Mitigating the risk of artificial intelligence bias in cardiovascular care.

Digital health technologies can generate data that can be used to train artificial intelligence (AI) algorithms, which have been particularly transformative in cardiovascular health-care delivery. However, digital and health-care data repositories that are used to train AI algorithms can introduce bias when data are homogeneous and health-care processes are inequitable. AI bias can also be introduced during algorithm development, testing, implementation, and post-implementation processes. The consequences of AI algorithmic bias can be considerable, including missed diagnoses, misclassification of disease, incorrect risk prediction, and inappropriate treatment recommendations. This bias can disproportionately affect marginalised demographic groups. In this Series paper, we provide a brief overview of AI applications in cardiovascular health care, discuss stages of algorithm development and associated sources of bias, and provide examples of harm from biased algorithms. We propose strategies that can be applied during the training, testing, and implementation of AI algorithms to mitigate bias so that all those at risk for or living with cardiovascular disease might benefit equally from AI.

Characteristics of Canadian physicians and their associations with practice patterns: a scoping review.

Background: Physician characteristics such as education and sociodemographic attributes are associated with particular practice patterns, such as practice in rural settings. Understanding the Canadian context of such associations can inform medical school recruitment and health workforce decision-making. Objective: The objective of this scoping review was to report the nature and extent of the literature on associations between characteristics of physicians in Canada and physicians' practice patterns. Eligibility criteria: We included studies reporting associations between 1) the education or sociodemographic attributes of practicing physicians or residents in Canada and 2) practice patterns, including career choice, practice setting, and populations served. Methods: We searched five electronic databases (MEDLINE (R) ALL, Embase, ERIC, Education Source and Scopus) for quantitative primary studies and reviewed reference lists of included studies for additional studies. Data were extracted using a standardized data charting form. Results: Our search yielded 80 studies. Sixty-two examined education, evenly divided between undergraduate and postgraduate. Fifty-eight examined physicians' attributes, most focusing on sex/gender. The majority of studies focused on the outcome of practice setting. We found no studies examining race/ethnicity or socioeconomic status. Conclusion: Many studies in our review found positive associations between (i) rural training or rural background and rural practice setting and (ii) location of training or physicians' origin and practice in that location, consistent with previous literature. Associations for sex/gender were mixed, suggesting it may be a less useful target for workforce planning or recruitment aiming to address gaps in health care provision. More research is needed on the association of characteristics, particularly race/ethnicity and socioeconomic status, with career choice and populations served.

Contexte: Il existe un lien entre les caractéristiques des médecins, comme leur formation et leur profil sociodémographique, et des cadres de pratique particuliers, comme la pratique en milieu rural. La compréhension de ces liens dans le contexte canadien peut éclairer les stratégies d'admission dans les facultés de médecine et la planification de la main-d'œuvre dans le secteur de la santé. Objectif: L'objectif de cette revue exploratoire était de faire état de la nature et de l'étendue de la littérature sur les liens entre les caractéristiques des médecins au Canada et leurs cadres de pratique. Critères de sélection : Nous avons inclus les études établissant des liens entre 1) la formation ou le profil sociodémographique des médecins ou des résidents en exercice au Canada et 2) les cadres de pratique, y compris le choix de carrière, le milieu de pratique et les populations desservies. Méthodes: Nous avons effectué des recherches dans cinq bases de données électroniques (MEDLINE (R) ALL, Embase, ERIC, Education Source et Scopus) pour trouver des études quantitatives primaires et avons examiné les listes de références bibliographiques des articles retenus pour repérer d'autres études. Les données ont été extraites à l'aide d'un formulaire normalisé. Résultats: Notre recherche a permis de recenser 80 études. Soixante-deux d'entre elles portaient sur l'éducation, réparties de manière égale entre le premier cycle et le cycle de spécialisation. Cinquante-huit examinaient les caractéristiques des médecins, la plupart portant sur le sexe/genre. La majorité des études étaient focalisées sur le critère du milieu de pratique. Nous n'avons trouvé aucune étude portant sur la race/ethnicité ou le statut socio-économique. Conclusion: En accord avec des travaux antérieurs de nombreuses études de notre revue trouvent des associations positives entre (i) la formation en milieu rural ou l'origine rurale et un cadre de pratique rural et entre (ii) le lieu de formation ou l'origine du médecin et une pratique dans ce lieu. Les associations relatives au sexe/genre étaient mitigées, ce qui porte à croire qu'il s'agit peut-être d'une cible moins utile pour la planification ou le recrutement d'une main-d'œuvre visant à combler les lacunes dans la prestation des soins de santé. Des travaux supplémentaires sont nécessaires sur les liens entre le profil des médecins, en particulier la race/ethnie et le statut socio-économique, d'une part, et le choix de carrière et les populations desservies d'autre part.

Identifying Primary Care Models of Dementia Care that Improve Quality of Life for People Living with Dementia and their Care Partners: An Environmental Scan.

Dementia is a growing concern in Canada, affecting peoples' health and raising the cost of care. Between June and October 2019, we conducted an environmental scan to identify primary care models, strategies, and resources for dementia care from 11 pre-selected countries and assess their impact on quality-of-life measures. Search strategies included a rapid scoping review, grey literature search, and discussions with stakeholders. Eighteen primary care-based models of dementia care were identified. Common factors include team-based care, centralized care/case coordination, individual treatment plans, a stepped-care approach, and support for care partners. Five provinces had released a dementia strategy. Evidence of positive outcomes supported primary care-based models for dementia care, although only one model demonstrated evidence of impact on quality of life. Although these findings are encouraging, further research is needed to identify primary care-based models of dementia care that demonstrably improve quality of life for people living with dementia and their care partners.

Social accountability in undergraduate medical education: A narrative review.

Background: Medical schools have been increasingly called upon to augment and prioritize their social accountability (SA). Approaches to increasing SA may include reorienting and focusing curricular activities on the priority health needs of the region that they serve. To inform the undergraduate medical education (UGME) curriculum renewal at our school, we examined how SA has been expressed in medical education across several countries and the impacts of SA activities on medical student experience and community-level outcomes. Methods: We conducted a narrative literature review using two electronic databases and searched for studies that reported on SA UGME activities implemented in Canada, Australia, New Zealand, the United States, and the United Kingdom. Studies were screened for inclusion based on predetermined eligibility criteria. Results: We included 40 studies for descriptive analysis and categorized UGME activities into five categories: (1) distributed medical education and community-specific placements/services (32; 80%), (2) community engagement and advocacy activities (23; 58%), (3) international elective preparation and experiences (8; 20%), (4) classroom-based learning of SA-related concepts (17; 43%), and (5) student engagement in SA UGME activities (6; 15%). We categorized impact into four main outcomes: student experience (21; 53%), student competencies (11; 28%), future career choice/practice setting (15; 38%), and community feedback (7; 18%). Student experience was most frequently examined, followed by future career choice/practice setting. Discussion: SA was primarily expressed in UGME activities through placement/service activities and most frequently assessed through student experiences. Student experiences of SA UGME activities have been reported to be largely positive, with benefits also reported for student competencies and influences on future career choice/practice setting. The expression of SA through community engagement in the development of curricular activities indicates a positive shift from social responsibility to SA, but a highly socially accountable curriculum would increasingly consider measures of community impact.

The Provincial Spread and Scale of the Ontario eConsult Service: Evaluation of the First 2 Years.

This paper reports on a multimethod cross-sectional study of the Ontario electronic consultation (eConsult) service. Utilization and closeout survey data from July 2018 through June 2020 were analyzed using the Reach Effectiveness Adoption Implementation Maintenance (RE-AIM) framework. Requesting clinicians submitted 60,474 eConsults, and monthly cases increased from 1,487 in July 2018 to 4,179 in June 2020. The median specialist response time was 1 day. An originally contemplated referral was avoided in 51% of cases. Ontario eConsult showed successful uptake across Ontario, demonstrating continued spread and scale, and offering a template for trailblazers looking to implement digital health innovations in their own jurisdictions.

Neighborhood-Level Influences and Adolescent Health Risk Behaviors in Rural and Urban Sub-Saharan Africa: A Systematic Review.

The impact of neighborhoods on adolescent engagement in health-risk behaviors (HRBs), such as substance use and sexual activity, has been well documented in high-income countries; however, evidence from low and middle-income country settings is limited, particularly in sub-Saharan African (SSA) countries. Neighborhoods and communities in SSA continue to experience urbanization, epidemiologic transition, and the simultaneous presence of large populations living in rurality and urbanicity. This is a dynamic context for exploring adolescent health challenges. This review seeks to identify and summarize existing literature that investigates adolescent engagement in HRBs when compared across rural and urban neighborhoods across SSA. We performed searches using three electronic databases, targeted grey literature searches and scanned reference lists of included studies. Following dual-screening, our search yielded 23 relevant studies that met all inclusion criteria. These were categorized into six broad themes including studies on: (1) sexual risk taking, (2) injury-related, (3) violence, (4) eating and/or exercise-related, (5) substance use, and (6) personal hygiene. We found that neighborhood factors relating to accessibility and availability of health information and care impacted adolescent engagement in HRBs in rural and urban areas. Urbanization of areas of SSA plays a role in differences in engagement in HRBs between rural and urban dwelling adolescents.

Primary Care Providers' Perspectives on the Ontario eConsult Program.

Background: Electronic consultation (eConsult) allows asynchronous virtual communication between primary care providers (PCPs) and specialists regarding patient care. Introduction: The Ontario eConsult Program enables timely and equitable access to specialist advice for Ontarians. This study examines clinicians' perspectives and experiences with the program. Materials and Methods: We conducted an anonymous survey of PCPs registered for the Ontario eConsult Program. The survey ran from June to August 2019 and included questions on PCPs' experiences with the service, opinions on remuneration, and recommendations for enhancement. Results: One thousand two hundred fifty-six PCPs completed the survey (response rate of 16%). Seventy-eight percent had submitted an eConsult, of whom 67% were active users (i.e., had submitted ≥3 eConsults in the past 6 months). The majority of PCPs stated that their user experience was very good (57%) or good (31%), 74% agreed that eConsult improved their referral decision making, and 73% agreed that eConsult increased their ability to manage a broader array of diagnoses. Thirty-seven percent felt adequately compensated for using eConsult, 30% wanted higher rates of remuneration, and 31% were not compensated or were unaware of the fee code. Discussion: The majority of PCPs who use eConsult had positive experiences with the service. Nevertheless, improvements to further streamline the service's use, particularly through electronic medical record integration, were broadly cited as a desirable improvement. Conclusions: PCPs expressed an overall positive experience with the Ontario eConsult Program, citing prompt response times and improved care delivery as chief benefits.

Key factors for national spread and scale-up of an eConsult innovation.

BACKGROUND: Expanding healthcare innovations from the local to national level is a complex pursuit requiring careful assessment of all relevant factors. In this study (a component of a larger eConsult programme of research), we aimed to identify the key factors involved in the spread and scale-up of a successful regional eConsult model across Canada. METHODS: We conducted a constant comparative thematic analysis of stakeholder discussions captured during a full-day National eConsult Forum meeting held in Ottawa, Canada, on 11 December 2017. Sixty-four participants attended, representing provincial and territorial governments, national organisations, healthcare providers, researchers and patients. Proceedings were recorded, transcribed and underwent qualitative analysis using the Framework for Applied Policy Research. RESULTS: This study identified four main themes that were critical to support the intentional efforts to spread and scale-up eConsult across Canada, namely (1) identifying population care needs and access problems, (2) engaging stakeholders who were willing to roll up their sleeves and take action, (3) building on current strategies and policies, and (4) measuring and communicating outcomes. CONCLUSIONS: Efforts to promote innovation in healthcare are more likely to succeed if they are based on an understanding of the forces that drive the spread and scale-up of innovation. Further research is needed to develop and strengthen the conceptual and applied foundations of the spread and scale-up of healthcare innovations, especially in the context of emergent learning health systems across Canada and beyond.

How long are Canadians waiting to access specialty care? Retrospective study from a primary care perspective.

OBJECTIVE: To calculate patient wait times for specialist care using data from primary care clinics across Canada. DESIGN: Retrospective chart audit. SETTING: Primary care clinics. PARTICIPANTS: A total of 22 primary care clinics across 7 provinces and 1 territory. MAIN OUTCOME MEASURES: Wait time 1, defined as the period between a patient's referral by a family physician to a specialist and the visit with said specialist. RESULTS: Overall, 2060 referrals initiated between January 2014 and December 2016 were included in the analysis. The median national wait time 1 was 78 days (interquartile range [IQR] of 34 to 175 days). The shortest waits were observed in Saskatchewan (51 days; IQR = 23 to 101 days) and British Columbia (59 days; IQR = 29 to 131 days), whereas the longest were in New Brunswick (105 days; IQR = 43 to 242 days) and Quebec (104 days; IQR = 36 to 239 days). Median wait time 1 varied substantially among different specialty groups, with the longest wait time for plastic surgery (159 days; IQR = 59 to 365 days) and the shortest for infectious diseases (14 days; IQR = 6 to 271 days). CONCLUSION: This is the first national examination of wait time 1 from the primary care perspective. It provides a picture of patient access to specialists across provinces and specialty groups. This research provides decision makers with important context for developing programs and policies aimed at addressing the largely ignored stage of the wait time continuum from the time of referral to eventual appointment time with the specialist.

Peer-led Self-management Interventions and Adherence to Antiretroviral Therapy Among People Living with HIV: A Systematic Review.

Adherence to antiretroviral therapy (ART) is essential to reduce morbidity and mortality among people living with HIV (PLWH). However, adherence remains suboptimal, and PLWH may benefit from more self-management support to address the complexities of chronic illness. Our objective was to identify the impact of peer-led self-management interventions on adherence and patient-reported outcomes (PROs) among PLWH. We searched MEDLINE, PubMed, Embase, PsycINFO, and CINAHL for English language publications from 1996 to March 2018, and included controlled intervention studies. Additional articles were handsearched, risk of bias assessed, and narrative syntheses outlined. Thirteen studies met inclusion criteria. Findings demonstrate unclear effectiveness for peer-led self-management interventions improving ART adherence; however evidence was limited with only seven studies measuring this outcome and some risk of bias. Many PROs were measured, with limited consistent findings. Future research is needed to strengthen the evidence regarding effects of peer-led self-management interventions on adherence and PROs among PLWH.

Assessment of the Generalizability of an eConsult Service through Implementation in a New Health Region.

INTRODUCTION: Excessive wait times for specialist care are a significant issue in many countries. Electronic consultation (eConsult) services have demonstrated the ability to improve access to specialist care. In this article, we evaluated the implementation of a successful eConsult service in a new jurisdiction to test its generalizability. METHODS: We used a multimethod approach to evaluate the Champlain Building Access to Specialists through eConsultation eConsult service's implementation in the South East Local Health Integration Network of Ontario, Canada. Our quantitative analysis drew on use data collected automatically by the service and survey responses completed between February 1, and June 15, 2017. For our qualitative analysis, we conducted a thematic analysis of 3 focus groups with primary care providers and specialists participating in the pilot study. RESULTS: Forty-nine out of the potential 219 primary care providers in Kingston submitted 301 cases to 24 specialty groups during the study period. Monthly case volume grew from 15 in February to 90 in May. The most frequently requested specialties included dermatology (n = 59), cardiology (n = 27), and gastroenterology (n = 26). Specialists responded in a median of 2 days, and a referral was originally contemplated but ultimately avoided in 40% of cases. Providers spoke positively of the service, citing high levels of satisfaction, enhanced collegiality, increased trust, and improved patient flow. CONCLUSIONS: Adoption of the eConsult service in the South East Local Health Integration Network was successful. The service exceeded all adoption targets, and the number of completed cases demonstrated a consistently upward trend, suggesting continued growth beyond the study's duration. The service's rate of adoption, high levels of satisfaction, and use data similar to other regions all demonstrate eConsult's generalizability.

A Systematic Review of Asynchronous, Provider-to-Provider, Electronic Consultation Services to Improve Access to Specialty Care Available Worldwide.

BACKGROUND: Electronic consultation (eConsult) is an asynchronous electronic communication tool allowing primary care providers to obtain a specialist consultant's expert opinion in a timely manner, thereby offering a potential solution to excessive wait times for specialist care, which remain a serious concern in many countries. INTRODUCTION: Our 2014 review of eConsult services demonstrated feasibility and high acceptability among patients and providers. However, gaps remain in knowledge regarding eConsult's impact on system costs and patient outcomes. MATERIALS AND METHODS: Following the PRISMA guidelines, we conducted a systematic review in May 2017 of English and French literature on OVID Medline, EMBASE, ERIC, and CINAHL databases, examining all studies on eConsult services published since our previous review. The Quadruple Aim Framework was used to synthesize outcomes. Articles reporting on the impact of eConsult on access, patient safety and satisfaction, utilization rates, clinical workflow, and continuing medical education were analyzed using a narrative synthesis approach. RESULTS: The initial search yielded 1,021 results, 50 of which were included on abstract and received a quality assessment and full text review. Of these, 43 were included in our final analysis. Results demonstrated the worldwide presence of eConsult services in North America and countries beyond, including Brazil, Australia, Spain, and The Netherlands. The breadth of specialty services offered has greatly expanded beyond dermatology and includes cardiology, nephrology, and hematology among others. Overall impact on access measures, acceptability, cost, and provider satisfaction remain positive. There is limited research on population health outcomes of morbidity and mortality. CONCLUSIONS: The availability of eConsult services has spread both geographically and in terms of specialty services offered. By allowing for a greater population to be served, access to care is being improved; however, long-term impact should continue to be assessed with a focus on patient safety, morbidity, mortality, and cost effectiveness metrics.

Understanding Patient Referral Wait Times for Specialty Care in Ontario: A Retrospective Chart Audit.

CONTEXT: When examining wait times for specialist care, the duration between a patient's referral and specialist visit (wait time one) is poorly understood. OBJECTIVES: To calculate wait time one in primary care clinics across Ontario using chart audit. METHODS: We conducted a retrospective chart audit at five Ontario-based primary care clinics in 2014-2015. RESULTS: We analyzed 461 referrals. Median wait time one for non-urgent and urgent referrals was 79 and 49 days, respectively. Gastroenterology, obstetrics/gynecology, and ear, nose and throat received the most referrals. Wait times were longest for dermatology (112 days) and shortest for general surgery (32 days). CONCLUSION: Wait times vary substantially by referral urgency and specialty type in Ontario. Calculating wait time from primary care clinics directly offers new perspectives on wait time one and enables clinics to target improvement efforts to best meet patient needs. Our findings will be relevant to providers and policy makers interested in implementing strategies to reduce wait times.