Risk Factors for COVID-19 versus non-COVID-19 related in-hospital and community deaths by Local Authority District in Great Britain

  1. Samuel Paul Leighton
  2. Danielle Jane Leighton
  3. James Herron
  4. Rachel Upthegrove
  5. Jonathan Cavanagh
  6. Georgios Gkoutos
  7. Breda Cullen
  8. Pavan K Mallikarjun

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

Objectives: To undertake a preliminary hypothesis-generating analysis exploring putative risk factors for coronavirus diseae 2019 (COVID-19) population-adjusted deaths, compared with non-COVID-19 related deaths, at a local authority district (LAD) level in hospital, care homes and at home. Design: Ecological retrospective cohort study Setting Local authority districts (LADs) in England, Scotland and Wales (Great Britain (GB)). Participants All LAD deaths registered by week 16 of 2020. Main Outcome Measures Death registration where COVID-19 is mentioned as a contributing factor per 100,000 people in all settings, and in i) cares homes, ii) hospitals or iii) home only, in comparison to non-COVID-19 related deaths. Results Across GB by week 16 of 2020, 20,684 deaths had been registered mentioning COVID-19, equivalent to 25.6 per 100,000 people. Significant risk factors for LAD COVID-19 death in comparison to non-COVID-19 related death were air pollution and proportion of the population who were female. Significant protective factors were higher air temperature and proportion of the population who were ex-smokers. Conversely, for all COVID-19 unrelated deaths in comparison to COVID-19 deaths, higher rates of communal living, higher population rates of chronic kidney disease, chronic obstructive pulmonary disease, cerebrovascular disease deaths under 75 and dementia were predictive of death, whereas, higher rates of flight passengers was protective. Looking at individual setttings, the most notable findings in care homes was Scotland being a significant risk factor for COVID-19 related deaths compared to England. For hospital setting, the proportion of the population who were from black and Asian minority ethnic (BAME) groups significantly predicted COVID-19 related death. Conclusions This is the first study within GB to assess COVID-19 related deaths in comparison to COVID-19 unrelated deaths across hospital, care homes and home combined. As an ecological study, the results cannot be directly extrapolated to individuals. However, the analysis may be informative for public health policy and protective measures. From our hypothesis-generating analysis, we propose that air pollution is a significant risk factor and high temperature a significant protective factor for COVID-19 related deaths. These factors cannot readily be modelled at an individual level. Scottish local authorities and local authorities with a higher proportion of individuals of BAME origin are potential risk factors for COVID-19 related deaths in care homes and in hospitals, respectively. Altogether, this analysis shows the benefits of access to high quality open data for public information, public health policy and further research.

Article activity feed

  1. ScreenIT

    SciScore for 10.1101/2020.05.21.20108936: (What is this?)

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

    Table 1: Rigor

    Institutional Review Board Statementnot detected.
    Randomizationnot detected.
    Blindingnot detected.
    Power Analysisnot detected.
    Sex as a biological variablenot detected.

    Table 2: Resources

    Software and Algorithms
    SentencesResources
    Alternatively, using G*Power,[46] at alpha 0.05 and power 0.80, the sample size and number of predictors is sufficient to detect an overall model effect size of f2 = 0.07 (equivalent to R2 = 0.07).
    G*Power
    suggested: (G*Power, RRID:SCR_013726)

    Results from OddPub: Thank you for sharing your code and data.

    Results from LimitationRecognizer: We detected the following sentences addressing limitations in the study:
    Limitations: We have reported aggregate data related to local authorities in GB. We acknowledge that such data should not be used to make inferences about individual outcomes (the ecological fallacy). Indeed, one recent study observed that living further from the European Union headquarters was associated with reduced risk of SARS-CoV-2 infection[68]. However, ecological studies do allow us to make inferences about population for public health intervention and are important for hypothesis generation.[69] Further, our study may provide clarity regarding some factors which are only relevant at a local or population level when controlled for societal confounders, such as pollution and climate.[69] In multivariable models such as ours in which all predictors are modelled together, the interpretation that each predictor is ‘mutually adjusted’ for the others is problematic when some predictors in fact form a mediating pathway towards the outcome. Although we have identified certain risk and protective factors whose association with the outcome is statistically significant even when other covariates are in the model, we have not attempted to conceptualise these in a causal order. Future research using individual-level data should consider plausible hypotheses to be tested in path models, to identify the likely temporal order of risk factors and the optimum point along the path at which an intervention could be targeted. If data regarding a predictor was available, and measured in th...

    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 10.1101/2020.05.21.20108936v1 on medRxiv