UBC Faculty Research and Publications

Controlled trial of the impact of a BC adult mental health practice support program (AMHPSP) on primary… Lauria-Horner, Bianca; Beaulieu, Tara; Knaak, Stephanie; Weinerman, Rivian; Campbell, Helen; Patten, Scott Nov 28, 2018

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-12875_2018_Article_862.pdf [ 1.13MB ]
JSON: 52383-1.0374921.json
JSON-LD: 52383-1.0374921-ld.json
RDF/XML (Pretty): 52383-1.0374921-rdf.xml
RDF/JSON: 52383-1.0374921-rdf.json
Turtle: 52383-1.0374921-turtle.txt
N-Triples: 52383-1.0374921-rdf-ntriples.txt
Original Record: 52383-1.0374921-source.json
Full Text

Full Text

RESEARCH ARTICLE Open AccessControlled trial of the impact of a BC adultmental health practice support program(AMHPSP) on primary health careprofessionals’ management of depressionBianca Lauria-Horner1* , Tara Beaulieu1, Stephanie Knaak2,5, Rivian Weinerman3, Helen Campbell4and Scott Patten5AbstractBackground: Depression affects over 400 million people globally. The majority are seen in primary care. Barriers inproviding adequate care are not solely related to physicians’ knowledge/skills deficits, but also time constraints, lackof confidence/avoidance, which need to be addressed in mental health-care redesign. We hypothesized that familyphysician (FP) training in the Adult Mental Health Practice Support Program (AMHPSP) would lead to greaterimprovements in patient depressive symptom ratings (a priori primary outcome) compared to treatment as usual.Methods: From October 2013 to May 2015, in a controlled trial 77 FP practices were stratified on the total numberof physicians/practice as well as urban/rural setting, and randomized to the British Columbia AMHPSP a multi-component contact-based training to enhance FPs’ comfort/skills in treating mild-moderate depression(intervention), or no training (control) by an investigator not operationally involved in the trial. FPs with a validlicense to practice in NS were eligible. FPs from both groups were asked to identify 3–4 consecutive patients > 18years old, diagnosis of depression, Patient Health Questionnaire (PHQ-9) score ≥ 10, able to read English, intactcognitive functioning. Exclusion criteria: antidepressants within 5 weeks and psychotherapy within 3 months ofenrollment, and clinically judged urgent/emergent medical/psychiatric condition. Patients were assigned to thesame arm as their physician. Thirty-six practices recruited patients (intervention n = 23; control n = 13). The studywas prematurely terminated at 6 months of enrollment start-date due to concomitant primary health-caretransformation by health-system leaders which resulted in increased in-office demands, and recruitment failure. Weused the PHQ-9 to assess between-group differences at baseline, 1, 2, 3, and 6 months follow-up. Outcomecollectors and assessors were blind to group assignment.Results: One hundred-and-twenty-nine patients (intervention n = 72; control n = 57) were analysed. A significantimprovement in depression scores among intervention group patients emerged between 3 and 6 months, time bytreatment interaction, likelihood ratio test (LR) chi2(3) = 7.96, p = .047.Conclusions: This novel skill-based program shows promise in translating increased FP comfort and skills managingdepressed patients into improved patient clinical outcomes even in absence of mental health specialists availability.Trial registration: #NCT01975948.Keywords: Primary care, Skill-based training, Depressive disorders, Mental disorders, Patient-centered outcomes* Correspondence: Bianca.Horner@dal.ca1Department of Psychiatry, Dalhousie University, Halifax, Nova Scotia, CanadaFull list of author information is available at the end of the article© The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Lauria-Horner et al. BMC Family Practice          (2018) 19:183 https://doi.org/10.1186/s12875-018-0862-yBackgroundThe World Health Organization (WHO) recognizesmental illness as becoming the number one cause ofyears-lived with disability worldwide by 2020 [1–3].Depression is one of the most prevalent and costly con-ditions, affecting over 400 million people globally [4–6].Concurrent with other physical and psychiatricconditions, there is higher morbidity and cost to thehealthcare system [7]. These facts underscore the needfor evidence-based strategies that promote early recogni-tion and treatment, thereby improving patient outcomes[1, 6]. Thus, integration of mental health in primary careis ideal and has been an area of focus in mental health-care redesign that strengthens patient-centered,evidence-based, sustainable care. Family physicians (FPs)see over 85% of these patients, and the majority can behandled early and effectively in this setting [8–15]. Mostpatients experience less stigma and increased comfortsharing problems with their doctor with whom they haveestablished trust [16–18]. However, even when the diag-nosis is made, less than 20% receive adequate treatment[19]. In one study conducted in 21 countries, respon-dents who met DSM-IV criteria for major depressive dis-order (MDD) within 12 months before the interview,only 16.5% received minimally adequate treatment as de-fined by evidenced-based guidelines [i.e. receiving eitherpharmacotherapy (for a minimum of 1 month, plus 4visits with any type of medical doctor) or psychotherapy(for a minimum of 8 visits with any professional includ-ing religious or spiritual advisor, social worker orcounsellor)]. Other studies have shown that althoughFPs commonly prescribe antidepressants, [20] in mild/moderate cases antidepressants are not necessarily asso-ciated with improved long-term outcomes, [21–24].Studies suggest that cognitive behaviour therapy (CBT)has an enduring effect with lower rates of relapse, thatmany patients prefer non-drug options, [25] and whereclinically appropriate, patient choice of evidence-basedoptions improves outcomes [26].The extent to which training programs lead to prac-tice changes or improved clinical outcomes remainsquestionable [27–31]. Barriers in providing adequatecare are not solely related to physicians’ knowledgedeficits, but as a result of complex interdependentfactors. There are physician factors, e.g., in-office timeconstraints, a large number of domains to which theyextend care, mental and physical health conditionsbeing intertwined and thereby confusing symptoms;systems support factors, e.g., lack of mental healthspecialists, ineffective interdisciplinary teams, funding,legislation [32, 33]; and patient factors, e.g., stigma asa barrier to help-seeking, patients lack of disclosuresensing physicians’ time pressures, costly uninsuredservices [10, 31, 34, 35].Training programs with top-down and bottom-upapproaches such as the British Columbia (BC) AdultMental Health Practice Support Program (AMHPSP)show the most promise as the program expands well be-yond a simple education program or workshop [36, 37].Physicians are trained to manage mild-moderate depres-sion and anxiety disorders on their own within officetime constraints by coaching patients through supportedCBT-based self-management strategies. They flexibly usetools and strategies with or without antidepressants,reflecting a “real world” scenario, [38–40] and the pro-gram is based on the quality improvement framework ofplan-do-study-act shown to effectively affect change[41–48] by offering FP opportunities to practice newskills immediately after the training sessions within thescope of their practice. A practice support coordinatorprovides on-site, in-practice support during actionperiods to help implement and sustain these changes[31, 49–51]. Patients are engaged through guidedself-management strategies, which is key for effectivechronic illness care [52–57]. Finally, physicians are com-pensated to attend training. In a realistic context wherephysicians are responsible on their own to manage thesepatients, or while patients wait to be seen by specialtyservices, we —were looking to evaluate—in a real-worldenvironment—if training in the AMHPSP improves pa-tient clinical outcomes. The training would provide phy-sicians another feasible non-drug treatment option intheir armamentarium.While qualitative evaluations of the AMHPSP con-ducted in BC report a positive impact on several keyoutcomes [37], the program has not been rigorously eval-uated through a controlled trial The NS Department ofHealth & Wellness and Mental Health Commission ofCanada therefore sponsored a trial to evaluate itseffectiveness.The first published part of our study shows a significantimprovement (diminishment) in intervention-group phy-sicians’ preference for social distance, and significant in-creases in perceived confidence and comfort managingmental illness [58]. In this part of the study, we sought todetermine whether training FPs in the AMHPSP wouldlead to greater improvements in patient depression scores(primary outcome in this part of the study) compared tothe control group. This paper also focuses on patients’ sat-isfaction with care received, and physicians’ antidepres-sants prescribing. A third part of the study consisting of ahealth economic analysis is underway.We chose a cluster RCT design (randomized practices)as the impact of intervention might bring about practicepattern changes. In the component of the study lookingat patient outcomes we planned an individual-level ana-lysis of depressive symptoms, accounting for the cluster-ing of patients within practices, since this seemed moreLauria-Horner et al. BMC Family Practice          (2018) 19:183 Page 2 of 12clinically salient than practice-level outcomes in this partof the analysis. An implication of this decision, however, isthat the full benefits of randomization are lost since theunit of analysis differs from the unit of randomization.MethodsAim, study design and settingThe aim of this study was to evaluate whether trainingfamily physicians (FPs) in the AMHPSP would lead togreater improvements in patients’ depression scores (pri-mary outcome) compared to the control group.This was a multicentre two-parallel group, controlledtrial in which practices were randomly allocated to theintervention arm (INT) or control arm and allocation wasmasked in the outcome assessment. Physicians in prac-tices assigned to the intervention group attended theAMHPSP training whereas physicians assigned to thecontrol group continued with treatment as usual (TAU).Participant eligibilityPhysician practicesSeventy-seven NS community-based family practicesidentified through associations, presentations, and pro-motional letters. Interested physicians provided writteninformed consent and enrolled between October 2013and January 2014 (111 FP). Physicians actively seeing pa-tients remunerated by any method (fee-for-service/alter-native funding program etc.) with a NS practice licensewere eligible to participate.Eligible patients were identified by their physician. In-clusion criteria included over 18 years of age, diagnosis ofdepression, PHQ-9 score ≥ 10, ability to read/speakEnglish (grade 6 level), intact cognitive functioning (phys-ician judgment). Exclusion criteria included active treat-ment with antidepressants within 5 weeks andpsychotherapy within 3months of enrollment, urgent/emergent medical or psychiatric condition (physicianjudgment). Patients were enrolled between June 2014 andMay 2015, with the last follow-up visit in November 2015.ProcedurePractice recruitment, randomization and maskingSeventy-seven practices (111 community-based FP) wereenrolled. Practice allocation was concealed at clusterlevel through a unique practice number (1–77) assignedby the principal Investigator (PI): B.L-H. Practices werestratified on number of physicians/practice, as well asurban and rural setting before randomization. An inves-tigator (SP) not involved in trial operations assigned therandomization sequence. STATA, version 12 (CollegeStation, TX, 2012) was used to generate therandomization sequence from a binomial distribution—probability of success of 0.5.Patients and physicians’—assigned unique identifiercodes—were linked to their corresponding practicenumber. The PI kept a master file containing participantnames with ID codes in a secure location. Study data-bases contained de-identified information. We could notconceal arm allocation from physicians for obvious rea-sons. However, the RC (data collector) and outcome as-sessors (independent researchers) were blinded to groupassignment. In addition—although we could not guaran-tee patient blinding—physicians were asked not to dis-close group allocation to their patients.Practices ranged from 1 to 6 physicians (INT), and 1–8physicians (Control). Thirty-nine practices were allocatedto the intervention group, and 38 to the control group.Nine percent of targeted practices withdrew prior to inter-vention delivery. The main reason; lack of time for studyspecific requirements. Only 23 of 36 (64%) of interventionand 13 of 34 (38%) of control practices recruited patientsfor the same reason (Fig. 1). Physician baseline character-istics such as age, gender, years of practice, pattern ofwork (full-time, part-time etc.), practice size, number ofunique annual patients were collected.InterventionBetween February–June 2014, physicians in practicesassigned to the intervention group attended threehalf-day interactive workshops delivered by the BC team,the Adult Mental Health Practice Support Program(AMHPSP). The AMHPSP consists of three half-dayinteractive workshops delivered by the BC team, inter-spaced by 5–6 week action periods during which physi-cians practiced what they learned. The programintroduces an organized approach that takes learnersthrough a problem and strength-based assessment to thedevelopment of an action plan. Key components in-cluded 3 supported self-management strategies for mildto moderate depression and anxiety meant to help shiftresponsibility from primary care provider to a shared re-sponsibility with the patients engaging them in illnessmanagement and their recovery; the Cognitive Behav-ioral Interpersonal Skills workbook incorporating a com-prehensive mental illness screening tool and patient userfriendly supported self-management handouts based onevidence-based cognitive behavioural therapy principles;the Canadian Mental Health Association (CMHA)Bounce Back program, a telephone guided CBT-basedservice; and an Antidepressant Skills Workbook [59].Contact-based education occurred through first-voiceadvocates sharing their journey [60]. Office staff mem-bers attended the training and Mental Health First Aid[61] to increase their comfort caring for patients. Localpsychiatrists and allied health professionals (counsellors,psychologists, clinicians etc.) attended training solely tofamiliarize themselves with the program. During actionLauria-Horner et al. BMC Family Practice          (2018) 19:183 Page 3 of 12periods, a coordinator offered on-site guidance on prac-tical office redesign to enhance the implementation ofnew learnings and tools, and shares experiences,challenges, and recommendations with stakeholders.Physicians had flexibility to use tools and strategies asclinically judged. Intervention group FPs received alearning stipend (CAN$3274.20) to participate in work-shops and action periods but remunerated through usualmethods to manage their patients. The control grouppractices received the training November 2015 (at studyend) including the learning stipend.Patient recruitmentOnce the intervention group completed the training(June 2014), all FPs from both intervention and controlgroups were asked to identify 3–4 consecutive patientswith a clinical diagnosis of depression, and a PatientHealth Questionnaire PHQ-9 score ≥ 10. To minimizedamaging the doctor/patient relationship (patients refus-ing or feeling-pressured to participate [62]), willing pa-tients’ were contacted by the research coordinator (RC)who described the study, administered and obtainedwritten consent, and assessed study eligibility. Eligiblepatients were assigned to the same arm as their phys-ician. Once enrolled, patients’ own FP continued tomanage the depression. A total of 169 patients were eli-gible to participate. Forty of 169 (24%) declined or didnot return calls to complete baseline data. We enrolled129 patients (intervention n = 72, control n = 57) all in-cluded in the analysis (Fig. 1).Outcome measuresFor depression scores (primary outcome), we used thePHQ-9. The PHQ-9 covers the nine DSM-5symptom-based criteria for major depressive disorder.Total score range: 0–27; 0–4 not depressed; 5–9 mild;10–14 moderate; 15–19 moderate-severe; 20–27 severe[63–66]. (Score ≥ 10; sensitivity 88%, specificity; 88%;positive likelihood ratio, 7.1). During scale development,criterion validity was assessed against the StructuredClinical Interview (SCID) [66, 67].Fig. 1 Participant Flow Diagram: Allocations, AttritionLauria-Horner et al. BMC Family Practice          (2018) 19:183 Page 4 of 12Occupational/general functioning (secondary out-comes) were assessed with the Lam EmploymentAbsence and Productivity Scale (LEAPS) and the Shee-han Disability Scale (SDS) respectively. The LEAPS a7-item scale assesses workplace impact of majordepression [68]. Total score range = 0–28; 0–5 = none--minimal impairment; 6–10 = mild; 11–16 = moderate;17–22 = severe; 23–28 = very severe. The SDS ratesdisruption (0–10) in each domain work/school; so-cial/leisure activities; family life/home responsibilities,(Total score range = 0–30; lower scores signify lessdisruption) [69].Physician frequency of antidepressant prescribing, pa-tient satisfaction, and quality of life (exploratory out-comes), were assessed with the Client Service ReceiptInventory, an extensively validated inventory, [70] the Cli-ent Satisfaction Inventory (CSI), and the Medical Out-comes Study Short-Form (SF-36) respectively [71–73].Outcome measures were collected at baseline, 1, 2, 3, 6months. All scales are validated as patient self-report, aswell as telephone-collection methods. Patients couldcomplete questionnaires over the phone, in writing ordigitally online.Early terminationThe study was prematurely terminated in May 2015 forthe following reasons. Patient recruitment completelystopped within 6months of the enrollment start-date dueto concomitant primary care health-care transformationby health-system leaders. This resulted in physicians’ re-ports of requiring more time for each type of appointmentin order to effectively provide quality comprehensive carewithin in-office time constraints, thus reducing availabletime they could allocate to the study. The research team’ssubstantial efforts to re-engage physicians throughevidenced-based recruitment “Best-Practices” had minimalimpact. We therefore choose to report results and ourchallenges as useful information for future study designs,and as an exploratory assessment of the intervention.Ethics statementThe NS Multisite and University of Calgary ResearchEthics Boards approved this study, registration www.clinicaltrials.gov #NCT01975948.Statistical methodsSample sizeWe calculated a sample size of 100 evaluable patients/arm to achieve 80% power to detect between-group dif-ferences in PHQ-9 mean of 2 points, with a significancelevel (α) of .05, 2-sided test, and assumed standard devi-ation of 5 points. A compensatory increase to 166 pa-tients/arm was needed using a design effect of 1.1 (intracluster correlation of 0.05 for patient outcomes based ona study by Murphey et al., [74] and 3 patients/practice onaverage), and the assumption that 66% of patients wouldprovide adequate follow-up data, accounting for attrition.AnalysisBetween-group mean differences of PHQ-9 scores dur-ing follow-up were assessed as a group-by-time inter-action. In order to take advantage of the multiplemeasures of data, we used a multi-level mixed-modelanalysis with patients clustered within practices andPHQ-9 ratings clustered within patients. As the patternof improvement over time was not expected to be linear,time points were represented in the model using indica-tor variables. The baseline assessment was not includedin the outcome assessment except as a covariate (eventhough it also included a PHQ-9 rating) since these rat-ings occurred prior to the intervention. We conducted amodified intention to treat analysis, including all respon-dents who provided at least one follow-up rating. Theeffect of the intervention was measured as series of 3study group by time interactions, using a likelihood ratio(LR) test to test the significance of the treatment by timeinteraction terms. This test produced the p-value usedto assess the statistical significance of observed differ-ences. Occupational, general functioning, quality of life,and client satisfaction were analyzed using the samemodel. Antidepressant use at any time during the 6month study period was treated as a binary variable (Y/N). Data was analysed (SP, SK) using STATA, version 14(College Station, TX, 2015).Since attrition occurred during follow-up, we con-ducted a sensitivity analysis using last observation car-ried forward imputation of post-attrition PHQ-9 ratingsand also conducted an analysis of completers only.ResultsPractice characteristicsOne between-group difference was observed. Physiciansassigned to intervention group were more likely to workin small practices due to one large practice being ran-domly assigned to the control group. Twenty-four of 36practices (67%) who recruited patients wereindividual-practice FPs (Table 1).Patient characteristicsPatients were predominantly female, employed, andpost-secondary educated. There were no clinically mean-ingful between-group differences, except that a higherproportion of control group participants were employedas compared to intervention group participants (Table 2).Also, participants with complete follow-up data tendedto be in smaller practices, had somewhat lower baselinePHQ-9 scores, and were more likely to identify as mar-ried/common-law.Lauria-Horner et al. BMC Family Practice          (2018) 19:183 Page 5 of 12Table 1 Practice CharacteristicsPatient recruitment by practices sizeIntervention (n = 23) Control (n = 13) p Yes (n = 36) No (n = 41) PPractice size1 physician 16 8 24 37 0.112 to 3 physicians 5 4 0.647 9 34 to 8 physicians 2 1 3 1Urban Setting14 9 0.727 23 29 0.627Percentage 39% 35% 64% 71%# patients recruited/ practice1 to 2 11 4 – – –3 to 4 7 5 0.252 – – –5 or more 5 4 – – –# patients with complete f/u data1 to 2 13 6 – – –3 to 4 6 3 0.538 – – –5 or more 4 4 – – –Table 2 Patient Baseline Characteristics: Intervention group, control group, completed all follow-up time points versus those whomissed one or more follow-up time pointsPatient Characteristics Intervention Group Control Group Completed all Follow-Up P^n = 72 n = 57* Yes (n = 62) No (n = 67)**GenderFemale 49 (68.1%) 43 (76.8%) 44 (71.0%) 48 (72.7%) .33, .85Marital statusMarried/common-law 38 (52.8%) 37 (66.1%) 42 (67.7%) 33 (50.0%) .15, .05Not married/single/divorced/ separated/widowed 34 (47.2%) 19 (33.9%) 20 (32.3%) 33 (50.0%)Age18–39 31 (43.1%) 31 (55.4%) 26 (41.9%) 36 (54.5%)40–59 26 (36.1%) 18 (32.1%) 26 (41.9%) 18 (27.3%) .33, .2060+ 15 (20.8%) 7 (12.5%) 10 (16.1%) 12 (18.2%)Employment statusEmployed (full or part time) 38 (52.8%) 43 (78.2%) 43 (69.4%) 38 (58.5%)Unemployed 9 (12.5%) 3 (5.5%) 4 (6.5%) 8(12.3%) .01, .37Retired/student/at-home/other 25 (34.7%) 9 (16.4%) 15 (24.2%) 19 (29.2%)EducationSome elementary or High school 9 (12.5%) 3 (5.4%) 5 (8.1%) 7 (10.6%)High school/some post-secondary 23(31.9%) 21 (37.5%) 18 (29.0%) 26 (39.4%) .40, .32Post-secondary diploma/degree + 40 (55.6%) 32 (75.0%) 39 (62.9%) 33 (50.0%)Mother tongueEnglish 66 (91.7%) 55 (98.2%) 59 (95.2%) 62 (93.9%)Other 6 (8.3%) 1 (1.8%) 3 (4.8%) 4 (6.1%) .14, 1.0*One respondent did not provide demographic data. Percentages based on n = 56. Two respondents did not provide employment status data**One respondent did not provide demographic data. Percentages based on n = 66. Two respondents did not provide employment status data^Fisher’s exact test was usedFirst p value pertains to INT/control comparison, the second for completed vs not comparisonLauria-Horner et al. BMC Family Practice          (2018) 19:183 Page 6 of 12Primary outcomePHQ-9 scores diminished in both groups (Fig. 2). Thetreatment group, however, had a progressive diminishmentin mean PHQ-9 scores, whereas the control group did notcontinue to decrease after the 3-month follow-up visit.The multi-level mixed model results were reflective of asignificant treatment effect (LR for treatment group bytime interactions, chi2(3) = 7.96, p = .047), which was es-sentially unchanged after adjusting for baseline depressivesymptoms (LR Chi2(3) = 9.04, p = 0.029). Addition of agecategories and sex to the model (the age categories weredepicted in the dataset using indicator variables represent-ing six age groups) did not change the result (LR for treat-ment group by time interactions terms, chi2(3) =10.40,p = 0.006). Consistent with the pattern seen in Fig. 2,which suggests that differences were only evident in thefinal time interval, removal of the first 2 time-by-treatmentgroup interactions did not significantly affect the fit of themodels, e.g. for the model with adjustments for age, sexand baseline depressive symptoms (LR comparing the ori-ginal to the reduced model, chi2(1) = 1.07, p = 0.301), indi-cating no significant loss of fit with removal of the twointeraction terms.Because of the imbalance in employment observed inTable 2, the analysis was repeated including an indicatorterm for full time employment status. The results wereunchanged: LR chi2(3) = 10.37, Prob > chi2 = 0.0157.Secondary outcomesBetween-group changes on LEAPS and SDS, SF-36, andCSI were not significantly different (See Additional file 1).Fig. 2 Between-group mean differences of PHQ-9 scores, group-by-time interactionLauria-Horner et al. BMC Family Practice          (2018) 19:183 Page 7 of 12Exploratory outcomesAntidepressant use at any time during the 6monthsstudy period was significantly lower in the interventiongroup compared to the control group Fisher’s exact test,p = 0.003. (Control group = 68.4%; Intervention group =41.7%). In a sensitivity analysis using LOCF imputationthere was no evidence of a treatment effect as assessedby the three group by time interactions (LR for the treat-ment group by time interactions, chi2(3) = 5.28, Prob >chi2 = 0.153). However, in the completer analysis, the ef-fect remained significant: LR for the treatment group bytime interactions chi2(3) = 7.96, Prob > chi2 = 0.047.Fifty-one percent of INT patients used program tools/strategies. The top 3 were CBT-based with BounceBackbeing highest (51% of patients who used tools, 26% ofINT patients) (Table 3).DiscussionResults of this study provide tentative evidence of theAMHPSP’s effectiveness in improving patient depressionscores. A large majority of INT patients (71%) reportedusing the program tools and strategies, the most fre-quently used was the BounceBack program (Table 3).Anecdotal FP reports suggested that BounceBack pro-gram offers increased support by shared patient manage-ment. PHQ-9 scores diminished in both groups frombaseline to 3 months due to the episodic nature of de-pression expected to cause regression to the mean, andpossibly as a result of placebo responders through re-peated encounters with the research data collector. Inpsychiatric disorders such as depression, thisnon-specific treatment effect is a well-recognized com-ponent in all psychotherapeutic encounters [75]. Thetreatment group, however, had a progressive diminish-ment in mean PHQ-9 scores, whereas the control groupdid not continue to decrease after the 3-monthfollow-up visit (Fig. 1).Further research is needed in this area as our studydid not achieve its pre-planned sample size, lending tothe possibility of Type II error. Also, the effect observedwas small only evident at a single time interval, between3 and 6months, a finding that may have occurred as aresult of Type I error. There was substantial attrition,and LOCF imputation did not result in preservation ofthe effect. However, the pattern of between-groupchange in mild/moderate depression scores over time issuggestive of a possible treatment effect. Furthermore,these changes occurred despite reduced prescribing ofmedication by FPs. (Table 3) Participants with completefollow-up data had somewhat lower baseline PHQ-9scores, which is consistent with the AMHPSP’s targetpopulation: mild to moderate depression, hence futurestudies should consider including only this group. Futurestudies should also strongly consider using apractice-level unit of analysis since randomization insuch trials must by necessity occur at the practice level.In a context where collaborative care has proven diffi-cult, (e.g., where primary care is delivered in independ-ent practice settings and/or no funding mechanisms forcollaborative care arrangements, we were looking atways of implementing time-efficient evidence-basedstrategies to increase physicians’ comfort and skills, de-creasing anxiety/avoidance treating highly prevalent de-pression conditions. We do not feel there are or willever be enough mental health physicians/clinicians tosupport FPs especially in rural and remote areas. There-fore, responsibility to care for the mentally ill largelyrests with family physicians. We hear time and againthat primary care knowledge gaps contribute tounrecognized and undertreated mental illness. However,training efforts aimed to improve mental health manage-ment and patient outcomes remain questionable [30].Training efforts need to include time efficient tools FPscan feasibly implement in the realistic context of a busypractice.Table 3 Antidepressant Prescribing at 6 months: Intervention and control groups Strategies used: InterventionAntidepressant Use (%) INT CONTROL41.7% 68.4%Specific tools used N % of participants (n = 37) % of all INT participants (n = 72)Referral to BB 19 51.4% 26.4% 0ASW resources 11 29.7% 15.3% 0CBIS resources 9 24.3% 12.5% 0BB DVD provided 6 16.2% 8.3% 0Physician telephone follow-up 3 8.1% 4.2% 0DAI administered 3 8.1% 4.2% 0Total* 51 116% 71% 0*patients used 1 or more tools*Mentioned tools: Diagnostic Assessment Interview (DAI) administered; Antidepressant Skills Workbook (ASW) provided; Referral to Bounce Back (BB); Bounce BackDVD provided; Cognitive Behavioural Interpersonal Skills (CBIS) resources provided; physician follow-up by telephoneLauria-Horner et al. BMC Family Practice          (2018) 19:183 Page 8 of 12The AMHPSP cannot address all barriers, however it isnovel recognizing that FPs need in-practice support that ex-pand well beyond a simple education program or work-shop. It includes many theoretically crucial ingredientsrequired to fill knowledge gaps, through organized trainingthat creates a substantial sustainable change, in the deliveryof patient-centered respectful care, by FPs that fits theirbusy schedule, targets stigma, and supports physicians andpatients through incentives, resources, and tools engagingpatients in recovery efforts. Most patients want to partici-pate in their recovery. Where better to learn the skills thanin a trusted, safe place with their physician, even in settingswhere access to specialty care is difficult or unavailable?To our knowledge, this study is the first RCT of its kindspecifically evaluating a practice support program focusingon mental health as it is meant to be used in a real-worldenvironment. We believe this provides insight on the po-tential impact of the AMHPSP training on patients’ clin-ical outcomes and practical applicability of delivering theprogram in primary care. Furthermore improvement indepression scores occurred despite reduced prescribing ofmedication by FPs which suggests possible cost offsetsdue to reduced prescribing costs (Table 3). Also, prioranalyses indicated reductions in social distance prefer-ences and increased confidence and comfort with thetraining, beneficial effects not captured by the analysis ofpatient-reported outcomes presented here58.LimitationsFirst, this was a controlled trial, but randomization oc-curred at the practice level rather than individual patients.Because the analysis occurred at the level of patients, thebenefits of randomization (equal distribution of confound-ing variables) cannot be assumed to have fully manifestedin the analysis. Indeed, as some practices failed to recruitpatients at all, the patient data analysis may have been in-fluenced by selection bias despite the randomization.Whereas the random assignment would help ensure thatthe practices assigned to each group had similar character-istics, including unmeasured ones, it does not ensure thatthe ones that recruited patients were comparable on suchcharacteristics. The significant difference in PHQ-9 scoreswas observed later in follow-up, at which point somephysician and patient attrition had occurred, and this at-trition may have caused bias. The lack of significant effectsin a sensitivity analysis using LOCF imputation heightensthese concerns. In addition, the there was an imbalance inpatient group sizes (intervention n = 72, Control n = 57)which could have further affected the specific effects ofthe intervention compared to TAU (Fig. 1). However, themixed model analysis methods permits the use of all data(e.g. data points from earlier in follow-up even when lostto follow-up), and may help to control bias that may arisedue to attrition.As noted above, future studies randomizing such inter-ventions by practice should consider the use ofpractice-level outcome measures that would allow theanalysis to derive full benefit from the randomization, alarger sample size and employ effective strategies to pre-vent attrition, especially since the benefits of this type ofintervention may unfold over a time frame of severalmonths. Third, physician participants were volunteersreceiving a learning stipend, therefore may not be repre-sentative of FPs in general, may have a special interest inthe subject matter. Although some aspects of the studycould not be blinded, performance bias was minimizedby data collector and outcome assessors blinding as wellas the use of widely validated scales more specificallythe PHQ-9 for our primary outcome. We took measuresalso to minimize the risk of patient selection bias by ask-ing physicians to identify 3 consecutive patients with adiagnosis of depression.Finally, a longer patient follow-up period may be im-portant to shed light if divergence of depression scoreswould further increase, or reveal a significant differencein quality of life/occupational functioning as these out-comes may take longer to improve. Since the study sam-ple consisted mostly of white, married, middle-agefemales, there should be caution in generalizing resultsbeyond this group. However, the approach is applicableto all prepaid health insurance plans in the US and toCanada’s single payer system [42].In assessing satisfaction, most patients reported highbaseline physician satisfaction ceiling effect. An out-come measure specific to mental healthcare would havebeen a better choice.ConclusionThis study provides preliminary evidence thatwell-designed novel skills-based PSPs that promote inte-gration of mental health into primary care may contrib-ute to mental health strategies in improving mentalhealth care. It also highlights the difficulties in evaluat-ing community-delivered interventions.Additional filesAdditional file 1: Tables: Mean, standard deviation (SD) and number ofnon-missing observations. (PDF 246 kb)Additional file 2: Data access document. (DOCX 12 kb)AbbreviationsAMHPSP: Adult mental health practice support program; BC: BritishColumbia; CBT: cognitive behaviour therapy; CMHA: Canadian mental healthassociation; CSI: Client satisfaction inventory; FP: Family physician;LEAPS: Lam employment absence and productivity scale; LOCF: lastobservation carried forward; LR: likelihood ratio t; NS: Nova Scotia;PHQ-9: Patient health questionnaire-9; PI: Principal investigator; RC: Researchcoordinator; SCID: Structured clinical interview; SDS: Sheehan disability scale;SF-36: Medical outcomes study short-form; TAU: Treatment as usualLauria-Horner et al. BMC Family Practice          (2018) 19:183 Page 9 of 12AcknowledgementsThe authors would like to thank the primary care providers and patients whoparticipated as part of this trial.Upon completion of the project Ms. Beaulieu’s affiliation changed. She heldthe position of Research Coordinator, British Columbia Centre for Excellencein HIV/AIDS.FundingThis project was made possible through funding from the Nova ScotiaDepartment of Health and Wellness and the Mental Health Commission ofCanada. The DOH&W covered the time and expenses of the researchcoordinator/practice support coordinator, participants, psychiatrists, mentalhealth clinicians and first voice individuals. The DOH&W also covered costsassociated with the Bounce Back program, the Mental Health First Aidtraining, and costs related to training facilities, equipment, and materials. TheMHCC covered travel expenses related to meetings as well as the remunerationfor RW, HC and those responsible for delivery of the intervention. As the leadresearch institution, Dalhousie University, Department of Psychiatry contributedin-kind services for the Principal Investigator, BLH, as well as related institutionaloverhead costs. The MHCC and the DOH&W did not influence any aspect ofthe research or publication.Availability of data and materialsThe data are archived on the University of Calgary Web Server at:http://people.ucalgary.ca/~patten/Datasets/deidentified_dataset.dtaThe data can be accessed from within Stata using the following command:use http://people.ucalgary.ca/~patten/Datasets/deidentified_dataset.dta(See Additional file 2)Authors’ contributionsBLH led the design and implementation of the research study, wrote themain paper, and was responsible for the approval of the final versionsubmitted for publication. BLH had full access to all the data in the studyand responsible for the integrity of the data. SK (employed by the MHCC)and SP were responsible for the accuracy of data analysis, and preparation ofevaluation reports. TB was central to ongoing collaboration acrossprofessionals, research team members, and stakeholders, and contributed incritically revising the manuscript for important intellectual content. RW, HCand their team led the design of the Adult Mental Health Practice SupportProgram. All authors contributed substantially to critically revising the paper.All authors read and approved the final manuscript.Ethics approval and consent to participateThe Nova Scotia Multisite and University of Calgary Research Ethics Boardsapproved this study, registration #NCT01975948. Written informed consentwas obtained from all participants.Consent for publicationNot applicable.Competing interestThe authors declare that they have no competing interests.Publisher’s NoteSpringer Nature remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.Author details1Department of Psychiatry, Dalhousie University, Halifax, Nova Scotia, Canada.2Opening Minds Anti-Stigma Initiative, Mental Health Commission of Canada,Ottawa, ON, Canada. 3University of Bristish Columbia, Medical Staff HonoraryStatus Island Health Authority, Victoria, Canada. 4Department of Psychiatry,University of British Columbia, Vancouver, Canada. 5University of Calgary,Calgary, Alberta, Canada.Received: 6 February 2018 Accepted: 5 November 2018References1. Murray CJL, Barber RM, Foreman KJ, Ozgoren AA, Abd-Allah F, Abera SF, etal. Global, regional, and national disability-adjusted life years (DALYs) for 306diseases and injuries and healthy life expectancy (HALE) for 188 countries,1990–2013: quantifying the epidemiological transition. Lancet. 2015;386:2145–91.2. Whiteford HA, Degenhardt L, Rehm J, Baxter AJ, Ferrari AJ, Erskine HE, et al.Global burden of disease attributable to mental and substance usedisorders: findings from the global burden of disease study 2010. Lancet.2013;382:1575–86.3. Depression in Europe [Internet]. 2012 [cited 2018 May 26]. Available from:http://www.euro.who.int/en/health-topics/noncommunicable-diseases/mental-health/news/news/2012/10/depression-in-europe4. Lim K-L, Jacobs P, Ohinmaa A, Schopflocher D, Dewa CS. A new population-based measure of the economic burden of mental illness in Canada.Chronic Dis Can. 2008;28:92–8.5. Kessler RC, Aguilar-Gaxiola S, Alonso J, Chatterji S, Lee S, Ormel J, et al. Theglobal burden of mental disorders: an update from the WHO world mentalhealth (WMH) surveys. Epidemiol Psichiatr Soc. 2009;18:23–33.6. Principaux repères sur la dépression [Internet]. World Health Organ. [cited2018 May 25]. Available from: http://www.who.int/fr/news-room/fact-sheets/detail/depression7. Naylor C. SP0115 the link between long-term conditions and mental health.Ann Rheum Dis. 2013;71:28–9.8. Barr MS. The need to test the patient-centered medical home. JAMA. 2008;300:834–5.9. Iglehart JK. No place like home--testing a new model of care delivery. NEngl J Med. 2008;359:1200–2.10. Unützer J, Park M. Strategies to improve the management of depression inprimary care. Prim Care. 2012;39:415–31.11. Kauye F, Jenkins R, Rahman A. Training primary health care workers inmental health and its impact on diagnoses of common mental disorders inprimary care of a developing country, Malawi: a cluster-randomizedcontrolled trial. Psychol Med. 2014;44:657–66.12. Butler M, Kane RL, McAlpine D, Kathol RG, Fu SS, Hagedorn H, et al.Integration of mental health/substance abuse and primary care. Evid RepTechnol Assess. 2008:1–362.13. Unutzer J, Katon WJ, Fan M-Y, Schoenbaum MC, Lin EHB, Della Penna RD, etal. Long-term cost effects of collaborative care for late-life depression. Am JManag Care. 2008;14:95–100.14. WHO | Integrating mental health into primary care: a global perspective[Internet]. WHO. [cited 2018 May 26]. Available from: http://www.who.int/mental_health/policy/services/integratingmhintoprimarycare/en/15. Mental Health and Addictions Strategy | novascotia.ca [Internet]. [cited 2018May 26]. Available from: https://novascotia.ca/dhw/mental-health/mental-health-addiction-strategy.asp16. Orleans CT, George LK, Houpt JL, Brodie HK. How primary care physicianstreat psychiatric disorders: a national survey of family practitioners. Am JPsychiatry. 1985;142:52–7.17. Azrin ST. Integrated care: high-impact mental health-primary care researchfor patients with multiple comorbidities. Psychiatr Serv Wash DC. 2014;65:406–9.18. Jorm AF, Korten AE, Jacomb PA, Christensen H, Rodgers B, Pollitt P. “Mentalhealth literacy”: a survey of the public’s ability to recognise mental disordersand their beliefs about the effectiveness of treatment. Med J Aust. 1997;166:182–6.19. Mulsant BH, Whyte E, Lenze EJ, Lotrich F, Karp JF, Pollock BG, et al.Achieving long-term optimal outcomes in geriatric depression and anxiety.CNS Spectr. 2003;8:27–34.20. Kendrick T, King F, Albertella L, Smith PW. GP treatment decisions forpatients with depression: an observational study. Br J Gen Pract. 2005;55:280–6.21. Schulberg HC, Block MR, Madonia MJ, Rodriguez E, Scott CP, Lave J.Applicability of clinical pharmacotherapy guidelines for major depression inprimary care settings. Arch Fam Med. 1995;4:106–12.22. Schulberg HC, Block MR, Madonia MJ, Scott CP, Lave JR, Rodriguez E, et al.The “usual care” of major depression in primary care practice. Arch FamMed. 1997;6:334–9.23. Fournier JC, DeRubeis RJ, Hollon SD, Dimidjian S, Amsterdam JD, SheltonRC, et al. Antidepressant drug effects and depression severity: a patient-levelmeta-analysis. JAMA. 2010;303:47–53.24. Kirsch I, Deacon BJ, Huedo-Medina TB, Scoboria A, Moore TJ, Johnson BT.Initial severity and antidepressant benefits: a meta-analysis of datasubmitted to the Food and Drug Administration. PLoS Med. 2008;5:e45.Lauria-Horner et al. BMC Family Practice          (2018) 19:183 Page 10 of 1225. van Schaik DJF, Klijn AFJ, van Hout HPJ, van Marwijk HWJ, Beekman ATF, deHaan M, et al. Patients’ preferences in the treatment of depressive disorderin primary care. Gen Hosp Psychiatry. 2004;26:184–9.26. Sherbourne CD, Wells KB, Duan N, Miranda J, Unützer J, Jaycox L, et al.Long-term effectiveness of disseminating quality improvement fordepression in primary care. Arch Gen Psychiatry. 2001;58:696–703.27. van Os TW, Ormel J, van den Brink RH, Jenner JA, Van der Meer K, TiemensBG, et al. Training primary care physicians improves the management ofdepression. Gen Hosp Psychiatry. 1999;21:168–76.28. Kroenke K, Taylor-Vaisey A, Dietrich AJ, Oxman TE. Interventions to improveprovider diagnosis and treatment of mental disorders in primary care. Acritical review of the literature. Psychosomatics. 2000;41:39–52.29. Kalet AL, Gillespie CC, Schwartz MD, Holmboe ES, Ark TK, Jay M, et al. Newmeasures to establish the evidence base for medical education: identifyingeducationally sensitive patient outcomes. Acad Med J Assoc Am Med Coll.2010;85:844–51.30. Sikorski C, Luppa M, König H-H, van den Bussche H, Riedel-Heller SG. DoesGP training in depression care affect patient outcome? - a systematic reviewand meta-analysis. BMC Health Serv Res. 2012;12:10.31. Crabtree BF, Nutting PA, Miller WL, McDaniel RR, Stange KC, Jaen CR,et al. Primary care practice transformation is hard work: insights froma 15-year developmental program of research. Med Care. 2011;49(Suppl):S28–35.32. Phillips RL, Miller BF, Petterson SM, Teevan B. Better integration of mentalhealth care improves depression screening and treatment in primary care.Am Fam Physician. 2011;84:980.33. Chomienne M-H, Grenier J, Gaboury I, Hogg W, Ritchie P, Farmanova-Haynes E. Family doctors and psychologists working together: doctors’ andpatients’ perspectives. J Eval Clin Pract. 2011;17:282–7.34. Unützer J, Schoenbaum M, Druss BG, Katon WJ. Transforming mental healthcare at the interface with general medicine: report for the presidentscommission. Psychiatr Serv Wash DC. 2006;57:37–47.35. Pereira B, Andrew G, Pednekar S, Kirkwood BR, Patel V. The integration ofthe treatment for common mental disorders in primary care: experiences ofhealth care providers in the MANAS trial in Goa, India. Int J Ment HealthSyst. 2011;5:26.36. Fortney J, Enderle M, McDougall S, Clothier J, Otero J, Altman L, et al.Implementation outcomes of evidence-based quality improvement fordepression in VA community based outpatient clinics. Implement Sci IS. 2012;7:30.37. Weinerman R, Campbell H, Miller M, Stretch J, Kallstrom L, Kadlec H, et al.Improving mental healthcare by primary care physicians in British Columbia.Healthc Q Tor Ont. 2011;14:36–8.38. Greenhalgh T, Robert G, Macfarlane F, Bate P, Kyriakidou O. Diffusion ofinnovations in service organizations: systematic review andrecommendations. Milbank Q. 2004;82:581–629.39. Miller WL, Crabtree BF, McDaniel R, Stange KC. Understanding change inprimary care practice using complexity theory. J Fam Pract. 1998;46:369–76.40. Plsek P. Complexity and the Adoption of Innovation in Health Care. Paperpresented at Accelerating Quality Improvement in Health Care: Strategies toAccelerate the Diffusion of Evidence-Based Innovations. Washington, D.C.:National Institute for Healthcare Management Foundation and NationalCommittee for Quality in Health Care; 2003.41. Ruggles B. The improvement guide: a practical approach to enhancingorganizational performance. Qual Prog Milwaukee. 1998;31:108.42. MacCarthy D, Weinerman R, Kallstrom L, Kadlec H, Hollander MJ, Patten S.Mental health practice and attitudes of family physicians can be changed!Perm J. 2013;17:14–7.43. Alweis R, Greco M, Wasser T, Wenderoth S. An initiative to improveadherence to evidence-based guidelines in the treatment of URIs, sinusitis,and pharyngitis. J Community Hosp Intern Med Perspect [Internet]. 2014[cited 2018 May 25];4. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3937558/44. Ginsburg LR, Lewis S, Zackheim L, Casebeer A. Revisiting interaction inknowledge translation. Implement Sci IS. 2007;2:34.45. Stange KC, Goodwin MA, Zyzanski SJ, Dietrich AJ. Sustainability of apractice-individualized preventive service delivery intervention. Am J PrevMed. 2003;25:296–300.46. Parker LE, Kirchner JE, Bonner LM, Fickel JJ, Ritchie MJ, Simons CE, et al.Creating a quality-improvement dialogue: utilizing knowledge fromfrontline staff, managers, and experts to foster health care qualityimprovement. Qual Health Res. 2009;19:229–42.47. Rubenstein LV, Mittman BS, Yano EM, Mulrow CD. From understandinghealth care provider behavior to improving health care: the QUERIframework for quality improvement. Quality Enhancement ResearchInitiative. Med Care. 2000;38:I129–41.48. Mendel P, Meredith LS, Schoenbaum M, Sherbourne CD, Wells KB.Interventions in organizational and community context: a framework forbuilding evidence on dissemination and implementation in health servicesresearch. Admin Pol Ment Health. 2008;35:21–37.49. Walters ST, Matson SA, Baer JS, Ziedonis DM. Effectiveness of workshoptraining for psychosocial addiction treatments: a systematic review. J SubstAbus Treat. 2005;29:283–93.50. Tamburrino MB, Nagel RW, Lynch DJ. Managing antidepressants in primarycare: physicians’ treatment modifications. Psychol Rep. 2011;108:799–804.51. Davis DA, Thomson MA, Oxman AD, Haynes RB. Changing physicianperformance. A systematic review of the effect of continuing medicaleducation strategies. JAMA. 1995;274:700–5.52. Nicholas MK, Asghari A, Corbett M, Smeets RJEM, Wood BM, Overton S, etal. Is adherence to pain self-management strategies associated withimproved pain, depression and disability in those with disabling chronicpain? Eur J Pain Lond Engl. 2012;16:93–104.53. Turner A, Anderson JK, Wallace LM, Bourne C. An evaluation of a self-management program for patients with long-term conditions. Patient EducCouns. 2015;98:213–9.54. Sevick MA, Trauth JM, Ling BS, Anderson RT, Piatt GA, Kilbourne AM, et al.Patients with complex chronic diseases: perspectives on supporting self-management. J Gen Intern Med. 2007;22(Suppl 3):438–44.55. Theeboom T, Beersma B, Van Vianen A. Does coaching work? A meta-analysis on the effects of coaching on individual level outcomes in anorganizational context. J Posit Psychol. 2013;9(1):1–18.56. Linden A, Butterworth SW, Prochaska JO. Motivational interviewing-based healthcoaching as a chronic care intervention. J Eval Clin Pract. 2010;16:166–74.57. Butterworth S, Linden A, McClay W, Leo MC. Effect of motivationalinterviewing-based health coaching on employees’ physical and mentalhealth status. J Occup Health Psychol. 2006;11:358–65.58. Beaulieu T, Patten S, Knaak S, Weinerman R, Campbell H, Lauria-Horner B.Impact of skill-based approaches in reducing stigma in primary carephysicians: results from a double-blind, parallel-cluster, randomizedcontrolled trial. Can J Psychiatry Rev Can Psychiatr. 2017;62:327.59. Antidepressant Skills Workbook - CARMHA - Centre for Applied Research inMental Health & Addiction - Simon Fraser University [Internet]. [cited 2018May 25]. Available from: https://www.sfu.ca/carmha/publications/antidepressant-skills-workbook.html60. Healthy Minds Cooperative: [Internet]. [cited 2018 May 25]. Available from:https://www.healthyminds.ca/.61. Mental Health First Aid | Mental Health Commission of Canada [Internet].[cited 2018 May 25]. Available from: https://www.mentalhealthcommission.ca/English/resources/mental-health-first-aid62. Hellman S, Hellman DS. Of mice but not men. Problems of the randomizedclinical trial. N Engl J Med. 1991;324:1585–9.63. Mulrow CD, Williams JW, Gerety MB, Ramirez G, Montiel OM, Kerber C. Case-finding instruments for depression in primary care settings. Ann Intern Med.1995;122:913–21.64. Martin A, Rief W, Klaiberg A, Braehler E. Validity of the brief patient healthquestionnaire mood scale (PHQ-9) in the general population. Gen HospPsychiatry. 2006;28:71–7.65. Löwe B, Schenkel I, Carney-Doebbeling C, Göbel C. Responsiveness of thePHQ-9 to psychopharmacological depression treatment. Psychosomatics.2006;47:62–7.66. Kroenke K, Spitzer RL. The PHQ-9: a new depression diagnostic and severitymeasure. Psychiatr Ann. 2002;32:509–15.67. Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depressionseverity measure. J Gen Intern Med. 2001;16:606–13.68. Lam RW, Michalak EE, Yatham LN. A new clinical rating scale for workabsence and productivity: validation in patients with major depressivedisorder. BMC Psychiatry. 2009;9:78.69. Sheehan DV, Harnett-Sheehan K, Raj BA. The measurement of disability. IntClin Psychopharmacol. 1996;11(Suppl 3):89–95.70. Chisholm D, Knapp M, Knudsen H, Amaddeo F, Gaite L, Van Wijngaarden B.Client socio-demographic and service receipt inventory - European Version:Development of an instrument for international research - EPSILON Study 5.Br J Psychiatr. 2000;177:S28–33.Lauria-Horner et al. BMC Family Practice          (2018) 19:183 Page 11 of 1271. Ware JE, Sherbourne CD. The MOS 36-item short-form health survey (SF-36).I. Conceptual framework and item selection. Med Care. 1992;30:473–83.72. McHorney CA, Ware JE, Raczek AE. The MOS 36-item short-form healthsurvey (SF-36): II. Psychometric and clinical tests of validity in measuringphysical and mental health constructs. Med Care. 1993;31:247–63.73. McHorney CA, Ware JE, Lu JF, Sherbourne CD. The MOS 36-item short-formhealth survey (SF-36): III. Tests of data quality, scaling assumptions, andreliability across diverse patient groups. Med Care. 1994;32:40–66.74. Murphy AW, Esterman A, Pilotto LS. Cluster randomized controlled trials inprimary care: an introduction. Eur J Gen Pract. 2006;12:70–3.75. Schoemaker JH, Kilian S, Emsley R, Vingerhoets AJJM, Medical and ClinicalPsychology. Factors associated with placebo response in depression trials.Neurol Psychiatry Brain Res. 2017;30:12–21.Lauria-Horner et al. BMC Family Practice          (2018) 19:183 Page 12 of 12


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items