Prediction models of COVID-19 fatality in nine Peruvian provinces: A secondary analysis of the national epidemiological surveillance system.

ABSTRACT: There are initiatives to promote the creation of predictive COVID-19 fatality models to assist decision-makers. The study aimed to develop prediction models for COVID-19 fatality using population data recorded in the national epidemiological surveillance system of Peru. A retrospective cohort study was conducted (March to September of 2020). The study population consisted of confirmed COVID-19 cases reported in the surveillance system of nine provinces of Lima, Peru. A random sample of 80% of the study population was selected, and four prediction models were constructed using four different strategies to select variables: 1) previously analyzed variables in machine learning models; 2) based on the LASSO method; 3) based on significance; and 4) based on a post-hoc approach with variables co

SUBMITTER: Nieto-Gutierrez W 

PROVIDER: S-EPMC10824411 | biostudies-literature | 2024

REPOSITORIES: biostudies-literature