Are at-risk sociodemographic attributes stable across COVID-19 transmission waves?

COVID-19 health impacts and risks have been disproportionate across social, economic, and racial gradients. (Chen et al., 2021, Thompson et al., 2021, Mamuji et al., 2021, COVID-19 and Ethnicity 2020) By examining the first five waves of the pandemic in Ontario, we identify if Forward Sortation Areas (FSAs) based measures of sociodemographic status and their relationship to COVID-19 cases are stable or vary by time. COVID-19 waves were defined using a time-series graph of COVID-19 case counts by epi-week. Percent Black visible minority, percent Southeast Asian visible minority and percent Chinese visible minority at the FSA level were then integrated into spatial error models with other established vulnerability characteristics. The models indicate that area-based sociodemographic patterns associated with COVID-19 infection change over time. If sociodemographic characteristics are identified as high risk (increased COVID-19 case rates) increased testing, public health messaging, and other preventative care may be implemented to protect populations from the inequitable burden of disease.

Centralization and integration of public health systems: Perspectives of public health leaders on factors facilitating and impeding COVID-19 responses in three Canadian provinces.

The extent to which power, resources, and responsibilities for public health are centralized or decentralized within a jurisdiction and how public health functions are integrated or coordinated with health care services may shape pandemic responses. However, little is known about the impacts of centralization and integration on public health system responses to the COVID-19 pandemic. We examine how public health leaders perceive centralization and integration facilitated and impeded effective COVID-19 responses in three Canadian provinces. We conducted a comparative case study involving semi-structured interviews with 58 public health system leaders in three Canadian provinces with varying degrees of centralization and integration. Greater public health system centralization and integration was seen by public health leaders to facilitate more rapidly initiated and well-coordinated provincial COVID-19 responses. Decentralization may have enabled locally tailored responses in the context of limited provincial leadership. Opacity in provincial decision-making processes, jurisdictional ambiguity impacting Indigenous communities, and ineffectual public health investments were impediments across jurisdictions and thus appear to be less impacted by centralization and integration. Our study generates novel insights about potential structural facilitators and impediments of effective COVID-19 pandemic responses during the second year of the pandemic. Findings highlight key areas for future research to inform system design that support leaders to manage large-scale public health emergencies.

Complications Following Elective Major Noncardiac Surgery Among Patients With Prior SARS-CoV-2 Infection.

Importance: There is an urgent need for evidence to inform preoperative risk assessment for the millions of people who have had SARS-CoV-2 infection and are awaiting elective surgery, which is critical to surgical care planning and informed consent. Objective: To assess the association of prior SARS-CoV-2 infection with death, major adverse cardiovascular events, and rehospitalization after elective major noncardiac surgery. Design, Setting, and Participants: This population-based cohort study included adults who had received a polymerase chain reaction test for SARS-CoV-2 infection within 6 months prior to elective major noncardiac surgery in Ontario, Canada, between April 2020 and October 2021, with 30 days follow-up. Exposures: Positive SARS-CoV-2 polymerase chain reaction test result. Main Outcomes and Measures: The main outcome was the composite of death, major adverse cardiovascular events, and all-cause rehospitalization within 30 days after surgery. Results: Of 71 144 patients who underwent elective major noncardiac surgery (median age, 66 years [IQR, 57-73 years]; 59.8% female), 960 had prior SARS-CoV-2 infection (1.3%) and 70 184 had negative test results (98.7%). Prior infection was not associated with the composite risk of death, major adverse cardiovascular events, and rehospitalization within 30 days of elective major noncardiac surgery (5.3% absolute event rate [n = 3770]; 960 patients with a positive test result; adjusted relative risk [aRR], 0.91; 95% CI, 0.68-1.21). There was also no association between prior infection with SARS-CoV-2 and postoperative outcomes when the time between infection and surgery was less than 4 weeks (aRR, 1.15; 95% CI, 0.64-2.09) or less than 7 weeks (aRR, 0.95; 95% CI, 0.56-1.61) and among those who were previously vaccinated (aRR, 0.81; 95% CI, 0.52-1.26). Conclusions and Relevance: In this study, prior infection with SARS-CoV-2 was not associated with death, major adverse cardiovascular events, or rehospitalization following elective major noncardiac surgery, although low event rates and wide 95% CIs do not preclude a potentially meaningful increase in overall risk.

Governing partnerships with technology companies as part of the COVID-19 response in Canada: A qualitative case study.

Cross-sector partnerships are vital for maintaining resilient health systems; however, few studies have sought to empirically assess the barriers and enablers of effective and responsible partnerships during public health emergencies. Through a qualitative, multiple case study, we analyzed 210 documents and conducted 26 interviews with stakeholders in three real-world partnerships between Canadian health organizations and private technology startups during the COVID-19 pandemic. The three partnerships involved: 1) deploying a virtual care platform to care for COVID-19 patients at one hospital, 2) deploying a secure messaging platform for physicians at another hospital, and 3) using data science to support a public health organization. Our results demonstrate that a public health emergency created time and resource pressures throughout a partnership. Given these constraints, early and sustained alignment on the core problem was critical for success. Moreover, governance processes designed for normal operations, such as procurement, were triaged and streamlined. Social learning, or the process of learning from observing others, offset some time and resource pressures. Social learning took many forms ranging from informal conversations between individuals at peer organisations (e.g., hospital chief information officers) to standing meetings at the local university's city-wide COVID-19 response table. We also found that startups' flexibility and understanding of the local context enabled them to play a highly valuable role in emergency response. However, pandemic fueled "hypergrowth" created risks for startups, such as introducing opportunities for deviation away from their core value proposition. Finally, we found each partnership navigated intense workloads, burnout, and personnel turnover through the pandemic. Strong partnerships required healthy, motivated teams. Visibility into and engagement in partnership governance, belief in partnership impact, and strong emotional intelligence in managers promoted team well-being. Taken together, these findings can help to bridge the theory-to-practice gap and guide effective cross-sector partnerships during public health emergencies.

Determinants of Hospital Use and Physician Services Among Adults With a History of Homelessness.

Background: People experiencing homelessness have diverse patterns of healthcare use. This study examined the distribution and determinants of healthcare encounters among adults with a history of homelessness. Methods: Administrative healthcare records were linked with survey data for a general cohort of adults with a history of homelessness and a cohort of homeless adults with mental illness. Binary and count models were used to identify factors associated with hospital admissions, emergency department visits and physician visits for comparison across the 2 cohorts. Results: During the 1-year follow-up period, a higher proportion of people in the cohort with a mental illness used any inpatient (27% vs 14%), emergency (63% vs 53%), or physician services (90% vs 76%) compared to the general homeless cohort. People from racialized groups were less likely use nearly all health services, most notably physician services. Other factors, such as reporting of a regular source of care, poor perceived general health, and diagnosed chronic conditions were associated with higher use of all health services except psychiatric inpatient care. Conclusion: When implementing interventions for patients with the greatest health needs, we must consider the unique factors that contribute to higher healthcare use, as well as the barriers to healthcare access.

Mass testing for asymptomatic COVID-19 infection among health care workers at a large Canadian hospital

Background The perceived risk of coronavirus disease 2019 (COVID-19) infection for health care workers (HCWs) is high. Although testing has focused on symptomatic HCWs, asymptomatic testing is considered by some to be an important strategy to limit occupational spread. Evidence on the results of large asymptomatic testing strategies in health care is, however, limited. This study examines the uptake and positivity of COVID-19 testing in a voluntary asymptomatic testing campaign at a large Canadian hospital. Methods In addition to testing HCWs with symptoms, all asymptomatic staff were offered a COVID-19 test at Trillium Health Partners, a large Ontario hospital, from May 27 to June 15, 2020. Testing was offered in four waves, corresponding to the likelihood of exposure to COVID-19–positive patients. The mass asymptomatic testing campaign was offered when the hospital’s community test positivity rate had declined to 5%. Results Since March 16, the hospital has tested 51.3% of its 10,143-person workforce at least once. In the asymptomatic testing campaign for HCWs between May 27 and June 15, 27% of clinical and non-clinical staff received testing. No large differences were found in the proportions of clinical HCWs tested by their exposure to COVID-19–positive patients. In this campaign, 0.2% of asymptomatic HCWs tested positive. However, these individuals either had mild symptoms at testing and did not self-identify or became symptomatic after testing. Conclusions At this large hospital with declining community prevalence, a mass asymptomatic testing campaign of HCWs found they had a very low likelihood of testing positive for COVID-19.

Can the Healthcare System Improve Food Security? A Need for Collaborative Community Partnerships.

The COVID-19 pandemic has heightened the food insecurity crisis in Canada, and existing supports have been largely insufficient to meet the food needs of communities. In response to increasing reports of food insecurity among Toronto residents during the pandemic, the Food RX program was developed as a collaborative initiative between FoodShare Toronto - a local, community-based food justice organization - and the University Health Network, a large university-affiliated hospital network in downtown Toronto, ON. This commentary describes the Food RX program, highlights the lessons learned during its early implementation and offers a set of recommendations for building community partnerships moving forward.

Early Impacts of the COVID-19 Pandemic on Public Health Systems and Practice in 3 Canadian Provinces From the Perspective of Public Health Leaders: A Qualitative Study.

CONTEXT: The COVID-19 pandemic has impacted health systems worldwide. Studies to date have largely focused on the health care system with less attention to the impact on public health systems and practice. OBJECTIVE: To describe the early impacts of COVID-19 on public health systems and practice in 3 Canadian provinces from the perspective of public health system leaders and synthesize lessons learned. DESIGN: A qualitative study using semistructured virtual interviews with public health leaders between October 2020 and April 2021. The World Health Organization's essential public health operations framework guided data collection and analysis. SETTING: This study involved the Canadian provinces of Alberta, Ontario, and Québec. These provinces were chosen for their large populations, relatively high COVID-19 burden, and variation in public health systems. PARTICIPANTS: Public health leaders from Alberta (n = 21), Ontario (n = 18), and Québec (n = 19) in organizations with a primary mandate of stewardship and/or administration of essential public health operations (total n = 58). RESULTS: We found that the COVID-19 pandemic led to intensified collaboration in public health systems and a change in workforce capacity to respond to the pandemic. This came with opportunities but also challenges of burnout and disruption of non-COVID-19 services. Information systems and digital technologies were increasingly used and there was greater proximity between public health leaders and other health system leaders. A renewed recognition for public health work was also highlighted. CONCLUSIONS: The COVID-19 pandemic impacted several aspects of public health systems in the provinces studied. Our findings can help public health leaders and policy makers identify areas for further investment (eg, intersectoral collaboration, information systems) and develop plans to address challenges (eg, disrupted services, workforce burnout) that have surfaced.

“We cobble together a storyline of system performance using a diversity of things”: a qualitative study of perspectives on public health performance measurement in Canada

There have been longstanding calls for public health systems transformations in many countries, including Canada. Core to these calls has been strengthening performance measurement. While advancements have been made in performance measurement for certain sectors of the health care system (primarily focused on acute and primary health care), effective use of indicators for measuring public health systems performance are lacking. This study describes the current state, anticipated challenges, and future directions in the development and implementation of a public health performance measurement system for Canada. We conducted a qualitative study using semi-structured interviews with public health leaders (n = 9) between July and August 2021. Public health leaders included researchers, government staff, and former medical officers of health who were purposively selected due to their expertise and experience with performance measurement with relevance to public health systems in Canada. Thematic analysis included both a deductive approach for themes consistent with the conceptual framework and an inductive approach to allow new themes to emerge from the data. Conceptual, methodological, contextual, and infrastructure challenges were highlighted by participants in designing a performance measurement system for public health. Specifically, six major themes evolved that encompass 1) the mission and purpose of public health systems, including challenges inherent in measuring the functions and services of public health;2) the macro context, including the impacts of chronic underinvestment and one-time funding injections on the ability to sustain a measurement system;3) the organizational structure/governance of public health systems including multiple forms across Canada and underdevelopment of information technology systems;4) accountability approaches to performance measurement and management;and 5) timing and unobservability in public health indicators. These challenges require dedicated investment, strong leadership, and political will from the federal and provincial/territorial governments. Unprecedented attention on public health due to the coronavirus disease 2019 pandemic has highlighted opportunities for system improvements, such as addressing the lack of a performance measurement system. This study provides actionable knowledge on conceptual, methodological, contextual, and infrastructure challenges needed to design and build a pan-Canadian performance measurement system for public health.

School Reopening And COVID-19 In The Community: Evidence From A Natural Experiment In Ontario, Canada.

In December 2020, Ontario, Canada, entered a provincewide shutdown to mitigate COVID-19 transmission. A regionalized approach was taken to reopen schools throughout early 2021 without any other opening of the economy, offering a unique natural experiment to estimate the impact of school reopening on community transmission. Estimated increases of 0.07, 0.08, 0.07, and 0.13 percentage points in community COVID-19 case growth rates occurred 11-15, 16-20, 21-25, and 26-30 days, respectively, after schools reopened. Although small, these changes were particularly evident among children younger than age fourteen, increased over time, and were greater when lag periods were considered, which points to a likely causal effect between in-person classes and a small increase in transmission. These findings suggest that although additional COVID-19 cases are to be expected after the reopening of schools, these risks may be manageable with sufficient, layered mitigation policies.

CoQUAD: a COVID-19 question answering dataset system, facilitating research, benchmarking, and practice.

BACKGROUND: Due to the growing amount of COVID-19 research literature, medical experts, clinical scientists, and researchers frequently struggle to stay up to date on the most recent findings. There is a pressing need to assist researchers and practitioners in mining and responding to COVID-19-related questions on time. METHODS: This paper introduces CoQUAD, a question-answering system that can extract answers related to COVID-19 questions in an efficient manner. There are two datasets provided in this work: a reference-standard dataset built using the CORD-19 and LitCOVID initiatives, and a gold-standard dataset prepared by the experts from a public health domain. The CoQUAD has a Retriever component trained on the BM25 algorithm that searches the reference-standard dataset for relevant documents based on a question related to COVID-19. CoQUAD also has a Reader component that consists of a Transformer-based model, namely MPNet, which is used to read the paragraphs and find the answers related to a question from the retrieved documents. In comparison to previous works, the proposed CoQUAD system can answer questions related to early, mid, and post-COVID-19 topics. RESULTS: Extensive experiments on CoQUAD Retriever and Reader modules show that CoQUAD can provide effective and relevant answers to any COVID-19-related questions posed in natural language, with a higher level of accuracy. When compared to state-of-the-art baselines, CoQUAD outperforms the previous models, achieving an exact match ratio score of 77.50% and an F1 score of 77.10%. CONCLUSION: CoQUAD is a question-answering system that mines COVID-19 literature using natural language processing techniques to help the research community find the most recent findings and answer any related questions.

Prevalence and health care costs of mitochondrial disease in Ontario, Canada: A population-based cohort study.

BACKGROUND: Mitochondrial disease prevalence has been estimated at 1 in 4000 in the United States, and 1 in 5000 worldwide. Prevalence in Canada has not been established, though multi-linked health administrative data resources present a unique opportunity to establish robust population-based estimates in a single-payer health system. This study used administrative data for the Ontario, Canada population between April 1988 and March 2019 to measure mitochondrial disease prevalence and describe patient characteristics and health care costs. RESULTS: 3069 unique individuals were hospitalized with mitochondrial disease in Ontario and eligible for the study cohort, representing a period prevalence of 2.51 per 10,000 or 1 in 3989. First hospitalization was most common between ages 0-9 or 50-69. The mitochondrial disease population experiences a high need for health care and incurred high costs (mean = CAD$24,023 in 12 months before first hospitalization) within the single-payer Ontario health care system. CONCLUSIONS: This study provides needed insight into mitochondrial disease in Canada, and demonstrates the high health burden on patients. The methodology used can be adapted across jurisdictions with similar routine collection of health data, such as in other Canadian provinces. Future work should seek to validate this approach via record linkage of existing disease cohorts in Ontario, and identify specific comorbidities with mitochondrial disease that may contribute to high health resource utilization.

Human factors applications in the design of decision support systems for population health: a scoping review.

INTRODUCTION: Public health professionals engage in complex cognitive tasks, often using evidence-based decision support tools to bolster their decision-making. Human factors methods take a user-centred approach to improve the design of systems, processes, and interfaces to better support planning and decision-making. While human factors methods have been applied to the design of clinical health tools, these methods are limited in the design of tools for population health. The objective of this scoping review is to develop a comprehensive understanding of how human factors techniques have been applied in the design of population health decision support tools. METHODS AND ANALYSIS: The scoping review will follow the methodology and framework proposed by Arksey and O'Malley. We include English-language documents between January 1990 and August 2021 describing the development, validation or application of human factors principles to decision support tools in population health. The search will include Ovid MEDLINE: Epub Ahead of Print, In-Process and Other Non-Indexed Citations, Ovid MEDLINE Daily and Ovid MEDLINE 1946-present; EMBASE, Scopus, PsycINFO, Compendex, IEEE Xplore and Inspec. The results will be integrated into Covidence. First, the abstract of all identified articles will be screened independently by two reviewers with disagreements being resolved by a third reviewer. Next, the full text for articles identified as include or inconclusive will be reviewed by two independent reviewers, leading to a final decision regarding inclusion. Reference lists of included articles will be manually screened to identify additional studies. Data will be extracted by one reviewer, verified by a second, and presented according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for scoping reviews. ETHICS AND DISSEMINATION: Ethics approval is not required for this work as human participants are not involved. The completed review will be published in a peer-reviewed, interdisciplinary journal.

Large-scale implementation of rapid antigen testing system for COVID-19 in workplaces.

The transmission of coronavirus disease 2019 (COVID-19) in workplaces has been a persistent issue throughout the pandemic. In response, a not-for-profit initiative emerged to mitigate COVID-19 workplace transmission in Canada. We report the process for establishing a workplace frequent rapid antigen test (RAT) program. The screening program identified 473 asymptomatic individuals who tested positive on the RAT and confirmed positive by a nasopharyngeal polymerase chain reaction (PCR) diagnostic test. One in 4300 RATs was presumptive positive but later tested PCR negative, and thus, false positives did not meaningfully disrupt workplace operations. Most employers rated the program highly and felt strongly that the program contributed to workplace and community safety. The findings describe a sustained and scalable implementation plan for establishing a frequent workplace testing program. High-frequency testing programs offer the potential to break chains of transmission and act as an extra layer of protection in a comprehensive public health response.

The Use of Cremation Data for Timely Mortality Surveillance During the COVID-19 Pandemic in Ontario, Canada: Validation Study.

BACKGROUND: Early estimates of excess mortality are crucial for understanding the impact of COVID-19. However, there is a lag of several months in the reporting of vital statistics mortality data for many jurisdictions, including across Canada. In Ontario, a Canadian province, certification by a coroner is required before cremation can occur, creating real-time mortality data that encompasses the majority of deaths within the province. OBJECTIVE: This study aimed to validate the use of cremation data as a timely surveillance tool for all-cause mortality during a public health emergency in a jurisdiction with delays in vital statistics data. Specifically, this study aimed to validate this surveillance tool by determining the stability, timeliness, and robustness of its real-time estimation of all-cause mortality. METHODS: Cremation records from January 2020 until April 2021 were compared to the historical records from 2017 to 2019, grouped according to week, age, sex, and whether COVID-19 was the cause of death. Cremation data were compared to Ontario's provisional vital statistics mortality data released by Statistics Canada. The 2020 and 2021 records were then compared to previous years (2017-2019) to determine whether there was excess mortality within various age groups and whether deaths attributed to COVID-19 accounted for the entirety of the excess mortality. RESULTS: Between 2017 and 2019, cremations were performed for 67.4% (95% CI 67.3%-67.5%) of deaths. The proportion of cremated deaths remained stable throughout 2020, even within age and sex categories. Cremation records are 99% complete within 3 weeks of the date of death, which precedes the compilation of vital statistics data by several months. Consequently, during the first wave (from April to June 2020), cremation records detected a 16.9% increase (95% CI 14.6%-19.3%) in all-cause mortality, a finding that was confirmed several months later with cremation data. CONCLUSIONS: The percentage of Ontarians cremated and the completion of cremation data several months before vital statistics did not change meaningfully during the COVID-19 pandemic period, establishing that the pandemic did not significantly alter cremation practices. Cremation data can be used to accurately estimate all-cause mortality in near real-time, particularly when real-time mortality estimates are needed to inform policy decisions for public health measures. The accuracy of this excess mortality estimation was confirmed by comparing it with official vital statistics data. These findings demonstrate the utility of cremation data as a complementary data source for timely mortality information during public health emergencies.

Pivoting data and analytic capacity to support Ontario's COVID-19 response.

Introduction: Health care systems have faced unprecedented challenges due to the COVID-19 pandemic. Access to timely population-based data has been vital to informing public health policy and practice. Methods: We describe how ICES, an independent not-for-profit research and analytic institute in Ontario, Canada, pivoted existing research infrastructure and engaged health system stakeholders to provide near real-time population-based data and analytics to support Ontario's COVID-19 pandemic response. Results: Since April 2020, ICES provided the Ontario COVID-19 Provincial Command Table and public health partners with regular and ad hoc reports on SARS-CoV-2 testing and COVID-19 vaccine coverage. These reports: 1) helped identify congregate care/shared living settings that needed testing and prevention efforts early in the pandemic; 2) provided early indications of inequities in testing and infection in marginalized neighbourhoods, including areas with higher proportions of immigrants and visible minorities; 3) identified areas with high test positivity, which helped Public Health Units target and evaluate prevention efforts; and 4) contributed to altering the province's COVID-19 vaccine roll-out strategy to target high-risk neighbourhoods and helping Public Health Units and community organizations plan local vaccination programs. In addition, ICES is a key component of the Ontario Health Data Platform, which provides scientists with data access to conduct COVID-19 research and analyses. Discussion and Conclusion: ICES was well-positioned to provide rapid analyses for decision-makers to respond to the evolving public health emergency, and continues to contribute to Ontario's pandemic response by providing timely, relevant reports to health system stakeholders and facilitating data access for externally-funded COVID-19 research.

Outcomes in patients with and without disability admitted to hospital with COVID-19: a retrospective cohort study.

BACKGROUND: Disability-related considerations have largely been absent from the COVID-19 response, despite evidence that people with disabilities are at elevated risk for acquiring COVID-19. We evaluated clinical outcomes in patients who were admitted to hospital with COVID-19 with a disability compared with patients without a disability. METHODS: We conducted a retrospective cohort study that included adults with COVID-19 who were admitted to hospital and discharged between Jan. 1, 2020, and Nov. 30, 2020, at 7 hospitals in Ontario, Canada. We compared in-hospital death, admission to the intensive care unit (ICU), hospital length of stay and unplanned 30-day readmission among patients with and without a physical disability, hearing or vision impairment, traumatic brain injury, or intellectual or developmental disability, overall and stratified by age (≤ 64 and ≥ 65 yr) using multivariable regression, controlling for sex, residence in a long-term care facility and comorbidity. RESULTS: Among 1279 admissions to hospital for COVID-19, 22.3% had a disability. We found that patients with a disability were more likely to die than those without a disability (28.1% v. 17.6%), had longer hospital stays (median 13.9 v. 7.8 d) and more readmissions (17.6% v. 7.9%), but had lower ICU admission rates (22.5% v. 28.3%). After adjustment, there were no statistically significant differences between those with and without disabilities for in-hospital death or admission to ICU. After adjustment, patients with a disability had longer hospital stays (rate ratio 1.36, 95% confidence interval [CI] 1.19-1.56) and greater risk of readmission (relative risk 1.77, 95% CI 1.14-2.75). In age-stratified analyses, we observed longer hospital stays among patients with a disability than in those without, in both younger and older subgroups; readmission risk was driven by younger patients with a disability. INTERPRETATION: Patients with a disability who were admitted to hospital with COVID-19 had longer stays and elevated readmission risk than those without disabilities. Disability-related needs should be addressed to support these patients in hospital and after discharge.

Risk stratification for COVID-19 hospitalization: a multivariable model based on gradient-boosting decision trees.

BACKGROUND: The COVID-19 pandemic has led to an increased demand for health care resources and, in some cases, shortage of medical equipment and staff. Our objective was to develop and validate a multivariable model to predict risk of hospitalization for patients infected with SARS-CoV-2. METHODS: We used routinely collected health records in a patient cohort to develop and validate our prediction model. This cohort included adult patients (age ≥ 18 yr) from Ontario, Canada, who tested positive for SARS-CoV-2 ribonucleic acid by polymerase chain reaction between Feb. 2 and Oct. 5, 2020, and were followed up through Nov. 5, 2020. Patients living in long-term care facilities were excluded, as they were all assumed to be at high risk of hospitalization for COVID-19. Risk of hospitalization within 30 days of diagnosis of SARS-CoV-2 infection was estimated via gradient-boosting decision trees, and variable importance examined via Shapley values. We built a gradient-boosting model using the Extreme Gradient Boosting (XGBoost) algorithm and compared its performance against 4 empirical rules commonly used for risk stratifications based on age and number of comorbidities. RESULTS: The cohort included 36 323 patients with 2583 hospitalizations (7.1%). Hospitalized patients had a higher median age (64 yr v. 43 yr), were more likely to be male (56.3% v. 47.3%) and had a higher median number of comorbidities (3, interquartile range [IQR] 2-6 v. 1, IQR 0-3) than nonhospitalized patients. Patients were split into development (n = 29 058, 80.0%) and held-out validation (n = 7265, 20.0%) cohorts. The gradient-boosting model achieved high discrimination (development cohort: area under the receiver operating characteristic curve across the 5 folds of 0.852; validation cohort: 0.8475) and strong calibration (slope = 1.01, intercept = -0.01). The patients who scored at the top 10% captured 47.4% of hospitalizations, and those who scored at the top 30% captured 80.6%. INTERPRETATION: We developed and validated an accurate risk stratification model using routinely collected health administrative data. We envision that modelling such risk stratification based on routinely collected health data could support management of COVID-19 on a population health level.