Neighbourhood characteristics associated with the geographic variation in laboratory confirmed COVID-19 in Ontario, Canada: a multilevel analysis

  1. Tristan Watson
  2. Jeffrey C. Kwong
  3. Kathy Kornas
  4. Sharmistha Mishra
  5. Laura C. Rosella

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

There is limited information on the role of individual- and neighbourhood-level characteristics in explaining the geographic variation in the novel coronavirus 2019 (COVID-19) between regions. This study quantified the magnitude of the variation in COVID-19 rates between neighbourhoods in Ontario, Canada, and examined the extent to which neighbourhood-level differences are explained by census-based neighbourhood measures, after adjusting for individual-level covariates (i.e., age, sex, and chronic conditions).

Methods

We conducted a multilevel population-based study of individuals nested within neighbourhoods. COVID-19 laboratory testing data were obtained from a centralized laboratory database and linked to health-administrative data. The median rate ratio and the variance partition coefficient were used to quantify the magnitude of the neighbourhood-level characteristics on the variation of COVID-19 rates.

Results

The unadjusted median rate ratio for the between-neighbourhood variation in COVID-19 was 2.22. In the fully adjusted regression models, the individual- and neighbourhood-level covariates accounted for about 44% of the variation in COVID-19 between neighbourhoods, with 43% attributable to neighbourhood-level census-based characteristics.

Conclusion

Neighbourhood-level characteristics could explain almost half of the observed geographic variation in COVID-19. Understanding how neighbourhood-level characteristics influence COVID-19 rates can support jurisdictions in creating effective and equitable intervention strategies.

Article activity feed

  1. ScreenIT

    SciScore for 10.1101/2021.04.06.21254988: (What is this?)

    Please note, not all rigor criteria are appropriate for all manuscripts.

    Table 1: Rigor

    Institutional Review Board StatementConsent: ICES is an independent, non-profit research institute whose legal status under Ontario’s health information privacy law allows it to collect and analyze health care and demographic data, without consent, for health system evaluation and improvement.
    RandomizationSecond, a multilevel Poisson regression model was fit (model 2) with neighbourhood-specific random intercepts that added seven individual-level variables: age, sex, asthma, diabetes, hypertension, CHF, and COPD.
    Blindingnot detected.
    Power Analysisnot detected.
    Sex as a biological variablenot detected.

    Table 2: Resources

    Software and Algorithms
    SentencesResources
    SAS Enterprise Guide v.8.15 (SAS Institute Inc, Cary, NC) was used to create the data set and conduct the multilevel regression analyses.
    SAS Institute
    suggested: (Statistical Analysis System, RRID:SCR_008567)
    Stata/MP v15 (StataCorp, College Station, TX) was used to conduct the ICC-VPC analyses.
    StataCorp
    suggested: (Stata, RRID:SCR_012763)

    Results from OddPub: We did not detect open data. We also did not detect open code. Researchers are encouraged to share open data when possible (see Nature blog).

    Results from LimitationRecognizer: We detected the following sentences addressing limitations in the study:
    Limitations include the potential for misclassification of neighbourhood measures as some neighbourhood features might have changed since 2016. The people with positive tests in our cohort may not be representative of Ontario because the population was not tested at random. Early in the pandemic, there was a bias towards testing symptomatic patients and people working in certain occupations (e.g., health care workers). This could lead to a type of selection bias known as collider bias (or detection bias), where some segments of the population are over or underrepresented in the analytic cohort [30–32]. By conditioning the results on those who were tested and symptomatic, this could lead to an over- or under-estimation of the MRR. In addition, about 15% cases of COVID-19 captured were not captured in the OLIS; therefore, our study under reports the number COVID-19 cases in the observation window compared to the numbers officially reported. However, given that most cases were captured in our analysis and missing cases don’t appear to cluster in a few select neighbourhoods, we do not expect that the missing cases would affect the MRR results in a meaningful way. This study shows that socio-demographic neighbourhood-level factors explain almost half of the observed geographic variation in COVID-19 rates in Ontario, Canada. The research highlights the importance of neighbourhood-level characteristics in explaining the geographic variation in COVID-19 rates and suggests a need to a...

    Results from TrialIdentifier: No clinical trial numbers were referenced.

    Results from Barzooka: We did not find any issues relating to the usage of bar graphs.

    Results from JetFighter: We did not find any issues relating to colormaps.

    Results from rtransparent:
    • Thank you for including a conflict of interest statement. Authors are encouraged to include this statement when submitting to a journal.
    • Thank you for including a funding statement. Authors are encouraged to include this statement when submitting to a journal.
    • No protocol registration statement was detected.

    About SciScore

    SciScore is an automated tool that is designed to assist expert reviewers by finding and presenting formulaic information scattered throughout a paper in a standard, easy to digest format. SciScore checks for the presence and correctness of RRIDs (research resource identifiers), and for rigor criteria such as sex and investigator blinding. For details on the theoretical underpinning of rigor criteria and the tools shown here, including references cited, please follow this link.

  2. Version published to 10.1101/2021.04.06.21254988v2 on medRxiv
  3. Version published to 10.1101/2021.04.06.21254988v1 on medRxiv