Mortality from COVID in Colombia and Peru: Analyses of Mortality Data and Statistical Forecasts

  1. Patrick E Brown
  2. Zoë R Greenwald
  3. Luis Ernesto Salinas
  4. Gabriel Aguirre Martens
  5. Leslie Newcombe
  6. Peter S Rodriguez
  7. Hellen Gelband
  8. Jeremy Veillard
  9. Prabhat Jha

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

National predictions of the course of COVID mortality can be used to plan for effective healthcare responses as well as to support COVID policymaking. We developed the Global COVID Assessment of Mortality (GCAM), a statistical model with continually improving precision that combines actual mortality counts with Bayesian inference, to predict COVID trends, currently until December 1, 2020. In Colombia, the GCAM analysis found the peak of COVID mortality around August 12 and an expected total of COVID deaths of 24,000-31,000, or 48%-92% over the total through August 21. In Peru, a first mortality peak occurred around May 24, and given the current trajectory, a second peak is predicted around September 6. Peru can expect 29,000-43,000 COVID deaths, representing an increase of 7%-55% over COVID deaths through August 21. GCAM projections are also used to estimate medical surge capacity needs. To gauge the reliability of COVID mortality forecasts, we compared all-cause mortality from January through June 2020 with average all-cause mortality in previous years in Colombia and Peru, and found that the excesses were consistent with GCAM forecast, most notably a doubling of overall mortality from May 25-June 7 th of weeks in Peru. The GCAM results predict that as a percentage of all adult deaths in previous years, Colombia can expect about 13% excess from COVID deaths, whereas Peru can expect 34% excess. Comparisons of GCAM analyses of several other countries with Colombia and Peru demonstrate the extreme variability that characterizes COVID mortality around the world, emphasizing the need for country-specific analyses and ongoing monitoring as more mortality data become available.

Article activity feed

  1. ScreenIT

    SciScore for 10.1101/2020.08.24.20181016: (What is this?)

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

    Table 1: Rigor

    NIH rigor criteria are not applicable to paper type.

    Table 2: Resources

    No key resources detected.

    Results from OddPub: Thank you for sharing your code.

    Results from LimitationRecognizer: We detected the following sentences addressing limitations in the study:
    Notably, a key limitation across countries is the absence of population-representative serological surveys on random samples of populations (Mallapaty, 2020) to help establish the true infection-fatality rate. This underscores the need for improved mortality surveillance systems in Colombia and Peru, with the capacity to report mortality data with minimal time lags and by age and place of death (such as long-term care homes). Colombia and Peru would also benefit by launching COVID antibody sero-surveys in the near future to provide nationally representative estimates of the total number of people with a history of infection in each setting. The availability of serological data can substantially enhance the accuracy of modelling, beyond relying on reported cases alone, and can therefore better inform and monitor intervention strategies to minimize harm caused by COVID (Metcalf et al., 2020). All models, including GCAM, will produce some forecasts that turn out to be wrong. This is due, in part, to undercounts and delayed reporting of COVID deaths, which are now known to be common and variable (Katz et al., 2020; Montagano, 2020; ONS, 2020; The Economist, 2020). For example, the undercount in UK COVID deaths might have been about 40% for some weeks (ONS, 2020). In particular, undercounts in nursing home deaths have emerged as a key issue (Condon et al., 2020). Systematic examination of undercounts across settings is needed to establish if the undercounts have accelerated during...

    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 10.1101/2020.08.24.20181016v2 on medRxiv
  3. Version published to 10.1101/2020.08.24.20181016v1 on medRxiv