Selected Conference Proceedings
5th Annual Family Medicine Research Day

Published online: 21 July 2019

Message from Family Medicine Research Day

This special issue features the top 15 abstracts from the 5th Annual Family Medicine Graduate Student Research Symposium held on May 27, 2019. The Research Symposium was jointly hosted by the Family Medicine Graduate Students’ Society and the Department of Family Medicine. The event brought together students, faculty and community members to learn about current research and advances in primary care, community health, medical education and health policy. We received 42 abstract submissions. All abstracts were reviewed by 2 reviewers using a blinded, peer-review process. Scores were carefully reviewed and confirmed by the abstract review committee, which was composed of six graduate students from the Department of Family Medicine at McGill University. Special thanks to the abstract reviewers, presentation judges, organizing committee, the Hirsch Rosenfeld Fund, Post-Graduate and Graduate Students' Society (PGSS), the Faculty of Medicine, the Department of Family Medicine Research Division, and the McGill Alumni Association, without whom this event would not have been possible.

Feasibility of Platform Trials Design for the development of HIV specific Electronic Patient Reported Outcomes: A Scoping Review

Asma Aqib1, Tibor Schuster1, Bertrand Lebouché1
1Department of Family Medicine, McGill University

Background: About 36.9 million People are infected with Human Immune Deficiency Virus (HIV) worldwide. Adherence to Anti-Retroviral Treatment (ART) is a major challenge for HIV care. The I-Score study aims to implement a platform trial (PT) approach for the development of an electronic patient reported outcome (e-PRO) that helps identifying barriers to ART adherence.

Objectives: The scoping review aims to summarize the ethical, pragmatic and technical issues associated with the PT, Adaptive randomizations, and Multi-arm multi-stage designs in clinical settings. The review aims to confirm the feasibility and acceptability of PT for e-PRO development and other e-Health applications.

Methods: Following PRISMA-ScR guidelines, a scoping review is being conducted using two databases (Medline via PubMed and Embase via Scopus) based on pre-defined search strategies. The search included original articles, reviews, letters, as well as commentaries. Studies published from 1947 to 2018 were eligible for inclusion, as were any published empirical studies (qualitative, quantitative, clinical trials) describing PT. All technical and methodological papers without real-world data applications in a medical/patient setting or conducted in study subjects other than humans, and published in languages other than English were excluded.

Expected Results: We anticipate that the results of scoping review will confirm the feasibility and applicability PT for the development of an e-PRO application for HIV patients. Additionally, will lay a novel foundation for the implementation of PT in the context of health apps’ development.

Conclusion: Platform trials are innovative and cost-effective study designs with a potential utility in the evaluation of multiple treatments and interventions like e-health applications

Corresponding author: Asma Aqib,

The Usability of Blended Learning Programs in Health Professions Education: A Scoping Review

Anish Arora1, Charo Rodriguez1, Tamara Carver1, Matthew Hacker Teper1, Tibor Schuster1
1Department of Family Medicine, McGill University

Objective: To develop a foundational understanding of how Blended Learning Programs (BLPs) are assessed in the field of health professions education with respect to the concept of usability.

Methods: A scoping review was conducted, guided by Arksey & O’Malley, 2005’s framework. The PRISMA-ScR checklist was corroborated. Databases searched were Scopus and ERIC. Searches were conducted in an iterative manner, where new terms were identified and added between September 17 and 21 of 2018. Screening was conducted via a questionnaire guide and in significant collaboration with academic librarians and co-authors. Charted data was validated by a co-author. Thematic analysis was conducted by two independent reviewers.

Results: The search strategy identified 8626 studies after removal of duplicates. Title and abstract screening identified 508 studies to be full-text reviewed. Full text revision yielded 53 studies that met eligibility criteria. 47/53 studies utilized a survey/questionnaire or feedback instrument. 33/47 studies did not indicate if their instrument was reliability tested, standardized, or validated. 4 overarching themes were identified: (1) Definitions and their Interrelation; (2) Key Dimensions of Usability that are Consistently Assessed in BLPs; (3) Variability in Research Approaches to Evaluation; and (4) The Discussed Benefits of Blended Learning.

Conclusion: Results suggest that consensus exists among researchers regarding which aspects of BLPs must be assessed, but not on how BLPs should be assessed. The findings of this review can be used to inform future BLP evaluation studies, as well as studies that apply the concept of usability outside the domain of integrated technology.

Corresponding author: Anish Arora,

Towards neuropsychiatric symptom improvement in patients with moderate to severe Alzheimer’s Disease: Adapting art therapy to address visuoperceptual deficits

Elena Guseva1,2, Ovidiu Lungu2,3, Isabelle Vedel1, Sondra Goldman4, Machelle Wilchesky1,2,5
1Department of Family Medicine, McGill University
2Centre for Research in Aging, Donald Berman Maimonides Geriatric Centre; Département de psychiatrie, Université de Montréal
3Centre de recherche de l'Institut universitaire de gériatrie de Montréal
4International Art Hives Network, Concordia University, Donald Berman Maimonides Geriatric Centre
5Lady Davis Institute for Medical Research, Jewish General Hospital

Background: Nonpharmacological clinical management strategies for neuropsychiatric symptoms (NPS) in dementia have limited effectiveness and require considerable resources. While art therapy (AT) is a candidate non-pharmacological management option, the progressive visuoperceptual deficits associated with Alzheimer’s disease (AD) restrict the range and effectiveness of available AT interventions.

Objective: To provide proof-of-concept evidence that visually-enhanced AD-adapted AT (VEAT-AD) may be more effective than non-adapted AT interventions (NAAT) in reducing agitation/aggression and apathy

Methods: A randomized crossover study will be conducted to compare VEAT-AD and NAAT in reducing our two outcomes. NPS will be evaluated using the Neuropsychiatric Inventory via interviews with long-term care nursing staff. Wearable devices will provide both emotion recognition data using physiological signals (e.g. changes in electrodermal activity, heart rate) during AT sessions, and behavioral change data (e.g. heart rate, sleep-wake cycle, activity levels) of participants between sessions. AT sessions will be video recorded, coded, and analyzed using the Engagement of a Person with Dementia Scale, and cross validated with physiological response data.

Expected Results: We anticipate that: 1) VEAT-AD will result in greater NPS reduction as compared with NAAT in this population; and 2) wearable technologies will provide reliable objective data to assess intervention effects.

Conclusion: This will be the first dementia AT study that to collect and analyze objective neurophysiological measurements obtained using wearable devices. Our results will inform a future large clinical trial to rigorously assess the efficacy of VEAT-AD in reducing NPS in moderate to advanced dementia.

Corresponding author: Elena Guseva,

An Ethical Analysis of Health Canada’s Policy Proposals for Reimbursing Gamete Donors: A Patient-Centered Care Perspective

Mary Henein1,2, Carolyn Ells1,2,3
1Department of Family Medicine, McGill University
2Biomedical Ethics Unit, McGill University
3Department of Medicine, McGill University

Objective: While financial compensation for donor gametes is illegal in Canada, reimbursing gamete donors for donor-related expenses is permitted. In 2017, Health Canada (HC) proposed a policy to regulate gamete donor reimbursement, which has implications for gamete donors and recipient’s wellbeing. This research aimed to analyze the ethical implications of HC’s proposed policies for reimbursing gamete donors and determine whether they align with a patient-centered care (PCC) framework.

Design: The research objective was achieved by conducting a critical ethical analysis using a PCC framework. The PCC framework consists of four overarching values: communication, respect for autonomy, respect for dignity, and care. The proposed policy was extracted from HC’s report “Towards a Strengthened Assisted Human Reproduction Act”. The extracted data was analyzed by outlining the proposed regulations and then applying the PCC framework to analyze the proposed policy.

Results: HC’s policy proposals align with the PCC framework by including costs related to caring for dependants as an eligible expense, and by seeking stakeholder feedback. However, the policies fell short of upholding PCC by enforcing a narrow list of eligible expenses, such as excluding loss of work-related income, and by delaying the development of gamete donor reimbursement regulations from the initial time that commercial gamete donation was criminalized in 2004.

Conclusion: The policy proposals by HC contribute to advancing Canadian reproductive policy. However, the results from this research raises concern that the policies are restrictive to unique circumstances of gamete donors and recipients. Overall, this research provides ethical input on gamete donor reimbursement policy in Canada.

Corresponding author: Mary Henein,

Knowledge, Attitudes and Practices regarding Noncommunicable Diseases among Community Health Workers in the Eastern Region of São Paulo, Brazil

Catherine Ji1, Tibor Schuster1, Julie Silvia Martins2, Yves Bergevin1
2Department of Family Medicine, McGill University
3Atenção Primária à Saúde Santa Marcelina

Objectives: With the growing burden of noncommunicable diseases (NCDs) and shortage of physicians/nurses in Brazil, community health workers (CHWs) have the potential of improving chronic disease management at the primary healthcare level, but more NCD-tailored training is needed. This study aims to determine the knowledge, attitudes and practices (KAP) of CHWs in São Paulo regarding NCDs, as a preliminary step to improve training curricula.

Methods: A cross-sectional survey study was conducted among CHWs working for the Santa Marcelina Network in the Eastern Region of São Paulo using an online self-administered KAP questionnaire, developed based on previously validated instruments, between January 28 and February 22, 2019. Descriptive analysis was performed for sociodemographic characteristics and all KAP questions, and scores were calculated from the sum of correct knowledge answers.

Results: 1071 CHWs completed the questionnaire (85% response rate). The respondents were mainly women (96%), had completed secondary school (67%) and less than 5 years of work experience (61%). The mean overall knowledge score was 62% (95%CI: 61.4-62.5), with the highest score for cardiovascular disease questions (80.2%, 95%CI: 79.5-81.0) and lowest for chronic pulmonary disease (28%, 95%CI: 26.1-29.9) and cervical cancer (32.9%, 95%CI: 31.5-34.2). Interestingly, 71.7% reported always discussing cervical cancer screening with community members and 87.1% felt confident/very confident about their counselling on this topic.

Conclusion: Despite high levels of self-reported confidence, CHWs’ knowledge was moderate to poor for many NCD topics. Understanding CHWs’ knowledge gaps and current engagement in NCD management will help optimize their training and clinical activities to better tackle the NCD epidemic.

Corresponding author: Catherine Ji,

Stakeholder recommendations to increase access of marginalized young women to government support programs in Botswana

David Loutfi1, Susan Law2, Neil Andersson1, Ann Cockcroft1
1Department of Family Medicine, McGill University
2Institute of Health Policy, Management, and Evaluation, University of Toronto

Background: Young women in southern Africa are at high risk of contracting HIV due poor education, poverty, and gender-based violence. The INSTRUCT trial in Botswana helps young women access government support programs that improve their livelihoods and may reduce their reliance on risky transactional sex. Recruiting the most vulnerable young women was challenging; our goal was to explore novel methods to reach and engage this group.

Methods: Using deliberative dialogue, we presented findings from our 2017 social network survey of young women to groups of stakeholders and discussed strategies for reaching young women with information about support programs. In two urban and two rural locations, we held nine discussion groups in Setswana (four with young women (n=46), four with community leaders (n=43), and one with service providers (n=4)). We used thematic analysis to explore their proposed strategies for engaging young women.

Findings: Suggestions for sharing information included forming local committees of young women, using existing village committees, and using social media. Social media may be more effective among young women in urban areas (with better internet connectivity); committees of young women may be more effective in rural areas where young women were better connected to each other. Trust was an important factor in determining who young women sought information from.

Conclusions: A network approach suggests strengthening existing ties among community members and building trust to form new ties to service providers. The participatory process of sharing findings with stakeholders and discussing local solutions could be relevant to work with other hard-to-reach populations.

Corresponding author: David Loutfi,

Identifying primary healthcare research priorities in East São Paulo, Brazil: A Delphi study among key stakeholders

Amanda Marcinowska1, Yves Bergevin1, Isabelle Vedel1, Tibor Schuster1
1Department of Family Medicine, McGill University

Objectives: To elicit and prioritize primary healthcare research needs in East São Paulo based on key stakeholders’ experiential knowledge within the domains of child and maternal health, communicable diseases, non-communicable diseases, mental health, urgencies and emergencies, patient safety, and care coordination.

Methods: The Delphi method, a group communication process consisting of a series of questionnaire rounds, will be used. The Delphi questionnaires will be web-based surveys, distributed across a broad range of stakeholders of the Santa Marcelina Network that span the spectrum of primary healthcare. The participants will be recruited through purposeful and snowball sampling. In round one, participants will list important primary healthcare research topics in each domain. Thematic analysis of the data will be used to identify topics for the round two questionnaire. In round two, participants will rate the level of importance of each topic on a Likert scale. In round three, participants will reflect on and revise or maintain their ratings in light of the statistical group response. Data from rounds two and three will be analyzed using descriptive statistics.

Anticipated Results: A prioritized list of research topics within the primary healthcare domains will be obtained with an assessment of importance for each topic, reflecting the needs of East São Paulo in rank-order.

Conclusion: The primary healthcare research needs identified by key stakeholders will help inform future research in East São Paulo and direct resources to the most essential research needs. This will stimulate research that can have an impact on practice, education, and policy in East São Paulo.

Corresponding author: Amanda Marcinowska,

Psychotropic medications and healthcare utilization in seniors with dementia: a population-based study

Eva Margo-Dermer1, Isabelle Vedel1, Catherine Hudon2
1Department of Family Medicine, McGill University
2Département de médecine de famille et de médecine d'urgence, Université de Sherbrooke

Background and Objective: Canada’s rising incidence of dementia poses challenges to healthcare systems. Persons with dementia have twice the acute hospital use as older persons without dementia. Factors associated with dementia patients’ healthcare utilization (e.g.: polypharmacy, comorbidities) have been evaluated individually, but the combination of factors that characterize different types of healthcare utilization by people with dementia is poorly understood. The objective of this study is to evaluate profiles of healthcare utilization by people with dementia.

Methods: Data for this retrospective study will be collected from the INSPQ provincial administrative database; data for an incident dementia cohort aged 65+ from March 31st, 2014-March 31st, 2015 will be analyzed using latent class analysis. Patients will be classified based on their healthcare utilization over this one-year period: utilization components will be primary care visits, emergency department visits, hospitalizations and long-term care transfers. A descriptive analysis of factors that characterize these latent classes (including sex, age, psychotropic prescriptions, other prescriptions, socio-economic status, comorbidities and mortality) will be conducted.

Anticipated Results: We hypothesize that latent classes will be characterized according to utilization of different healthcare services rather than low to high general health services use. We also predict there will be a higher prevalence of psychotropic prescriptions and proportion of women in groups with more hospital visits.

Implications: Patients’ transitions between healthcare utilization classes may be an indication for clinicians to implement changes to their care. This study is expected help minimize unnecessary healthcare utilization by people with dementia and to contribute to the development of sustainable dementia management strategies.

Corresponding author: Eva Margo-Dermer,

Association between outdoor air pollutants and systolic blood pressure: Results from a cross-sectional analysis of Montreal adolescents

Erica Marrone1, Cristina Longo2, Jennifer O’Loughlin3, Paul Villeneuve4, Michael Zappitelli5, Tracie A. Barnett1,6,7
1Department of Family Medicine, McGill University
2Department of Respiratory Medicine, Amsterdam Medical Centre
3School of Public Health, University of Montreal
4Department of Health Sciences, Carleton University
5Department of Pediatrics, Toronto Hospital for Sick Kids
6Department of Epidemiology and Biostatistics, INRS-Institut Armand Frappier
7CHU Sainte-Justine Children's Hospital Research Centre

Objective: In adults, air pollution is associated with elevated blood pressure (BP), but few studies have examined this relationship in youth. The objective of this cross-sectional study was to quantify the association between air pollutants [fine particulate matter (PM2.5) and nitrogen dioxide (NO2)] and systolic BP (SBP) among grade 11 adolescents from Montreal, Canada.

Methods: Participants were students enrolled in the Nicotine Dependence in Teens study with available anthropometric and postal code data. Exposure data were provided by the Canadian Urban Environmental Health Research Consortium and matched to participants’ residential postal code. Elevated SBP was defined as SBP≥90th percentile (computed using BP reference data from the American Academy of Pediatrics). Binary logistic regression models were used to estimate adjusted ORs for each pollutant with respect to elevated SBP.

Results: 508 adolescents were included (mean age: 16.9, 46% male). 4% had SBP measures ≥90th percentile. The odds of elevated SBP increased by 33% (OR 1.33, 95% CI 0.64 to 3.05) for every interquartile range (IQR) increase in residential PM2.5 levels. Similarly, the odds of elevated SBP increased by 17% (OR: 1.17, 95% CI 0.47 to 2.70) for every IQR increase in residential NO2 levels.

Conclusion: These data support the presence of a relationship between exposure to air pollutants (PM2.5 and NO2) and SBP in adolescents, which warrants further investigation using longitudinal data. Family physicians who treat adolescents should be aware of this relationship since the prevalence of primary hypertension (i.e. elevated BP with no identifiable cause) among this population is increasing.

Corresponding author: Erica Marrone,

Family doctor, educator, and learner: Being a clinical teacher in an international blended education program for faculty development

David Ortiz-Paredes1, Charo Rodríguez1, Torsten Risør2, Tamara Carver1, Peter Nugus1,3
1Department of Family Medicine, McGill University
2Department of Public Health, The Artic University of Norway
3Centre for Medical Education, McGill University

Introduction and Objective: Clinical teachers perform overlapping tasks and have to juggle many professional identities (PIs) such as educator and clinician. Little is known about how they manage and make sense of their PIs. The present research explored the lived experiences of clinical teachers in negotiating their PIs during a faculty development program (FDP).

Methodology: This study employed interpretative phenomenological analysis; a methodology that explores in-depth how people conceptualize their personal and social worlds adopting an insider’s perspective. Six Brazilian teachers in family medicine participating in an FDP were purposively sampled. In-depth semi-structured individual interviews were conducted and thematically analyzed.

Results: Each participant had a unique lived experience, but commonalities across them are identified. All participants made sense of their multiple PIs by identifying a single identity comprising others. Participants integrated their PIs in agreement with their personal identities, values, and beliefs. Participants understood their craft as a relational process by which they weave themselves into their context and entangle their experience with that of others. For some, the FDP was an opportunity to understand a collective identity and recognizing a community of practice. Participants, however, diverged when recognizing who their peers are: whereas some named a single group (family physicians), others have a more comprehensive view including even patients.

Conclusion: Clinical teachers have multifaceted identities, to which they give sense, manage and integrate into their daily practice in diverse ways. Understanding these endeavours enriches current perspectives of what it is like to be a teacher in family medicine. Faculty developers should consider these perspectives to tailor these programs adequately.

Corresponding author: David Ortiz-Paredes,

Factors associated with short birth interval (kunika) in low- and middle-income countries: preliminary results from a systematic review

Juan Pimentel1, Umaira Ansari2, Neil Andersson1,2, Ann Cockcroft1,2
1Department of Family Medicine, McGill University
2Centro de Investigación de Enfermedades Tropicales (CIET), Universidad Autónoma de Guerrero

Objectives: Studies have reported associations between short birth interval and adverse maternal and child health outcomes. To our knowledge, no previous systematic review has collated evidence on the factors associated with short birth interval. We undertook a systematic review of factors associated with short birth interval in low- and middle-income countries.

Methods: The review searched PubMed, Embase, LILACS, and Popline databases for empirical studies on the topic. We included documents in English, Spanish, French, Italian, and Portuguese, without date restriction. Two independent reviewers screened the articles and extracted the data. We present here interim, descriptive findings.

Results: After screening, 30 of an initial 2802 documents remained for review, 14 of them published after 2010, and 28 of them observational. The studies included 330,804 women. 15 studies were from Asia, 12 from Asia, and three from Latin America. Some 17 studies examined associations between maternal education and short birth interval, 15 examined age, 10 breastfeeding, 10 socioeconomic level, 9 contraception use, 9 previous pregnancy outcomes, 7 parity, 4 culture and ethnicity, 4 sex of the previous child. Definition of factors was heterogeneous and study quality was variable. For most factors, studies reported both positive and negative associations with short birth interval. Longer breastfeeding and male sex of the previous child were consistently associated with longer birth interval.

Conclusions: Preliminary results suggest consistent associations between breastfeeding, sex of the child, and birth interval. The results of the completed review will inform dialogue groups that will address short birth interval in northern Nigeria.

Corresponding author: Juan Pablo Pimentel-Gonzalez,

Analysis of big primary care data by Machine Learning: Random Forest to evaluate patients’ weight and other features as predictors of antidepressant prescribing

Svetlana Puzhko1, Gillian Bartlett1, Tibor Schuster1
1Department of Family Medicine, McGill University

Objectives: Treatment response to certain antidepressants (AD) may depend on patient’s body weight, sex, and other individual genetic and non-genetic factors. We applied a Random Forest (RF) machine learning algorithm to a large Canadian primary care database analysis to evaluate weight and other patient’s characteristics available in Electronic Medical Records (EMR) for prediction of AD prescribing, and to compare performance of RF model to a conventional approach (multivariable logistic regression model).

Methods: Source. EMR from the national Canadian Primary Care Sentinel Surveillance Network (CPCSSN) for 2011-2016; adult patients (18 years and older) with depression. Measures. AD prescribing, weight, age, sex, network identification number (ID), and comorbidities. Analysis. RF classification model with the number of trees set to 300 and multivariable binary logistic regression (MLR).

Results: Among 61699 patients with depression, 41389 were prescribed AD. Five most important predictors of AD prescribing with RF were ranked as follows: network ID (Mean Decrease Accuracy [MDA] = 77.8%), age (MDA = 32.3%), epilepsy (MDA = 31.5%), hypertension (MDA = 21%), and weight (MDA = 13.8%). In the RF model, out-of-bag prediction error = 34%; sensitivity = 93.4%, specificity = 11.6%. Areas Under the Curve were 57.2% and 58.5% for the RF and the MLR, respectively.

Conclusions: RF model showed high sensitivity but low specificity, and its performance was not superior as compared to the MLR model; however, applying RF to analyse large primary care database allowed to determine the importance of socio-demographic and health characteristics available in EMR, including weight, in prediction of AD prescribing.

Corresponding author: Svetlana Puzhko,

Suboptimal immunization coverage among Canadian rheumatology patients in routine clinical care

Tedi Qendro1, María Laura de la Torre2, Pantelis Panopalis3, Elizabeth Hazel3, Brian J. Ward4, Inés Colmegna3, Marie Hudson5
1Department of Family Medicine, McGill University
2Internal Medicine, Rheumatology and Immunology
3Division of Rheumatology, McGill University Health Center
4Infectious Diseases and Immunity in Global Health Program, Research Institute of the McGill University Health Center
5Division of Rheumatology and Lady Davis Institute for Medical Research, Jewish General Hospital

Objective: To assess vaccination coverage and predictors of vaccination among a Canadian population of rheumatology patients in routine clinical care.

Methods: In this cross-sectional study, consecutive adult patients presenting to a tertiary rheumatology clinic at the McGill University Health Center between May and September 2015 were asked to fill a survey on vaccination. Patients self-identified as having rheumatoid arthritis (RA), systemic autoimmune rheumatic diseases (SARD), spondyloarthropathies (SpA), or other diseases (OD). Multivariate logistical regression analyses were performed to evaluate patient and physician factors associated with vaccination (influenza, pneumococcal, hepatitis B virus [HBV]). Published Quebec general population influenza and pneumococcal vaccination rates in those aged ≥65 were used as comparative baseline rates.

Results: 352 patients were included in the analysis (RA:136, SARD:113, SpA:47, OD:56). Vaccination rates were reported as: (1) influenza: RA:48.5%, SARD:42.0%, SpA:31.9%, OD:88.9%, Quebec general population:58.5%; (2) pneumococcal: RA:42.0%, SARD:37.8%, SpA:29.7%, OD:33.3%, Quebec general population:53.2%; (3) HBV: RA:33.6%, SARD:55.6%, SpA:73.5%, OD:36.8%; and (4) herpes zoster: RA:5.6%, SARD:28.6%, SpA:25.0%, OD:16.7%. Physician recommendation was the strongest independent predictor of vaccination across all vaccine types (influenza: OR 8.56, 95% CI 2.80-26.2, p<0.001; pneumococcal: OR 314, 95% CI 73.0-1353, p<0.001; HBV: OR 12.8, 95% CI 5.27-31.1, p<0.001). Disease group, disease duration, comorbidities, treatment type, and being followed by a primary care physician were not significantly associated with vaccination.

Conclusion: There is suboptimal immunization coverage among ambulatory rheumatology patients. An important role for patient and physician education is highlighted from our study, especially as physician recommendation of vaccination was strongly predictive of vaccine uptake.

Corresponding author: Tedi Qendro,

Medical reversals in primary health care: Towards improved clinical decision-making and education of healthcare professionals

Christian Ruchon1, Roland Grad1, Pierre Pluye1, Kristian B. Filion2, David C. Slawson3, Allen F. Shaughnessy4, Tibor Schuster1, Emelie Braschi5, Mathieu Rousseau1, Soumya B. Sridhar6, Anupriya Grover7, Jennifer R.-S. Cheung7, Mark H. Ebell8
1Department of Family Medicine, McGill University
2Departments of Medicine and of Epidemiology, Biostatistics, and Occupational Health, McGill University
3Department of Family Medicine, University of North Carolina at Chapel Hill
4Department of Family Medicine, Tufts University School of Medicine
5Department of Physical Medicine and Rehabilitation, University of Ottawa
6Department of Family Medicine, University of Rochester
7Cambridge Health Alliance, Cambridge
8Department of Epidemiology & Biostatistics, College of Public Health, University of Georgia

Objective: The efficacy of clinical interventions is challenged as new randomized controlled trials (RCTs) emerge. A medical reversal (MR) occurs when new evidence shows an established medical practice to be less effective than originally claimed and contributes to practice change. We seek to identify characteristics of RCTs associated with MR in primary healthcare.

Methods: We obtained 960 synopses, called Patient-Oriented Evidence that Matters (POEMs), published 2002-2007. These POEMs, written to educate clinicians, summarize RCTs of interventions selected for relevance to primary healthcare. The years 2002-2007 were chosen to allow for new RCTs to emerge.
Step 1: Evidence about the efficacy of each intervention will be extracted; first from each POEM-RCT (E1) and then from DynaMed, a knowledge resource (E2). Teams of two physician-raters will independently compare the initial (E1) and updated (E2) evidence of efficacy and categorize each POEM as (1) reversed or (2) not reversed, in 2019.
Step 2: From each POEM-RCT, factors potentially associated with MR such as sample size will be extracted. A random forest and a logistic regression analysis, with MR as the dependent variable, will identify which independent variables are most strongly associated with MR.

Preliminary Results:Of 408 POEMs analyzed:
Step 1: 34 POEMs (8.3%; 95% CI, 6.0 to 11.4) were categorized as reversed.
Step 2: Mean trial sample size in POEMs identified as reversed, compared to not reversed, was: (x=1870, sd = 6811) versus (x= 2414, sd = 5801).

Conclusion: Knowing which factors are associated with MR will enhance our ability to think more critically about RCTs relevant to primary health care, potentially improving clinical decision-making and the education of health professionals.

Corresponding author: Christian Ruchon,

Assessing the reproductive healthcare needs and priorities of women living with HIV in Canada: a study protocol

Lashanda Skerritt1, Angela Kaida2, Alexandra de Pokomandy1,3
1Department of Family Medicine, McGill University
2Faculty of Health Science, Simon Fraser University
3Chronic Viral Illness Service, McGill University Health Centre

Objectives: In Canada, women living with HIV (WLWH) have poorer HIV-related health outcomes compared to men and experience unmet reproductive healthcare needs. To date, a systematic evaluation has yet to determine whether WLWH have access to reproductive healthcare that aligns with their needs and priorities. Hence, the objectives of this research are to 1) identify determinants of reproductive discussions between WLWH and their healthcare providers, 2) investigate how the reproductive intentions of WLWH change over 18-months, and 3) identify determinants of satisfaction with HIV care and the relative importance of reproductive healthcare from the perspective of WLWH.

Methods: Guided by critical feminist theory and community-based participatory approaches, this research will draw on data from the 1,422 participants of the Canadian HIV Women’s Sexual and Reproductive Health Cohort Study (CHIWOS). Data collected between 2013 and 2018 will be analyzed using descriptive statistics, causal mediation analysis and confirmatory factor analysis. Fuzzy Cognitive Mapping will be used to capture women’s perspectives on factors associated with satisfaction with HIV care. These identified factors will inform the analysis of CHIWOS data to determine how WLWH prioritize reproductive healthcare among other aspects of their HIV care.

Expected Results:This research is expected to highlight the unmet needs for reproductive healthcare for WLWH and identify factors that facilitate access to counselling and support.

Conclusion: Findings may be used to review current guidelines, improve training for healthcare providers, and emphasize the importance of addressing the priorities of WLWH in research and healthcare.

Corresponding author: Lashanda Skerritt,