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

Can’t sign in? Forgot your password?
Enter your email address below and we will send you the reset instructions
If the address matches an existing account you will receive an email with instructions to reset your password.
Can’t sign in? Forgot your username?
Enter your email address below and we will send you your username
If the address matches an existing account you will receive an email with instructions to retrieve your username
Your password has been changed

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.
The role of school closures as a mitigation policy for the COVID-19 pandemic has been hotly debated. Although closures may slow the spread of respiratory viruses, the impact of these measures on curbing COVID-19 remains less clear.1,2 Children are much less prone to developing severe COVID-19 than older people, although the long-term effects of infection remain unknown.14 In contrast, it has been widely shown that in-person classes substantially benefit child well-being and learning, as well as having wider societal benefits in terms of allowing parents to engage in work.15 Many have argued that schools are the essential work of children, and they should close only when absolutely necessary to prevent catastrophic overburdening of the health care system; they should be the last aspect of society to close and the first to reopen.1 Accordingly, some jurisdictions have elected to preferentially keep schools open throughout most, if not all, of the pandemic.1
As reported in a several systematic reviews, studies have reached mixed conclusions regarding the impact of school closure and subsequent reopening on the spread of COVID-19.2,3,5 This inconsistency of findings may be a result of the high risk for bias in most studies. For example, in many regions the initial decision to close schools for in-person instruction occurred concomitantly with the institution of other nonpharmaceutical interventions (such as masking and distancing), stay-at-home strategies, and economic closures, making it challenging to disentangle the individual impact of school closures.2 Further, many studies examining SARS-CoV-2 transmission in children have been limited to day care or summer camps and might not be generalizable to other settings.2 Understanding the impact associated with school reopening allows for the development of additional policies and mitigation strategies to reduce transmission risk.
On December 26, 2020, a provincewide shutdown order came into effect in Ontario, Canada.6 The many broad economic closures and social gathering restrictions included prohibiting nonessential travel, all in-person gatherings, in-person operation of nonessential businesses, and in-person school instruction. Throughout January and February 2021 a regionalized approach was taken to reopen schools for in-person instruction.7 Notably, all regions opted to wait until after schools reopened before relaxing other gathering restrictions. This offered a unique natural experiment to estimate what impact, if any, the reopening of schools had on community COVID-19 growth rates, particularly as the timing of reopening was exogenously determined (by provincial authorities) and was not exclusively based on epidemiological factors (that is, the timing of reopening decisions was, to some degree, random). Specifically, we assessed the percentage-point change in COVID-19 growth rates 1–5, 6–10, 11–15, 16–20, 21–25, and 26–30 days after schools reopened, relative to the four days before and the day of reopening. In addition, we controlled for other time-varying nonpharmaceutical intervention policies, daily testing rates, sociodemographic differences between communities, and other differences that may have influenced COVID-19 growth rates. We provide age-stratified estimates to assess potential differential changes in COVID-19 growth rates after school reopening.
Ontario is Canada’s most populous province, with approximately 14.5 million residents. For the 2020–21 school year, a mix of in-person and virtual learning options was offered to the more than two million students enrolled in public elementary and secondary schools.8 When and where in-person instruction was offered, in-person attendance was voluntary. Before the start of the 2020–21 school year, several policies were applied to mitigate COVID-19 transmission in schools, including mandatory masking, daily symptom checks, social distancing, and cohorting (see the online appendix for details).8,9 At the start of the 2020–21 school year face masks were required indoors for all students in grade 4 and above, as well as for staff and visitors. Also, on January 12, 2021, face masks became mandatory for students in grades 1–3 and for all students (grades 1–12) when outdoors.8
Polymerase chain reaction (PCR) diagnostic tests for SARS-CoV-2 completed during the study period (December 26, 2020–March 8, 2021, inclusive) were identified from linked laboratory data sources. These data sets captured more than 90 percent of tests completed in Ontario and 100 percent of confirmed cases.10 Universal screening testing was not conducted in the community or schools, but some asymptomatic testing initiatives occurred in a small number of schools (fewer than 1 percent), mostly in November and December 2020.11
We calculated the cumulative number of laboratory tests that were positive for SARS-CoV-2 for each of Ontario’s forty-nine census division areas; census divisions are provincially legislated counties, municipalities, or equivalent geographic areas, with an average population size of approximately 300,000 residents. School reopening dates were designated by public health units and were publicly reported.8 Except for those serving large, dense municipalities (for example, Toronto), most public health units contain multiple census divisions.
Additional details regarding the study data are in appendix A.9 The data included in this study were linked using unique encoded individual-level identifiers and analyzed at ICES (formerly the Institute for Clinical Evaluative Sciences).
We conducted an event study regression examining whether the reopening of elementary and secondary schools was associated with a change in community COVID-19 case growth rates. Specially, we focused on the period starting December 26, 2020 (that is, when the provincewide shutdown was enacted), through March 8, 2021 (that is, three weeks after schools in the final public health units were permitted to reopen), to provide sufficient time to observe changes in case growth rates after reopening.
Event study regression is similar to a difference-in-differences design, but instead of another comparison group, in our study the event study design allowed for the comparison of COVID-19 growth rates immediately before school reopening (across all communities) as the counterfactual. Importantly, this design allowed for the adjustment of confounders, including regional differences in testing, population demographics, and time-variant public health measures, as well as pre-event trends in case growth rates (that is, in the days and weeks preceding school reopening).
We estimated the impact on case growth using the approach described by Wei Lyu and George Wehby.12 Specifically, we estimated the impact on daily census division–level COVID-19 growth rates, defined as the difference in the natural log of the cumulative number of people testing positive for SARS-CoV-2 on a given day minus the natural log of the cumulative number of cases on the day before. We multiplied this value by 100 to provide percentage-point changes.
The reference period used in this analysis included the four days before school reopening through the day schools reopened (that is, days −4 to 0) for all census divisions included in the analysis. Relative to this period, we estimated how case growth rates changed during the following six periods: 1–5, 6–10, 11–15, 16–20, 21–25, and 26–30 days after schools reopened. Our modeling approach also considered pre-reopening trends by considering four windows before schools reopened: 5–9, 10–14, 15–19, and 20–30 days. Although some census divisions had longer pre- or postevent observation periods because of regional school closures, our primary analysis was restricted to the thirty days before and after reopening to ensure that our results were representative of the entire province and were most likely to reflect the effects associated with school reopening (that is, we focused on the most temporally proximate changes).
Census division–level changes in COVID-19 growth rates were estimated using ordinary least squares regression. Provincial effect estimates were calculated by weighting the census division–level estimates by the census division’s 2020 population. The 95% confidence intervals were estimated using heteroskedasticity robust standard errors, clustered by public health unit to account for the geographic level at which government-level policies, including school reopening and economic restrictions, were instituted. We further estimated the number of cases attributable to school reopening during the study period—that is, the counterfactual number of cases that would have happened had schools not reopened for in-person instruction. Two rural regions opted to close schools again on March 1, 2021,13,14 so we estimated the counterfactual number of cases up to February 28, 2021.
Several covariates were included in the final, adjusted event study regression to adjust for baseline trends in case growth. First, we included the number of days since the provincewide shutdown came into effect, as the number of confirmed cases dramatically declined after the introduction of this order (appendix B).9 Similarly, we included a term accounting for the growth rate of the cumulative number of SARS-CoV-2 tests completed in each census division, as case growth is highly correlated with testing. We further controlled for the timing and relaxation of commercial and other gathering restrictions in each public health unit (that is, according to Ontario’s color-coded reopening framework) and changes in census division workplace mobility (to reflect changes in behavior and movement).15 The model also controlled for several census division–level fixed effects describing regional sociodemographic differences, including population density, the average number of people living in a dwelling, and the proportion of the population working in an essential occupation (that is, for which working in person is required). Finally, the model controlled for the effects of weekends (given effects on testing availability and demand) and holidays or events in the five to ten days before (given impacts of large gatherings on transmission), including the weekends of New Year’s, Family Day (a provincewide civic holiday in February), and the NFL’s Super Bowl Sunday.
All analyses were performed using SAS, version 9.4. The use of data in this project was authorized under Section 45 of Ontario’s Personal Health Information Protection Act and thus was exempt from research ethics board review. ICES is an independent, nonprofit research institute whose legal status under Ontario’s health information privacy law allows it to collect and analyze health care data, without consent, for health system evaluation.
We planned a priori to assess the potential for a differential impact of school reopening on COVID-19 case growth rates by age. We stratified the main adjusted event regression model according to the age of confirmed cases, and we estimated the above adjusted model coefficients for elementary school–age children (younger than age 14); secondary school–age adolescents (ages 14–18); younger (ages 19–44) and older (ages 45–64) adults, both of which would include parents, and possibly grandparents, of school-age children; and seniors (ages 65 and older).
To assess the potential impact of heterogeneities in case growth patterns, testing availability, sociodemographics, and so on at smaller-scale geographies, we performed the event study regression at the level of the dissemination area, Canada’s smallest census area (approximately 400–700 people). Notably, zero new cases were reported on more than 90 percent of dissemination area–days during the study period, so our primary analysis relied on the larger geographic level of census divisions to offer more stable estimates.
The first SARS-CoV-2 variant of concern, Alpha, was identified in Ontario December 26, 2020; we examined the potential for differential growth rates, transmissibility, and epidemiological patterns.
Throughout the pandemic, Canada introduced a number of travel-related restrictions. Although the broadest such restrictions were in place before the study period, several new measures were introduced during the study period in response to variants of concern. To consider their potential impact, we included a daily provincewide indicator term describing the number of travel measures in place.
Other checks included considering community transmission levels in the two weeks before school reopening, based on the US Centers for Disease Control and Prevention’s school risk–level thresholds, as community transmission levels may have informed reopening timelines; excluding testing outlier days (that is, above the ninetieth or below the tenth percentile); considering differences in school-based symptom testing guidance; considering the delivery mode for secondary school classes (that is, in-person delivery only or mixed in-person and virtual delivery); one-, two-, three-, four-, and five-day lags in case and test growth rates; and considering additional holidays; see appendix C.9
As with other empirical studies of COVID-19 cases, this analysis was limited to confirmed cases, which yielded an underestimate of incident SARS-CoV-2 infections. This is particularly true for pediatric populations, for whom testing has been historically low, given that many are asymptomatic or mildly symptomatic.2,3 Furthermore, differences in testing propensity when students are at home versus in school, where testing may be required for return, may exacerbate differences in diagnostic detection. Moreover, a strength of event study regression, over traditional difference-in-differences designs, is the ability to simultaneously control for multiple time-variant and -invariant covariates, although the complexity and limited use of these models, particularly with respect to health outcomes, may result in more difficulty assessing the appropriateness of model specifications and interpretation of model coefficients. Our findings may be sensitive to these considerations, particularly the selection of specific confounders; however, we conducted a variety of robustness checks to assess the consistency of our findings over varying model specifications.
In addition, data regarding several important factors were limited. These include the incidence of variants of concern, in-person attendance rates, identifiers to link individual children to school boards (districts), and specific details regarding the behavior changes and sources of infection for school-age children and other household members. With these limited data, there is the potential for residual confounding. Notably, our age-stratified analysis, inclusion of multiple windows of time after reopening, and inclusion of changes in workplace mobility may offer some insights into the likely directionality of transmission, assuming that no changes in testing propensity occurred after the reopening of schools. In addition, in the weeks after the last region’s school reopening, there were numerous influential factors for which data were scarce, if available at all (for example, detailed information on the circulation of variants of concern and vaccination rates). Our study period thus was limited to a period with sufficient data to control for underlying case trends and for which pre- versus post-reopening comparisons would be most reliable. However, this was also a period marked by limited community transmission, thereby reducing the generalizability of these findings to contextually different pandemic waves.
Confirmed COVID-19 cases during the second wave of the pandemic in Ontario peaked in early January 2021, shortly after the shutdown order, and exhibited a steady decline throughout January and February (exhibit 1). Among the cases reported during the study period, 28.4 percent resided in Toronto, Ontario’s largest city, and another 18.2 percent and 10.4 percent, respectively, occurred in the adjacent, densely populated regions of Peel and York (data not shown); these were the last three public health units to reopen schools.
Exhibit 1 Absolute cumulative number and growth rate of confirmed COVID-19 cases in Ontario, Canada, December 26, 2020–March 8, 2021
Exhibit 2 illustrates the estimated changes in the daily growth rate of confirmed COVID-19 cases in the days preceding and following the reopening of schools, as obtained from the fully adjusted regression model using daily census division–level data from December 26, 2020, through March 8, 2021; model coefficients are in appendix B.9 We estimated a small increase in daily COVID-19 case growth rates after school reopening, relative to the four days before and the day of reopening (that is, days −4 to 0). Although there was evidence to suggest that the magnitude of this effect increased over time, no estimate was statistically significant. Specifically, after we controlled for myriad factors related to census division–level differences in daily census division–level COVID-19 case growth rates, we estimated increases of 0.07 percentage points (95% CI: −0.07, 0.21), 0.08 percentage points (95% CI: −0.09, 0.25), 0.07 percentage points (95% CI: −0.13, 0.27), and 0.13 percentage points (95% CI: −0.15, 0.41) in COVID-19 case growth in the 11–15, 16–20, 21–25, and 26–30 days, respectively, after school reopening. In contrast, a small decline in daily case growth rates was observed in the first ten days after reopening, relative to the four days before and the day of reopening: −0.09 (95% CI: −0.14, −0.03) and −0.02 (95% CI: −0.13, 0.09), respectively, 1–5 and 6–10 days after reopening.
Exhibit 2 Event study estimates of the changes in the daily census division–level growth rate of COVID-19 cases in the days before and after the reopening of schools in Ontario, Canada, December 26, 2020–March 8, 2021
In the absence of school reopening, we estimated that 213 (95% CI: −256, 672) fewer cases of COVID-19 would have occurred in Ontario between December 26, 2020, and February 28, 2021—that is, roughly 0.08 percent fewer cases than were observed (data not shown).
In age-stratified analyses (exhibit 3), we estimated that school reopening was followed shortly thereafter by a large increase in case growth rates in elementary school–age children (younger than age 14). In contrast, among older adults (ages 45–64) and seniors (ages 65 and older), growth rates were lower and lagged reopening. There were no discernable differences in case growth rates among adults ages 19–44 after the reopening of schools, and we observed a decline in case growth rates among adolescents ages 14–18.
Exhibit 3 Event study estimates of the effects of time since school reopening on the age-specific daily census division–level growth rate of COVID-19 cases in Ontario, Canada, December 26, 2020–March 8, 2021
We implemented multiple iterations of the main event study model to examine the robustness of our estimates to different model specifications and sampling choices. A detailed description of these robustness checks is in appendix C.9 The results were broadly comparable across these checks; that is, the increases in case growth rates after the reopening of schools were minimal and lacked statistical significance. For example, an informative result from our check using single-day effect estimates (that is, daily versus five-day bins) is that the apparent downward pre-reopening trend was concentrated in the earlier study period and largely disappeared in the four days before schools reopened. This increased our confidence that the failure to find a significant effect was not a result of residual confounding from pre trends. This is similarly demonstrated in appendix exhibit C3,9 which shows no discernible pattern in regional case growth rates, aside from weekday differences, before reopening; this further suggests that the decision to reopen schools was not exclusively based on a pattern of declining community incidence in the days before.
We also considered the impact of one-to-five-day lags on case (and testing) growth rates to account for reporting delays and the latency period of SARS-CoV-2 infection. In this analysis, much larger effect estimates were observed, with the effect of school reopening on case growth rates being magnitudes greater when longer lag periods were considered. When a five-day lag period was considered—approximately the median latency period for SARS-CoV-2 and reflective of the date of transmission—we estimated increases of 0.08 percentage points (95% CI: −0.04, 0.20), 0.18 percentage points (95% CI: 0.01, 0.34), 0.20 percentage points (95% CI: 0.01, 0.39), 0.20 percentage points (95% CI: −0.02, 0.42), and 0.27 percentage points (95% CI: −0.03, 0.57) in COVID-19 growth rates 1–5, 6–10, 11–15, 16–20, and 21–25 days, respectively, after reopening (appendix exhibit C1).9 This further increased our confidence in some causal link between school reopening and an increase in COVID-19 case growth rates, as this suggests that a larger effect would be estimated if the date of transmission could be determined and included in the analysis.
Given the critical role that schools play for children, families, and communities, and the currently limited understanding of the long-term health effects associated with COVID-19, the decision to close schools as a part of broader COVID-19 mitigation strategies has been a fervently debated topic. This study built on a growing literature showing that school reopening may lead to an increase in community COVID-19 growth rates.2,3 Specifically, we estimated a marginal increase in cases shortly after the resumption of in-person elementary and secondary school classes in Ontario, even within the context of the multiple layers of nonpharmaceutical interventions that were in place in schools (namely, masking, cohorting, and symptom screening) and the broader community (appendix exhibits D2 and D3) during a time in which variants of concern were not prevalent (appendix exhibit D4) and COVID-19 incidence was at a low point relative to other periods in the pandemic (appendix exhibit D5).9 However, there was variability across schools in the degree to which mitigation measures (for example, ventilation, class sizes, and asymptomatic and rapid testing) were implemented.
Our analysis suggests that an increase in cases should be expected in the weeks after the reopening of schools and that these effects may be seen first within pediatric populations, followed shortly thereafter by adult populations. These results align with a recent decision analytical modeling study of the Ontario population, which similarly reported schools playing a minor role in the transmission of SARS-CoV-2 relative to broader community transmission and predicted no noticeable increase in cases after school reopening if other nonpharmaceutical interventions were in place.16 Specifically, the authors estimated that only sixty-six additional cases occurred in Ontario between September 1 and October 31, 2020, as a result of schools being open (compared with a simulated counterfactual in which schools remained closed but all other community nonpharmaceutical interventions remained in place).
Imperatively, these results suggest that any increases in case growth after the reopening of schools may be manageable with appropriate mitigation policies, such as reducing contacts through smaller class sizes, masking, and cohorting,17,18 as well as prioritizing school reopening over other activities that increase contact in a community, such as social gatherings, recreational events, and in-person work.1,3,5 Critically, all public health units in Ontario did not relax gathering restrictions until after schools had reopened. Although these results may suggest that children being in schools was a source of increased transmission into homes and the broader community, an examination of workplace mobility patterns (appendix exhibit B1)9 shows that a sharp increase in workplace mobility also followed the reopening of schools, possibly a result of parents and school staff being able to return to work, which may similarly increase community transmission. As argued by others, if schools are deemed the essential work of children and policy makers wish to minimize disruptions to childhood education, these results support the notions that schools should be the last to close and first to reopen and that efforts to reduce mobility and social interactions should be maintained to minimize transmission and interruptions to in-person learning.1,5
Although we estimated that there would be small increases in case growth rates in the month after school reopening, these effects grew over time and were substantially larger in magnitude when a latency lag period was considered. Moreover, these results were qualitatively robust to a range of robustness checks, further suggesting that school reopening led to a small increase in COVID-19 cases. This underscores the crucial importance of considering the exponential nature of growth rates and the critical need to implement other mitigation measures to keep community transmission as low as possible before and after the reopening of schools. As seen after the introduction of Ontario’s shutdown order, such policies can quickly and drastically reduce transmission; this study was unable to ascertain whether these measures alone, without the need to also close schools, could have reduced case growth rates as effectively. We noted a differential pattern in case growth patterns after the reopening of schools for secondary school–age youth. As shown in other studies, contact patterns and behaviors among adolescents differ from those of younger children (for example, younger children require closer and more prolonged contact with educators, whereas older children and adolescents tend to engage in more social contacts outside of school).4 These differences have implications for determining the most impactful school policies, such as mask mandates, cohorting, and testing approaches (for example, test-to-stay), as well as the trade-offs between virtual and in-person learning.
Critically, this study period was during the second pandemic wave in Ontario, when transmission was low; the circulation of variants of concern was limited; COVID-19 vaccines were restricted predominantly to front-line health care workers and residents of congregate living settings, such as long-term care facilities; and multiple other nonpharmaceutical interventions were in place in schools and the broader community. Thus, these results might not be generalizable to other populations and pandemic waves—namely, those characterized by new variants of concern, which may be more transmissible and differentially affect younger populations, and a heavier burden of illness among unvaccinated people, including children. As the epidemiology of SARS-CoV-2 changes and as vaccines are rolled out more broadly, it is possible that the burden of infection may shift to different populations, which may affect the role that schools have in the chain of transmission.3
Notably, recent pandemic waves have been marked by the predominant circulation of highly transmissible Omicron subvariants. In response to a surge in hospitalizations in December 2021, Ontario reinstated additional community-based measures, including school closures, work-from-home policies, restaurant closures, and gathering restrictions.19,20 However, unlike in prior waves, these closures were time limited and purpose driven; in the case of schools, a two-week closure occurred after winter break (that is, January 3–17, 2022) to provide time to enhance mitigation efforts, such as the distribution of high-efficiency particulate air filters, higher-quality masks, and rapid and at-home diagnostic tests, and also prioritized vaccination for students ages 5–11 (primary doses) and educators (third doses).20 Impressively, at the time schools reopened, 82 percent of Ontarians ages five and older had received two vaccine doses, and among Ontarians ages twelve and older, coverage was 89 percent.21 Moreover, the reopening of schools preceded the gradual relaxation of other restrictions (starting January 31, 2022), excluding masking (March 21, 2022) and proof-of-vaccination policies (March 1, 2022).22 Although the substantial December 2021–January 2022 Omicron wave has subsided, the impact of the emerging BA.2 wave in Ontario had yet to be realized as of early April 2022. Current evidence (based on testing, wastewater, workplace surveillance, and other indicators) suggests that the province was experiencing a period of rapidly accelerating case growth in early April, beginning shortly after the relaxation of mask mandates, with a disproportionate number of cases (and hospitalizations) occurring in children.21,23 This underscores the critical role that community-based measures play in mitigating SARS-CoV-2 transmission, and it anecdotally reinforces the conclusions of our study suggesting that schools can reopen without a substantial increase in community transmission if adequate, layered mitigation measures are in place and the opening of schools is prioritized before the relaxation of other community-based measures.
This study examined a unique school reopening situation in a large Canadian province, allowing estimation of the impact of reopening on community COVID-19 cases. These results offer a realistic assessment of a manageable increase in the transmission of SARS-CoV-2 in the month after the reopening of schools in the context of wider societal interventions and school-based nonpharmaceutical intervention policies. Future waves of this pandemic must prioritize interventions and policies that reduce transmission in schools to reduce the broader community impacts of school closures.
This study was supported by ICES (formerly the Institute for Clinical Evaluative Sciences), which is funded by the Ontario Ministries of Health and Long-Term Care, as well as the Ontario Health Data Platform, a Province of Ontario initiative to support Ontario’s ongoing response to COVID-19 and its related impacts. Funding for the analyses was provided by an unrestricted grant to Astrid Guttmann by the SickKids Research Institute. The opinions, results, and conclusions are those of the authors and are independent from the funding source. No endorsement by ICES, the Ontario Ministries of Health and Long-Term Care, the Ontario Health Data Platform and its partners, or the Province of Ontario is intended or should be inferred. Parts of this report are based on data and information compiled and provided by the Ontario Ministry of Health, the Canadian Institute for Health Information, and Public Health Ontario. The authors acknowledge Public Health Ontario for access to case-level data from iPHIS/CCM Plus and COVID-19 laboratory data, as well as assistance with data interpretation. They also thank the staff of Ontario’s public health units, who are responsible for COVID-19 case and contact management and data collection within iPHIS/CCM Plus. Geographical data are adapted from Statistics Canada, Postal Code Conversation File + 2016 (Version 7B). This does not constitute endorsement by Statistics Canada of this project. The authors also thank Avi Goldfarb for his helpful comments and review of the manuscript. Laura Rosella and Astrid Guttman are joint senior authors.

Article Metrics

Published online 6 June 2022

© 2022 Project HOPE—The People-to-People Health Foundation, Inc.
This study was supported by ICES (formerly the Institute for Clinical Evaluative Sciences), which is funded by the Ontario Ministries of Health and Long-Term Care, as well as the Ontario Health Data Platform, a Province of Ontario initiative to support Ontario’s ongoing response to COVID-19 and its related impacts. Funding for the analyses was provided by an unrestricted grant to Astrid Guttmann by the SickKids Research Institute. The opinions, results, and conclusions are those of the authors and are independent from the funding source. No endorsement by ICES, the Ontario Ministries of Health and Long-Term Care, the Ontario Health Data Platform and its partners, or the Province of Ontario is intended or should be inferred. Parts of this report are based on data and information compiled and provided by the Ontario Ministry of Health, the Canadian Institute for Health Information, and Public Health Ontario. The authors acknowledge Public Health Ontario for access to case-level data from iPHIS/CCM Plus and COVID-19 laboratory data, as well as assistance with data interpretation. They also thank the staff of Ontario’s public health units, who are responsible for COVID-19 case and contact management and data collection within iPHIS/CCM Plus. Geographical data are adapted from Statistics Canada, Postal Code Conversation File + 2016 (Version 7B). This does not constitute endorsement by Statistics Canada of this project. The authors also thank Avi Goldfarb for his helpful comments and review of the manuscript. Laura Rosella and Astrid Guttman are joint senior authors.

source

Leave a Comment

%d bloggers like this: