Examining the impacts of sociotechnical systems factors on safety in clinical systems.
What we do
The Safety, Equity, & Design (SED) lab is directed by Dr. Myrtede Alfred, an assistant professor in the Mechanical and Industrial Engineering at the University of Toronto. The lab examines sociotechnical systems factors contributing to adverse events in clinical systems. Complementing the social determinants of health framework, the research lab also leverages human factors and systems engineering to examine clinical systems’ contributions to racial/ethnic disparities in health outcomes. We are currently conducting AHRQ and NSERC funded research investigating maternal health disparities, remote patient monitoring, and retained foreign objects.
Latest Publications
Advancing healthcare equity through human factors engineering
Inequities in health care and corresponding disparities in health outcomes are pressing issues across the world. Preventable differences in health outcomes, including life expectancy, prevalence of chronic diseases, and causes of mortality, have been documented in the United States (US), Canada, the United Kingdom (UK), Brazil, France, and other many other countries (Hughes et al., 2022, p. 19; Ray et al., 2018; Sheikh et al., 2022). Disparities are associated with patients’ socioeconomic status, immigration status, geographic region, disability, gender and sexual identity (National Academies of Sciences et al., 2017). Additionally, racial/ethnic disparities are well-established. In the US, for example, Black, Indigenous, and Hispanic populations report worse healthcare experiences and health outcomes compared to White populations (2023 National Healthcare Quality and Disparities Report, 2023.; Advancing Racial Equity in U.S. Health Carayon et al., 2020; Unequal Treatment, 2003). Resource availability, environmental conditions, access to health care, and structural influences on people’s ability to adhere to medical recommendations are significant drivers of population health (Braveman et al., 2011). At the same time, there is increasing recognition that the quality of medical care is a key contributor to disparities in health outcomes with prior research identifying differences in the quality of care and utilization of health services for racialized individuals (2023 National Healthcare Quality and Disparities Report, 2023; Unequal Treatment, 2003). Over the past twenty years, the US has made little progress on reducing racial and ethnic disparities in access to and quality of care and similar challenges persist with other marginalized communities, including those living with disabilities (Benjamin et al., 2024; Mitra et al., 2022). This work reaffirmed that improving medical care within and across settings is critical to addressing disparities in outcomes and cultivating conditions that support health and well-being of the diverse populations served by our healthcare system.
Spatial and Racial/Ethnic Variation in the Prevalence of Cesarean Delivery in a South Carolina Medical Center
Jeffrey T. Howard, Sam Lawton, Dulaney Wilson, Amartha Gore, Latha Hebbar, Christine Morton, Christopher Goodier & Myrtede Alfred
While racial/ethnic disparities in cesarean delivery have been noted in the literature, less is known about the intersection between individual-level race/ethnicity and community-level social vulnerability as factors in cesarean delivery. The goal was to use medical record data from a large medical center combined with census tract-level data to examine patterns of social vulnerability, racial population distribution, and prevalence of cesarean delivery.
Data were obtained from electronic medical records of patients from a large medical center in South Carolina from 2019 to 2020. The outcome variable was cesarean delivery (yes/no), and covariates included the year of delivery; age of patient; race/ethnicity; spoken language; BMI categories; clinical indications of anemia, hypertension, preeclampsia, and diabetes; and census tract Social Vulnerability Index (SVI). Generalized linear mixed models for multilevel binary logistic regression were used to test the main hypothesis that the census tract level Social Vulnerability Index is positively associated with cesarean delivery.
Among a total of 5011 patients, we found that non-Hispanic Black mothers were more likely to have cesarean deliveries compared with non-Hispanic White mothers. After controlling for census tract-level SVI, the individual-level race/ethnicity association was no longer significant. Significant spatial autocorrelation across census tracts was evident for cesarean delivery prevalence, non-Hispanic Black population, and SVI. A high prevalence of cesarean delivery tended to cluster with high SVI and a high non-Hispanic Black population.
We found that non-Hispanic Black mothers were more likely to have cesarean deliveries, which was explained by census tract differences in the SVI.
Escalation Pathways of Remote Patient Monitoring Programs for COVID-19 Patients in Canada and the United States: A Rapid Review
Introduction: During the COVID-19 pandemic, hospitals in North America were overwhelmed with COVID-19 patients and had limited capacity to admit patients. Remote patient monitoring (RPM) programs were developed to monitor COVID-19 patients at home and reduce disease transmission and the demand on hospitals. A critical component of RPM programs is effective escalation pathways. The purpose of this review is to synthesize the implementation of escalation pathways of RPM programs for COVID-19 patients in Canada and the United States.
Methods: The search identified 563 articles from Embase, PubMed, and Scopus. Following title and abstract screening, 131 were selected for full-text review, and 26 articles were included. Data were extracted on study location, patient eligibility and program size, data collection, monitoring team, escalation criteria, and escalation response.
Results: The included studies were published between 2020 and 2022; 3 in Canada and 23 in the United States. The RPM programs collected physiological vital signs and symptom data, which were inputted manually by patients and health care workers or synced automatically. Escalations were triggered automatically or following manual review by nurses and physicians when signs and symptoms were concerning or reached a specific threshold. Escalations included emergency department referrals, physician appointments, and increased monitoring.
Conclusion: Many decisions are required when designing RPM escalation pathways for patients with COVID-19, which is crucial to promptly address patients’ changing health statuses and clinical needs. Future research is needed to evaluate the effectiveness of escalation pathways for COVID-19 patients through performance metrics and patient and health care worker experience.
Patient safety event (PSE) reports, which document incidents that compromise patient safety, are fundamental for improving healthcare quality. Accurate classification of these reports is crucial for analyzing trends, guiding interventions, and supporting organizational learning. However, this process is labor-intensive due to the high volume and complex taxonomy of reports. Previous work has shown that machine learning (ML) can automate PSE report classification; however, its success depends on large manually-labeled datasets. This study leverages Active Learning (AL) strategies with human expertise to streamline PSE-report labeling. We utilize pool-based AL sampling to selectively query reports for human annotation, developing a robust dataset for training ML classifiers. Our experiments demonstrate that AL significantly outperforms random sampling in accuracy across various text representations, reducing the need for labeled samples by 24% to 69%. Based on these findings, we suggest that incorporating AL strategies into PSE-report labeling can effectively reduce manual workload while maintaining high classification accuracy.