Kaplan-Meier type survival curves for COVID-19: a health data based decision-making tool

TL;DR

This paper demonstrates that Kaplan-Meier survival curves, derived from inconsistent COVID-19 health data, can effectively reveal the disease's dynamics and differences across countries, serving as a useful decision-making tool.

Contribution

It introduces a robust, simple model using Kaplan-Meier curves to analyze COVID-19 progression despite data inconsistencies across countries.

Findings

Kaplan-Meier curves reflect differences in COVID-19 evolution among countries.

The model provides insights into disease dynamics despite non-uniform data.

Distinct curve properties can be interpreted to understand country-specific disease characteristics.

Abstract

Countries are recording health information on the global spread of COVID-19 using different methods, sometimes changing the rules after a few days. They are all publishing the number of new individuals infected, cured and dead, along with some supplementary data. These figures are often recorded in a non-uniform manner and do not match the standard definitions of these variables. However, in this paper we show that the Kaplan-Meier curves calculated with them could provide useful information about the dynamics of the disease in different countries. Our aim is to present a robust and simple model to show certain characteristics of the evolution of the dynamic process, showing that the differences of evolution among the countries is reflected in the corresponding Kaplan-Meier-type curves. We compare the curves obtained for the most affected countries so far, proposing possible interpretations of the properties that distinguish them.