerate

39/36 33/30 I: SM [Patient education(1 h one-to-one session,book) þ Behavioural

z3 h Individual 12 mo > 60% of Prescribeddoses taken (weight

of used canisters)

NA* NA NA 45 54

1216G.Moulle

cetal.

Page 7

Support (action plan basedon symptoms)]C: UC

Gallefossand Bakke,199942

Norway RCT Good 41/44 38/21 I: SM [Patient education(brochure, 40e80 min one-to-one session, 2*2 h groupsession) þ Behaviouralsupport (action plan basedon peak flow andsymptoms monitoring)]C: UC

z5 h Individual& Group

12 mo >75% of prescribeddoses of inhaledsteroid (refill audit)

Large, 0.52(0.05e1.00)

39 39 32 39

Jansonet al,200944

UnitedStates

RCTModerate

37/40 47/46 I: SM [Patient education(3*30 min individualsession) þ (action planbased on peak flow andsymptoms monitoring)]C: UC

1 h 30 Individual 6 mo % of prescribed ICSdoses taken(electronicmonitoring)

None, 0.09(�0.34e0.52)

45 39 35 30

Levy et al,200045

UK RCT Good 43/40 33/43 I: SM [Patient education(3*30 min individualsessions) þ Behaviouralsupport (action plan basedon peak flow andsymptoms monitoring)]C: UC

2 h Individual 6 mo Increased use ofinhalers for asthmaattacks (self-reported):(1) for mild

attacks;(2) for severe

attacks

(1) None, 0.51(�0.17e1.23)

(2) Large, 0.66(0.30e0.98)

103 108 86 95

Morice andWrench,200146

UK RCTModerate

NA 38/33 I: SM [Patient education(3*30 min individualsessions þ booklet þ ) þBehavioural support(action plan based on peakflow and symptomsmonitoring)]C: UC

z1 h30 Individual 6 mo Frequency ofinhaler use (self-reported)

None 0.44(�0.05e0.93)

40 40 35 30

Self-Management (Education D Behavioural D Motivational support)Put et al,200347

Belgium RCTModerate

43/48 58/38 I: SM [Patient education(work book, home workassignments) þBehavioural support (self-monitoring, stimuluscontrol, response control)þ Motivational support(cognitive restructuring)]C: UC

6 h Individual 6 mo Adherence Scale (1)self-report, 3 mo(2) self-report,

6 mo

(1) Large 0.75(0.23e1.02)

(2) Large 0.52(0.01e1.03)

13 12 12 11

(continued on next page)

Improve

adherence

toICSin

adult

asth

matics

1217

Page 8

Table 2 (continued)

Study Location Study

design

Study

quality

Mean

age,

y (I/C)

% Male

(I/C)

Interventions Duration

individual /

Group

Length of

follow-up

Adherence

measures

ES (95%CI) for

adherence

No. Enrolled or

randomized

No.

Completed

I C I C

Self-Management D Decision Support

Cote et al,200138

Canada RCTModerate

34/38 33/46 I: SM [Patienteducation þ Behaviouralsupport (action plan basedon peak flowmonitoring)] þ DS

(interaction providers/patients based on theoryof Bandura)C: UC

z2 h 12 mo Prescription forsteroid inhaler (self-reported)

None 0.44(�0.04 to0.92)

NA 45 33 35

Mehuys et al.,200851

Belgium RCT Good 36/35 46/48 I: SM [Patienteducation þ Behaviouralsupport (peak flowmonitoring)] þ DS

(pharmacists education onGINA guidelines)C: UC

NA 6 mo (1) % of prescribeddoses of inhaledsteroid (refillaudit e phar-

macy claims)(2) Prescription for

steroid inhaler(self-reported)

(1) Medium 0.43(0.10e0.75)

(2) None 0.27(�0.05 to0.60)

107 94 80 70

Self-Management D Delivery system Design

Bailey et al,199048

UnitedStates

RCT Good NA 39/29 I: SM [Patient education(1 h one-to-onecounselling session, workbook) þ Behaviouralsupport (peak flowmeter þ 2 groupmeetings þ 3 follow-upletters)] þ DSD

[structured telephonesupport (2 calls)]C: UC

z2 h30 Individual& Group

12 mo (1) Self-reportedadherence toinhaler (�5 of 6items eMoriskyscale)

(2) Self-reportedadherence toinhaler(all 6 items eMorisky scale)

(3) Adherence toinhaler (ratedby staff)

(1) Large, 0.72(0.45e0.99);

(2) Large 0.62(0.35e0.88)

(3) Medium, 0.60(0.33e0.87)

132 135 124 101

Bailey et al,199940

UnitedStates

RCT Good NA 30/30 I: SM [Patient education(1 h one-to-onecounselling session, workbook) þ Behaviouralsupport (peak flowmeter þ 2 group

z2 h30Individual& Group

24 mo Self-reportedadherence toinhaler (all 6 itemseMorisky scale)

(1) (12 mo)None, 0.05(�0.27e0.36)

78 78 NA NA

1218G.Moulle

cetal.

Page 9

meetings þ 3 follow-upletters)] þ DSD

[structured telephonesupport (2 calls)]C: SM [Patient education(pamphlets)]

(2) (24 mo):Medium, 0.40(0.09e0.72)

Bailey et al,199940

UnitedStates

RCT Good NA 32/30 I: SM [Patient education(15e20 min one-to-onecounselling session,shortened workbook) þ Behaviouralsupport (action plan basedon peak flowmonitoring þ 1 follow-upletter)] þ DSD [Structuredtelephone support (1 call)]C: SM [Patient education(pamphlets)]

z30 minIndividual

24 mo (1) Self-reportedadherence toinhaler (all 6items eMoriskyscale)

(1) (12 mo)None 0.00(�0.32to 0.32)

(2) (24 mo):None, 0.15(�0.16 to0.47)

76 78 NA NA

Chatkin et al,200650

Brazil RCTModerate

43/44 26/29 I: SM (Patienteducation) þ DSD

[Structured telephonesupport (6 calls)]C: UC

1 h Individual 3 mo % of prescribeddoses taken (diskrecords)

Medium 0.43(0.24e0.72)

140 131 NA NA

DeLarondeet al,200549

UnitedStates

RCTModerate

35/36 43/52 I: SM [Patient education(educational booklet andvideo)] þ DSD [structuredtelephone support (6calls)]C: UC

z1 he1 h30Individual

12 mo Ratio of number ofdispensed anti-inflammatory drugson (dispensed b2agonist þ anti-inflammatory drugs)(Pharmacy claims

data)

NA* 67 67 67 67

Windsor et al,199052

UnitedStates

RCTModerate

50/49 39/29 I: SM [Patient education(30 min one-to-onesession) þ Behaviouralsupport (action plan basedon peak flowmonitoring þ 1 h groupsession)] þ DSD

[Structured telephonesupport (2 calls]C: UC

z2 h Individual& Group

12 mo Self-reportedadherence toinhaler (all items� Morisky scale)

Large 0.60(0.34e0.88)

132 135 124 101

(continued on next page)

Improve

adherence

toICSin

adult

asth

matics

1219

Page 10

Table 2 (continued)

Study Location Study

design

Study

quality

Mean

age,

y (I/C)

% Male

(I/C)

Interventions Duration

individual /

Group

Length of

follow-up

Adherence

measures

ES (95%CI) for

adherence

No. Enrolled or

randomized

No.

Completed

I C I C

Self-Management D Decision Support D Delivery system Design D Clinical Information System

Bender et al,201054

UnitedStates

RCTModerate

40/43 40/32 I: SM (Patienteducation þ Behaviouralsupport) þ DS

(intervention based on thebenefit-risk model) þ DSD

[Structured telephonesupport (2-3 calls þ Freetelephone service] þ CIS

(database with automaticstorage of all callinformation)C: UC

z 30e45 minIndividual

10 wks % inhaler puffsprescribed(electronicmonitoring orweight of used

canisters)

Very large 0.91(0.32e1.49)

25 25 25 25

Wilson et al.,2010a55

UnitedStates

RCT Good 46/45 44/43 I: SM [Patienteducation þ Behaviouralsupport þ Motivationalsupport (e.g., barriers tomedication adherenceelicited andaddressed)] þ DS (Allmanagers trainedseparately þ use of GINAguidelines) þ DSD

[Structured telephonesupport (3calls) þ Scheduling offollow-up appointment þCoordination betweencare providers andpatient’s physicians] þ CIS

(Patient’s chartsdocumented by the casemanagers and available tothe patient’s physician)C: UC

z3 h Individual 24 mo (1) Continuousmedicationacquisitionindex (refillaudit)

(2) ICS canisterequivalent(weight of used

canisters)

(1) (12 mo):Large 0.58(0.38e0.77)

(2) (12 mo): Verylarge 1.12(0.91e1.33)

204 204 182 189

Wilson et al.,2010b55

UnitedStates

RCT Good 47/45 44/43 I: idem than above withoutthe motivational interviewC: UC

z2 h30 24 mo (1) Continuousmedicationacquisition

204 204 180 189

1220G.Moulle

cetal.

Page 11

index(refill

audit)

(2)ICS

canister

equivalent

(weightofused

canisters)

(1)(12mo):

Medium

0.36

(0.16e0.55

)(2)(12mo):

Large0.56

(0.42e0.71

)Wilsonetal.,

2010

c55

United

States

RCTGood46

/47

44/4

4I:(.

)withmotiva

tional

interview

C:(.

)without

motiva

tionalinterview

z3h

24mo

(1)Continuous

medication

acq

uisition

index(refill

audit)

(2)ICS

canister

equivalent

(weightofused

canisters)

(1)(12mo):

Medium

0.22

(0.03e0.42

)(2)(12mo):

Medium

0.25

(0.05to

0.44

)

204

204

182

180

Notes:

RCTZ

RandomisedControlledTrial;

NA

Znotindicatedorco

uld

notbeca

lculatedwithava

ilable

data;IZ

Intervention;SM

Zself-m

anage

ment;

DSZ

decisionsupport;

DSD

ZDelive

rysystem

design

;CIS

ZClinicalinform

ationsystem;CZ

Control;UCZ

usual

care;Studyquality(asperDownsandBlack

’sch

ecklist)Z

Poor:

<25

%oftotalpossible

score,

Moderate:25

e49

%oftotalpossible

score,High:�

50%oftotalpossible

score.

Improve adherence to ICS in adult asthmatics 1221

paradigm is well-accepted within the pharmaceuticalindustry and has produced efficacious therapies. A similarparadigm for behavioral medicine interventions has beenrecently suggested by the National Institutes of Health(NIH) in its innovative approaches to reduce obesity (i.e.,Obesity-Related Behavioural Intervention Trials, programORBIT).61 This model provides a clear framework fortranslating basic behavioural and social science discoveriesinto efficacious behavioural interventions. The currentreview that looks at better characterizing the key ingredi-ents (i.e. including CCM components) of behaviouralinterventions in asthmatic patients follows this essentialfirst step of the intervention development process. Thisfirst step consists of providing a scientific basis for theintervention, notably with a meta-analysis. The next stepwill be to optimize the identified key CCM components,notably in terms of dose/duration, mode of delivery,acceptability, and cultural appropriateness. Only whenthese important milestones are achieved will it be appro-priate to move towards larger “Phase III and IV” trialsassessing efficacy and effectiveness.

Recommendations for future studies

Choice of adherence measure

The studies included in this review used several methodsfor measuring adherence such as patient self-report diaries,validated questionnaires, canister weighting, assessment ofprescription refills, and electronic monitoring. They vary incost, practicability, accuracy, complexity, and objectiv-ity.62,63 Interestingly, in both studies51,55 where twomethods of assessing adherence were used and for whicha standardized ES could be calculated, the more objectivethe measure of adherence (self-report vs. pharmacyclaims51 and pharmacy claims vs. canister weighting55), thegreater the ES’s. Therefore, the method of adherencemeasurement is a potentially important moderator to takeinto account when evaluating interventions that target ICSadherence. More objective measures of adherence seem tobe more sensitive to capture changes following interven-tion, with self-reports being the least sensitive (perhapsdue to reporting bias where patients have been shown tooverestimate their medication intake51,59), and pharmacyrefills and electronic devices (the latter of which providesdetailed records of time and date for every inhaler actua-tion) being the most sensitive. This suggests that futureresearch would benefit from including more objectivemethods of assessing adherence in order to improve diag-nostic accuracy, optimize the likelihood of observingmeaningful differences in intervention effects, and toinform treatment decisions.

Choice of comparison group

The choice of an appropriate comparison group is importantto judge the relative efficacy of an intervention understudy.64 If the aim of a trial is to determine whether a novelintervention is superior to existing clinical practice, then ithas to be compared with those practices e knowing that“usual” and “standard of” care changes over time. This may

Page 12

Figure 2 Forest plot of studies. Notes: SM Z self-management; DS Z decision support; DSD Z delivery system design;CIS Z clinical information system; I2 Z heterogeneity index (0e100%); * Z random effect model.

1222 G. Moullec et al.

be why the two RCT’s by Bailey et al.,40,48 which despiteassessing the efficacy of the same intervention (i.e., self-management education program þ structured telephonefollow-up), had very different ES’s (i.e., 0.62 and 0.05).

Between 1990 when they published their initial study and1999 when they published they second, usual or standardclinical practice had changed and teaching asthma self-management skills became routine. Therefore, in their

Page 13

Improve adherence to ICS in adult asthmatics 1223

second study, they only evaluated the added value of anindividualized structured telephone follow-up (i.e.,considered as only one CCM element: delivery systemdesign) versus an education intervention alone. Thisenhancement of usual care is likely to have explained thedisparate ES’s found for the same intervention across thetwo studies.

Focus on key intervention ingredients

With many behavioral interventions including multiplecomponents, it is difficult to know which are more or lessimportant for changing the outcome of interest. However,Wilson et al.’s55 study highlights the specific contribution ofone ingredient of the CCM for improving ICS adherence:motivational support. This study included two groups; inthe first group, the choice of treatment was actively sharedand negotiated between the patient and the physician; inthe second group, the physician alone selected the treat-ment regimen. The results showed that the shared-decisionmaking strategy was associated with a significantimprovement in ICS adherence compared to the physician-only decision-making strategy (ES Z 0.25; 95%CI,0.05e0.44). This finding from a well-powered RCT (i.e., 200patients per arm) suggests the key role played by motiva-tional support within the context of self-managementinterventions, though more research is needed to confirmthis. Initiating and maintaining medication adherenceinvolves a complex series of behavioral changes. Behaviorchange is most likely to occur when individuals are intrin-sically motivated to engage in that behaviour and feelconfident in their ability to change65; two factors that arerarely considered or targeted in asthma adherence inter-ventions to date. Interventions that focus on improvingintrinsic motivation (which emphasizes engaging ina behaviour because it is consistent with personal values,such as having greater personal control over treatmentregimens), rather than extrinsic motivation (whichemphasized engaging in a behaviour to obtain externalrewards such as money or physician approval), may there-fore demonstrate greater efficacy for the improvement ofadherence outcomes, as demonstrated by Wilson et al. Inthis regard, interventions that emphasize shared decision-making and the use of motivational interviewing tech-niques,47,55 which are designed to increase intrinsic moti-vation and foster self-efficacy, may show the most promisefor the improvement of ICS adherence.

Limitations and strengths of the present review

First, this review may be criticized for the small number ofstudies included, particularly those that were adequatelypowered. A related limitation is the small number of CCMcomponent combinations tested, which limits our ability todetermine which one(s) was (were) most critical to success.However, the two trials that included the most (i.e., four)CCM components demonstrated the most significant effectson improving ICS adherence, suggesting the efficacy ofintervening within the CCM framework. Moreover, ourfindings are well supported by prior meta-analyses28,30

addressing the suitability of CCM implementation to

optimize health benefits in patients with various chronicdiseases including diabetes, congestive heart failure, COPDand depression. Second, the restriction for English languagestudies may have introduced a selection bias. However,during a period where redundant publication is commonpractice, the likelihood that important studies would beunpublished in English is low. Third, we did not assess thepresence of publication bias, though the presence ofnegative studies makes this bias less likely. Fourth, theinterventions provided in the studies reviewed were notalways well described. Therefore, inaccurate assignment ofCCM components to these studies may have beencommitted. Finally, we did not have access to data thatwould have allowed us to assess the optimal dose of CCMcomponents in the study interventions. It is possible thatthe interventions we studied had positive effects becausedoing trials requires energy and commitment to participatein the intervention.28 This commitment on the part ofparticipants may be the active ingredient of initial success,but this is difficult to estimate. Simplified comparisonsbased solely on the presence or absence of CCM compo-nents may mask important differences between studies.28

Further research is needed to refine the understanding ofmechanisms explaining the beneficial effects of multi-levelprograms.

Despite the above limitations, this review also hasa number of important strengths. First, to our knowledge,this is the first systematic review to specifically examine theefficacy of interventions designed to (directly or indirectly)improve ICS adherence among adult asthmatics. Second, byconducting a meta-analysis, we were able to objectivelyquantify the relative magnitude of the intervention effectsacross studies. Finally, presenting pooled effect sizes asa function of the number of CCM components used acrossthe studies may be considered a novel and clinically rele-vant way of summarizing the extant literature, as it providessupport for a doseeresponse relationship between thenumber of CCM components and improvements in ICSadherence among adult asthmatics that provides specificdirection for future intervention work in this area.

Conclusions

In conclusion, the inclusion of a greater number of CCMcomponents within interventions was associated withgreater effects on ICS adherence outcomes. Further, thisreview suggests that interventions that include motiva-tional support (i.e., motivational interviewing) may showthe greatest promise in improving adherence, thoughmore research is needed to confirm this. Finally, researchis needed in the development process of behavioralinterventions to improve adherence behavior amongasthmatics. Over the last two decades, several interven-tional studies have been conducted and publishedwithout any real coherent theoretical of developmentframework, and have “leapfrogged” important stages ofdevelopment. The challenge for future studies will be tobridge the gap between basic behavioral/social scienceresearch, behavioural medicine intervention studies(efficacy trials) and public health/community studies(effectiveness trials).

Page 14

1224 G. Moullec et al.

Acknowledgements

Grant support for this work was provided by a grant from theCanadian Institutes of Health Research (CIHR) (KLL), post-doctoral scholarship from Fonds de la Recherche en Sante duQuebec (FRSQ) (GM), FRSQ Research Scholar awards (SLB andKLL), and a CIHR New Investigator Award (SLB).

Conflict of interest statement

Dr. Moullec has no conflicts of interest to disclose; Ms Gour-Provencal has no conflicts of interest to disclose; Dr Baconhas no conflicts of interest to disclose; Dr Campbell has noconflicts of interest to disclose; Dr Lavoie has served on theadvisory board for Schering-Plough and received sponsor-ship for investigator-generated research from Glax-oSmithKline, lecture fees from GlaxoSmithKline, MerckFrosst, AstraZenaca, Pfizer and Air Liquide, and support foreducational materials from Merck Frosst.

References

1. Ernst P, Fitzgerald JM, Spier S. Canadian asthma consensusconference: summary of recommendations. Can Respir J 1996;3:89e100.

2. World Health Organization. Fact sheet 206: bronchial asthma[cited 2005 July 4]; Available from: http://www.who.int/mediacentre/factsheets/fs206/en/print.html; January 2000.

3. CDC National Center for Health Statistics. Asthma prevalence,health care use and mortality [cited July 4]; Available from:http://www.cdc.gov/nchs/products/pubs/pubd/hestats/asthma/asthma.htm; 2005.

4. Barnes NC. Can guideline-defined asthma control be achieved?:the gaining optimal asthma control study. Am J Respir CritCare Med 2004;170:830e1.

5. GINA Dissemination Committee. Dissemination and imple-mentation of asthma guidelines. Global Initiative For Asthma;2003.

6. Global Initiative for Asthma. Global strategy for asthmamanagement and prevention. NIH Publication No 02-3659 2005.

7. Chapman KR, Rea RM, Boulet LP. Primary care assessment ofasthma control; discordance with guideline recommendations.Am J Respir Crit Care Med 2005;171:A549.

8. Lemiere C, Bai T, Balter M, Bayliff C, Becker A, Boulet L-P,et alon behalf of the Canadian Adult Consensus Group of theCanadian Thoracic Society. Adult asthma consensus guidelinesupdate 2003. Can Respir J 2004;11:9Ae33A.

9. Chapman KR, Boulet LP, FitzGerald MJ, McIvor RA,Zimmerman S. Patient factors associated with suboptimalasthma control in Canada: results from the reality of asthmacontrol study. Am J Respir Crit Care Med 2005;172:A678.

10. Green RH, Brightling CE, Pavord ID, Wardlaw AJ. Managementof asthma in adults: current therapy and future directions.Postgrad Med J 2003;79:259e67.

11. Krishnan JA, Riekert KA, McCoy JV, Stewart DY, Schmidt S,Chanmugam A, Hill P, Rand CS. Corticosteroid use afterhospital discharge among high-risk adults with asthma. Am JRespir Crit Care Med 2004;170:1281e5.

12. Heaney LG, Robinson DS. Severe asthma treatment: need forcharacterising patients. The Lancet 2005;365:974e6.

13. Mintz M. Asthma update: part I. Diagnosis, monitoring, andprevention of disease progression. Am Fam Physician 2004;70:893e8.

14. Rowe BH, Edmonds ML, Spooner CH, Camargo CA. Evidence-based treatments for acute asthma. Respir Care 2001;46:1380.

15. Sin DD, Man J, Sharpe H, Gan WQ, Man SFP. Pharmacologicalmanagement to reduce exacerbations in adults with asthma:a systematic review and meta-analysis. JAMA 2004;292:367e76.

16. Colice GL. Categorizing asthma severity: an overview ofnational guidelines. Clin Med Res 2004;2:155e63.

17. Cerveri I, Locatelli F, Zoia MC, Corsico A, Accordini S, deMarco R. International variations in asthma treatmentcompliance: the results of the European Community Respira-tory Health Survey (ECRHS). Eur Respir J 1999;14:288e94.

18. Bender B, Milgrom H, Rand C. Nonadherence in asthmaticpatients: is there a solution to the problem? Ann AllergyAsthma Immunol 1997;79:177e85.

19. Bender BG, Long A, Parasuraman B, Tran ZV. Factors influ-encing patient decisions about the use of asthma controllermedication. Ann Allergy Asthma Immunol 2007;98:322e8.

20. Wraight JM, Cowan JO, Flannery EM, Town GI, DR T. Adherenceto asthma self-management plans with inhaled corticosteroidand oral prednisone: a descriptive analysis. Respirology 2002;7:133e9.

21. Barr RG, Somers SC, Speizer FE, Camargo Jr CA. Patient factorsand medication guideline adherence among older women withasthma. Arch Intern Med 2002;162:1761e8.

22. Wenzel S. Severe asthma in adults. Am J Respir Crit Care Med2005;172:149e60.

23. Kripalani S, Yao X, Haynes RB. Interventions to enhancemedication adherence in chronic medical conditions:a systematic review. Arch Intern Med 2007;167:540e50.

24. WHO. Adherence to long-term therapies. Geneva: WorldHealth Organization; 2003.

25. Graves KD. Social cognitive theory and cancer patients’ qualityof life: a meta-analysis of psychosocial intervention compo-nents. Health Psychol 2003;22:210e9.

26. Kazdin AE. Mediators and mechanisms of change in psycho-therapy research. Annu Rev Clin Psychol 2007;3:1e27.

27. BodenheimerT,Wagner EH,GrumbachK. Improving primary carefor patients with chronic illness. JAMA 2002;288:1775e9.

28. Tsai AC, Morton SC, Mangione CM, Keeler EB. A meta-analysisof interventions to improve care for chronic illnesses. Am JManag Care 2005;11:478e88.

29. Lemmens KM, Nieboer AP, Huijsman R. A systematic review ofintegrated use of disease-management interventions in asthmaand COPD. Respir Med 2009;103:670e91.

30. Adams SG, Smith PK, Allan PF, Anzueto A, Pugh JA, Cornell JE.Systematic review of the chronic care model in chronicobstructive pulmonary disease prevention and management.Arch Intern Med 2007;167:551e61.

31. Leickly FE, Wade SL, Crain E, Kruszon-Moran D, Wright EC,Evans R. Self-reported adherence, management behavior, andbarriers to care after an emergency department visit by innercity children with asthma. Pediatrics 1998;5:1e8.

32. Conn KM, Halterman JS, Fisher SG, Yoos HL, Chin NP,Szilagyi PG. Parental beliefs about medications and medicationadherence among urban children with asthma. Ambul Pediatr2005;5:306e10.

33. Downs SH, Black N. The feasibility of creating a checklist forthe assessment of the methodological quality both of rando-mised and non-randomised studies of health care interven-tions. J Epidemiol Community Health 1998;52:377e84.

34. Wilson DB, Lipsey MW. Practical meta-analysis. ThousandOaks: Sage Publications; 2001.

35. Walter SD, Yao X. Effect sizes can be calculated for studiesreporting ranges for outcome variables in systematic reviews. JClin Epidemiol 2007;60:849e52.

36. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuringinconsistency in meta-analyses. BMJ 2003;327:557e60.

Page 15

Improve adherence to ICS in adult asthmatics 1225

37. Cohen J. Statistical power analysis for the behavioralsciences. Hillsdale, NJ: Erlbaum; 1988.

38. Cote J, Bowie DM, Robichaud P, Parent J-G, Battisti L, Boulet L-P. Evaluation of two different educational interventions foradult patients consulting with an acute asthma exacerbation.Am J Respir Crit Care Med 2001;163:1415e9.

39. Cote J, Cartier A, Robichaud P, Boutin H, Malo JL, Rouleau M,et al. Influence on asthma morbidity of asthma educationprograms based on self-management plans following treatmentoptimization. Am J Respir Crit Care Med 1997;155:1509e14.

40. Bailey WC, Kohler CL, Richards Jr JM, Windsor RA, Brooks CM,Gerald LB, et al. Asthma self-management: do patienteducation programs always have an impact? Arch Intern Med1999;159:2422e8.

41. Berg J, Dunbar-Jacob J. An evaluation of a self-managementprogram for adults with. Clin Nurs Res 1997;6:225.

42. Gallefoss F, Bakke PS. How does patient education and self-management among asthmatics and patients with chronicobstructive pulmonary disease affect medication? Am J RespirCrit Care Med 1999;160:2000e5.

43. Couturaud F, Proust A, Frachon I, Dewitte JD, Oger E, Quiot JJ,Leroyer C. Education and self-management: a one-yearrandomized trial in stable adult asthmatic patients. J Asthma2002;39:493e500.

44. Janson SL, McGrath KW, Covington JK, Cheng SC, Boushey HA.Individualized asthma self-management improves medicationadherence and markers of asthma control. J Allergy ClinImmunol 2009;123:840e6.

45. Levy ML, Robb M, Allen J, Doherty C, Bland JM, Winter RJ. Arandomized controlled evaluation of specialist nurse educationfollowing accident and emergency department attendance foracute asthma. Respir Med 2000;94:900e8.

46. Morice AH, Wrench C. The role of the asthma nurse in treat-ment compliance and self-management following hospitaladmission. Respir Med 2001;95:851e6.

47. Put C, van den Bergh O, Lemaigre V, Demedts M, V G. Evalu-ation of an individualised asthma programme directed atbehavioural change. Eur Respir J 2003;21:109e15.

48. Bailey WC, Richards Jr JM, Brooks CM, Soong S-j, Windsor RA,Manzella BA. A randomized trial to improve self-managementpractices of adults with asthma. Arch Intern Med 1990;150:1664e8.

49. Delaronde S, Peruccio DL. Improving asthma treatment ina managed care population. Am J Manag Care 2005;11:361e8.

50. Chatkin J, Blanco D, Scaglia N, Wagner M, Fritscher C. Impact ofa low-cost and simple intervention in enhancing treatmentadherence in aBrazilian asthma sample. J Asthma 2006;43:263e6.

51. Mehuys E, Van Bortel L, De Bolle L, Van Tongelen I, Annemans L,Remon JP, et al. Effectiveness of pharmacist intervention forasthma control improvement. Eur Respir J 2008;31:790e9.

52. Windsor RA, Bailey WC, Richards Jr JM, Manzella B, Sen-Jaw S,Brooks M. Evaluation of the efficacy and cost effectiveness of

health education methods to increase medication adheranceamong adults with asthma. Am J Public Health 1990;80:1519e21.

53. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gotzsche PC,Ioannidis JPA, et al. The PRISMA statement for reportingsystematic reviews and meta-analyses of studies that evaluatehealthcare interventions: explanation and elaboration. BMJ2009;151:W65e94.

54. Bender BG, Apter A, Bogen DK, Dickinson P, Fisher L,Wamboldt FS, et al. Test of an interactive voice responseintervention to improve adherence to controller medicationsin adults with asthma. J Am Board Fam Med 2010;23:159e65.

55. Wilson SR, Strub P, Buist AS, Knowles SB, Lavori PW, Lapidus J,et althe Better Outcomes of Asthma Treatment Study Group.Shared treatment decision making improves adherence andoutcomes in poorly controlled asthma. Am J Respir Crit CareMed 2010;181:566e77.

56. Schaffer SD, Tian L. Promoting adherence. Clin Nurs Res 2004;13:69e89.

57. Morisky DE, Green LW, Levine DM. Concurrent and predictivevalidity of a self-reported measure of medication adherence.Med Care 1986;24:67e74.

58. Brooks CM, Richards JM, Kohler CL, Soong SJ, Martin B,Windsor RA, et al. Assessing adherence to asthma medicationand inhaler regimens: a psychometric analysis of adult self-report scales. Med Care 1994;32:298e307.

59. Lemmens KM, Nieboer AP, van Schayck CP, Asin JD, Huijsman R.A model to evaluate quality and effectiveness of diseasemanagement. Qual Saf Health Care 2008;17:447e53.

60. Lemmens KM, Nieboer AP, Rutten-Van Molken MP, vanSchayck CP, Asin JD, Dirven JA, et al. Application of a theo-retical model to evaluate COPD disease management. BMCHealth Serv Res 2010;10:81.

61. Czajkowski SM. NIH launches program to develop innovativeapproaches to combat obesity. [cited 12 October 2009].Available from: http://www.nihorbit.org.

62. Hansen RA, Kim MM, Song L, Tu W, Wu J, Murray MD.Comparison of methods to assess medication adherenceand classify nonadherence. Ann Pharmacother 2009;43:413e22.

63. Daniels T, Goodacre L, Sutton C, Pollard K, Conway S,Peckham D. Accurate assessment of adherence: self andclinician report versus electronic monitoring of nebulizers.Chest 2011;140:425e32.

64. Freedland KE, Mohr DC, Davidson KW, Schwartz JE. Usual andunusual care: existing practice control groups in randomizedcontrolled trials of behavioral interventions. Psychosom Med2011;73:323e35.

65. Rollnick S, Miller WR, Butler CC. Motivational interviewing inhealth care: helping patients change behavior. New York, NY:Gilford Press; 2008.