The DOI matches its Crossref record; 59 of the 60 references it cites were found in a registry or an index; 41 of the 59 sources its claims cite were checked for support.
Social inequalities in COVID-19 death by area-level income in 11.2 million people in Ontario, Canada: patterns over time and the mediating role of vaccination.
Open accessPublished version of a preprint
Abstract
Knowledge of patterns in COVID-19 deaths by area-level income over time and the mediating role of vaccination in inequality patterns remains limited. We used data from a population-based retrospective cohort of 11 248 572 adults in Ontario, Canada. Cause-specific hazard models were used examine the relationship between income (2016 Census at the dissemination area level) and COVID-19 deaths between March 1, 2020 and January 30, 2022, stratified by wave. We used regression-based causal mediation analyses to examine the mediating role of vaccination in the relationship between income and COVID-19 deaths during waves 4 and 5. After accounting for demographics, baseline health, and other social determinants of health, inequalities in COVID-19 deaths by income persisted over time (HR [95% CI] comparing lowest income vs highest income quintiles were 1.37 [0.98-1.92] for wave 1, 1.21 [0.99-1.48] for wave 2, 1.55 [1.22-1.96] for wave 3, and 1.57 [1.15-2.15] for waves 4 and 5). By the start of wave 4, 7 534 259 (67.7%) of those alive were vaccinated, with lower odds of vaccination in the lowest income vs highest income quintiles (0.71 [0.70-0.71]). This inequality in vaccination accounted for 56.9% [22.5%-91.3%] of inequalities in COVID-19 deaths between individuals in the lowest income vs highest income quintiles. Efforts are needed to address vaccination gaps and residual heightened risks associated with lower income to improve health equity in COVID-19 outcomes.
Overview
CoChat Summary
TL;DR
In a population-based cohort of 11.2 million Ontario adults followed across five pandemic waves (March 2020–January 2022), COVID-19 deaths remained higher in lower-income neighbourhoods after adjustment for demographics, baseline health, and area-level social determinants of health. During the later waves, 57.9% of the income inequality in COVID-19 deaths was attributable to differential vaccination coverage by income.
Key findings
- Of 11,248,572 adults, 7,044 (0.063%) experienced a COVID-19 death; deaths were concentrated in the lowest- vs. highest-income quintile (0.097% vs. 0.038%).
- Adjusted hazard ratios for COVID-19 death (lowest vs. highest income quintile) were 1.37 [0.98–1.92] in wave one, 1.21 [0.99–1.48] in wave two, 1.55 [1.22–1.96] in wave three, and 1.57 [1.15–2.15] in waves four and five, with a significant increasing trend over time (p=0.02).
- By the start of wave four, 7,534,259 (67.7%) of 11,122,816 adults were vaccinated; the lowest-income quintile had lower odds of vaccination than the highest (aOR 0.71 [0.70–0.71]; 59.1% vs. 74.1% coverage).
- Vaccination mediated 57.9% [21.9%–94.0%] of the income inequality in COVID-19 deaths during waves four and five; the direct effect aHR was 1.22 [0.88–1.68] and the indirect effect aHR 1.24 [1.24–1.25]. Sensitivity analyses attributed 29.8% [12.3%–?] to area-level essential work and 8.7% [−3.4%–20.8%] to high-density housing.
Methods
- Population-based retrospective cohort of community-dwelling adults aged ≥18 in Ontario as of March 1, 2020, followed to January 30, 2022, using linked administrative health data at ICES; exclusions included long-term care residents, age >105, non-Ontario postal codes, and missing covariates.
- Exposure: area-level median household income from the 2016 Census at the dissemination-area level, ranked within city and categorized into quintiles. Outcome: COVID-19 death (death within 30 days after or 7 days before a positive SARS-CoV-2 PCR test). Mediator: vaccination (≥1 dose Johnson & Johnson or ≥2 doses of other vaccines).
- Cause-specific hazard models (SAS 9.4) per wave with competing risks, adjusted for demographics, baseline health, and area-level SDOH, with robust sandwich covariance for DA clustering; regression-based causal mediation analysis (R 4.1.2, CMAverse) for waves four and five, plus E-value bias analysis.
Limitations
- Stated by the authors: area-level income was used in the absence of individual-level measures, which may underestimate inequalities; individuals without provincial health insurance were not captured; time-dependent vaccination status was not modeled (though the income–vaccination gap at wave five was similar to wave four); vaccine waning was not considered, which may underestimate the mediating role of vaccination; individual-level SDOH (occupation, housing, racialized/Indigenous identity) and contact/masking exposures were unavailable.
- Evident from the text: wave one and wave two adjusted hazard ratios for the lowest vs. highest income quintile had confidence intervals crossing 1, so the early-wave inequality estimates are imprecise; waves four and five were combined to increase power given small death counts; the mediation estimate for essential work has a malformed upper confidence bound in the provided text.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
18 Valid · 6 Not valid · 4 Not checked · 59 Own results and interpretations
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Limitations include our use of a area-level income measure in the absence of individual-level measures, which might result in an underestimation of social inequalities(65).”
What the sources say
- Partly supportedcites [65] Rao, 2021, A disproportionate epidemic: COVID-19 cases and deaths among essential workers in Toronto, Canada, Ann Epidemiol. sourceExperimental — accuracy not yet measured
The Methods describe neighbourhood-level census attributes, and the Conclusions identify ecological-level measures as a limitation because they cannot establish individual-level associations. The paper does not say that using area-level income without individual-level income might underestimate social inequalities; that proposed consequence is unsupported. (LLM, conf=0.83)
hedgedstates a possible cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Rests on an assumption not shown here.
- Partly supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“save more lives(45,50)”
What the sources say
- Not checked — no full text of the sourcecites [45] Dahabreh, 2024, Causal inference about the effects of interventions from observational studies in medical journals, JAMA. source
- Partly supportedcites [50] Ortiz, 2024, Socio-demographic determinants of COVID-19 vaccine uptake in Ontario: exploring differences across the health region model, Vaccine. sourceExperimental — accuracy not yet measured
The Introduction (p. 1) reports that COVID-19 vaccines have prevented millions of deaths, and the Conclusion (p. 9) calls for improving vaccine access for marginalized communities. The paper does not establish the comparative claim that equitable coverage saves more lives than less equitable coverage. (LLM, conf=0.82)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Date of death was determined using CCM and RPDB, which capture over 99.3 % of COVID-19 deaths in Ontario(3).”
What the sources say
- Partly supportedcites [3] Wang, 2022, Differential patterns by area-level social determinants of health in coronavirus disease 2019 (COVID-19)–related mortality and non–COVID-19 mortality: a population-based study of 11.8 million people in Ontario, Canada, Clin Infect Dis sourceExperimental — accuracy not yet measured
The Methods > Outcomes passage says the date of death was determined using CCM and RPDB. It estimates that the two databases capture 99.3% of COVID-19–related deaths, whereas the claim states as fact that they capture over 99.3%. (LLM, conf=0.87)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
The cited sources that could be checked couldn’t confirm or dispute this claim.
In the paper
“Our findings further support the potential of effect modification by vaccination on social inequalities in COVID-19 deaths, similar to findings from the previous ecological study in the UK(19).”
What the sources say
- Could not verifycites [19] Benchimol, 2015, The reporting of studies conducted using observational routinely-collected health data (RECORD) statement, PLoS Med. sourceExperimental — accuracy not yet measured
The RECORD statement’s Discussion (Limitations; page unavailable) addresses confounding in studies using routinely collected data, not whether vaccination modifies social inequalities in COVID-19 deaths. Searches for that finding and for a UK ecological study found no relevant passage in our copy. (LLM, conf=0.78)
hedgedstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: unclear — It is unclear why the paper says it holds.
- Could not verify
The cited sources that could be checked couldn’t confirm or dispute this claim.
In the paper
“Individuals living in lower-income neighbourhoods are more likely to reside in high density housing with multi-generational households”
What the sources say
- Could not verifycites [59] Cénat, 2022, A systematic review on vaccine hesitancy in black communities in Canada: critical issues and research failures, Vaccine. sourceExperimental — accuracy not yet measured
The Discussion discusses socioeconomic status and mentions household size as a predictor of vaccine hesitancy, but does not link lower-income neighbourhoods to housing density or multigenerational households. Searches for high-density housing, multigenerational households, and related housing terms found no passage addressing that link. (LLM, conf=0.72)
- Could not verifycites [60] Lavoie, 2022, Understanding national trends in COVID-19 vaccine hesitancy in Canada: results from five sequential cross-sectional representative surveys spanning April 2020–March 2021, BMJ Open. sourceExperimental — accuracy not yet measured
The Results section on sociodemographic predictors and the Discussion section on vaccine-hesitant Canadians address household income in relation to vaccine hesitancy, not housing conditions. Searches for lower-income neighbourhoods, housing density and multigenerational households found no passage addressing the claimed relationship; page numbers were not available in our copy. (LLM, conf=0.75)
hedgedstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Could not verify
46 more claims · 5 with their cited sources checked
Create a free account to see all 87 claims and to verify the 3 references still open.
- Probably not a substantial claim. 25
- It may state two separate claims as one. 12
- The quote only partly states this claim. 3
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated major: it would touch a result — a value a result rests on, reported and recomputed p-values on opposite sides of the significance level, two passages that conflict with no sign which is right, or a cited work retracted before this paper was published.
In the paper
§ Discussion
“| Prior infectoion n | 1121 (0.052%) | 901 (0.041%) | 854 (0.038%) | 751 (0.033%) | 682 (0.030%) |”
§ Results
“of whom, 30,356; 191,945; 176,438; 214,599 tested positive for SARS-CoV-2, resulting in 955, 2860, 2185, and 1044 COVID-19 deaths, during wave one, wave two, wave three, and waves four&five, respectively.”
Rated major: it would touch a result — a value a result rests on, reported and recomputed p-values on opposite sides of the significance level, two passages that conflict with no sign which is right, or a cited work retracted before this paper was published.
In the paper
§ Mediating and effect modifying role of vaccination during waves four&five
“The magnitude of association between income and COVID-19 deaths appeared to be slightly larger for those with than without vaccination”
§ Discussion
“Smaller magnitude of inequality in COVID-19 deaths by income among those vaccinated than non-vaccinated”
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ Introduction
“by the start of wave four (August 1, 2020)”
§ Discussion
“Vaccination status by the start of wave 4 (August 1, 2021)”
Citations
Cites
Factors associated with COVID-19-related death using OpenSAFELY
Williamson, E. J., Walker, A. J., Bhaskaran, K., Bacon, S., Bates, C., Morton, C. E., Curtis, H. J., Mehrkar, A., Evans, D., Inglesby, P., et al. (2020) · Nature · doi:10.1038/s41586-020-2521-4 · Cited by 6,674
Sensitivity analysis in observational research: Introducing the E-value
VanderWeele, T. J., & Ding, P. (2017) · Annals of Internal Medicine · doi:10.7326/m16-2607 · Cited by 6,350
Benchimol, E. I., Smeeth, L., Guttmann, A., Harron, K., Moher, D., Petersen, I., Sørensen, H. T., von Elm, E., Langan, S. M., & RECORD Working Committee (2015) · PLOS Medicine · doi:10.1371/journal.pmed.1001885 · Cited by 5,749
The Robust Inference for the Cox Proportional Hazards Model
Lin, D. Y., & Wei, L. J. (1989) · Journal of the American Statistical Association · doi:10.1080/01621459.1989.10478874 · Cited by 2,695
Introduction to the Analysis of Survival Data in the Presence of Competing Risks
Austin, P. C., Lee, D. S., & Fine, J. P. (2016) · Circulation · doi:10.1161/circulationaha.115.017719 · Cited by 2,653
Self-citationRobins, J. (1986) · Mathematical Modelling · doi:10.1016/0270-0255(86)90088-6 · Cited by 2,451
Socioeconomic Status in Health Research: One Size Does Not Fit All
Braveman, P. A., Cubbin, C., Egerter, S., Chideya, S., Marchi, K. S., Metzler, M., & Posner, S. (2005) · JAMA · doi:10.1001/jama.294.22.2879 · Cited by 2,242
Valeri, L., & VanderWeele, T. J. (2013) · Psychological Methods · doi:10.1037/a0031034 · Cited by 2,132
Global impact of the first year of COVID-19 vaccination: a mathematical modelling study
Watson, O. J., Barnsley, G., Toor, J., Hogan, A. B., Winskill, P., & Ghani, A. C. (2022) · The Lancet Infectious Diseases · doi:10.1016/s1473-3099(22)00320-6 · Cited by 1,769
Competing Risk Regression Models for Epidemiologic Data
Lau, B., Cole, S. R., & Gange, S. J. (2009) · American Journal of Epidemiology · doi:10.1093/aje/kwp107 · Cited by 1,403
Research works printed but not linked: 9
- Finch, , 2022, The Continuing Impact of COVID-19 on Health and Inequalities
- Oxfam International, 2022, Pandemic of Greed
- Gill, 2022, Why Global Vaccine Equity is the Prescription for a Full Recovery
- Marfo, 2022, Beyond Vaccine Hesitancy: Understanding Systemic Barriers to Getting Vaccinated
- Statistics Canada, 2017, Dictionary, census of population, 2016-dissemination area (DA)
- Canadian Institute for Health Information, 2018, Measuring health inequalities: a toolkit—area-level equity stratifiers using PCCF and PCCF+
- Ontario Data Catalogue, 2017, Permanent resident data (immigration, refugees and citizenship Canada)
- SAS Institute, 2013, SAS/ACCESS® 9.4 Interface to ADABAS
- R Core Team, 2022, R: A Language and Environment for Statistical Computing
Cited by
Sapin, M. E., Maxwell, C. J., Clarke, A. E., Cooper, C., So, M., Schwartz, K. L., Daneman, N., Mishra, S., & MacFadden, D. R. (2025) · Clinical Microbiology and Infection · doi:10.1016/j.cmi.2025.02.025 · Cited by 3
Self-citationWang, L., Swayze, S., Siddiqi, A., Baral, S., Sander, B., Sbihi, H., Kwong, J. C., & Mishra, S. (2026) · American Journal of Epidemiology · doi:10.1093/aje/kwag047 · Cited by 1
Self-citationDeterminants of COVID-19 mortality among the US law enforcement officers
Okunade, A. A., Olarewaju, F. O., Oyemakinde, B., & Lubiani, G. G. (2025) · Journal of Population Research · doi:10.1007/s12546-025-09393-y
Guṭu, R., Schaps, V., Wachtler, B., Beese, F., Hoebel, J., Alibone, M., & Wahrendorf, M. (2025) · Scandinavian Journal of Work Environment & Health · doi:10.5271/sjweh.4242 · Cited by 1
Metadata
- Authors
- Wang, L., Swayze, S., Bodner, K., Calzavara, A., Harrigan, S. P., Siddiqi, A., Baral, S. D., Austin, P. C., Smylie, J., Koh, M., Sbihi, H., Sander, B., Kwong, J. C., & Mishra, S.
- Affiliations
- St. Michael's Hospital (CA)
- Institute for Clinical Evaluative Sciences (CA)
- University of British Columbia (CA)
- University of North Carolina at Chapel Hill (US)
- Johns Hopkins University (US)
- Sunnybrook Health Science Centre (CA)
- University of Toronto (CA)
- University Health Network (CA)
- Corresponding author
- Mishra, Sharmistha
- ORCID
- Venue
- American Journal of Epidemiology
- Publisher
- Oxford University Press (OUP)
- Published
- 2025-03-07
- Volume
- 195
- Issue
- 2
- Pages
- 379-390
- ISSN
- 0002-9262 (ISSN-L)
- 1476-6256
- OpenAlex
- W4408342407
- Cited by
- 5
- References
- 50
- Citations by year
- 2026: 2 · 2025: 3
- FWCI
- 2.77
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- Locations
- American Journal of Epidemiology (publishedVersion, cc-by-nc)
- medRxiv (submittedVersion, cc-by-nc-nd)
- PubMed (publishedVersion)
- Funders
- Canada Research Chairs (CRC-950-232429, CRC-950-232643)
- Canadian Institutes of Health Research (GA1-177697, VS1-175536, VR5-172683)
- Topics
- Physical Sciences › Mathematics › Modeling and Simulation › COVID-19 epidemiological studies
- Social Sciences › Health › Vaccine Coverage and Hesitancy
- Social Sciences › Health › Health disparities and outcomes
- Keywords
- COVID-19 mortality, health disparities, socioeconomic status, COVID-19 vaccination, vaccine inequity, mediation analysis, social determinants of health, pandemic waves, Ontario
- MeSH terms
- Mediation Analysis, COVID-19, SARS-CoV-2, COVID-19 Vaccines, Socioeconomic Disparities in Health, Adult, Aged, Aged, 80 and over, Female, Humans, Income, Male, Middle Aged, Ontario, Registries, Retrospective Studies, Time Factors, Vaccination, Proportional Hazards Models, Young Adult, Social Determinants of Health
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-30, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-30 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Social inequalities in COVID-19 death by area-level income in 11.2 million people in Ontario, Canada: patterns over time and the mediating role of vaccination. https://app.cochat.ai/papers/10.1093/aje/kwaf051
- The paper
- Wang, L., Swayze, S., Bodner, K., Calzavara, A., Harrigan, S. P., Siddiqi, A., Baral, S. D., Austin, P. C., Smylie, J., Koh, M., Sbihi, H., Sander, B., Kwong, J. C., & Mishra, S. (2026). Social inequalities in COVID-19 death by area-level income in 11.2 million people in Ontario, Canada: patterns over time and the mediating role of vaccination. American Journal of Epidemiology. https://doi.org/10.1093/aje/kwaf051
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.