Probabilistic survival modeling in health research: an assessment using cohort data from hospitalized patients with COVID-19 in a Latin American city

ABSTRACT

Probabilistic survival methods have been used in health research to analyze risk factors and adverse health outcomes associated with COVID-19. The aim of this study was to employ a probabilistic model selected among three distributions (exponential, Weibull, and lognormal) to investigate the time from hospitalization to death and determine the mortality risks among hospitalized patients with COVID-19. A retrospective cohort study was conducted for patients hospitalized due to COVID-19 within 30 days in Londrina, Brazil, between January 2021 and February 2022, registered in the database for severe acute respiratory infections (SIVEP-Gripe). Graphical and Akaike Information Criterion (AIC) methods were used to compare the efficiency of the three probabilistic models. The results from the final model were presented as hazard and event time ratios. Our study comprised of 7,684 individuals, with an overall case fatality rate of 32.78%. Data suggested that older age, male sex, severe comorbidity score, intensive care unit admission, and invasive ventilation significantly increased risks for in-hospital mortality. Our study highlights the conditions that confer higher risks for adverse clinical outcomes attributed to COVID-19. The step-by-step process for selecting appropriate probabilistic models may be extended to other investigations in health research to provide more reliable evidence on this topic.

KEYWORDS:

• Survival analysis
• parametric models
• SARS-CoV-2
• risk factors
• Brazil

Acknowledgments

We acknowledge CAPES (Coordenação de Aperfeiçoamento de Pessoal de Nível Superior) for the financial support and the City Health Secretary of Londrina for providing the data.

Disclosure statement

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper

Additional information

Funding

The work was supported by the CAPES [001].