Risk Prediction of Death in Inpatient Adults With COVID-19 from Mexico

Area of Science:

• Epidemiology
• Public Health
• Infectious Diseases

Background:

• COVID-19 exhibits significant lethality variations globally, with pronounced regional differences in Mexico due to socioeconomic disparities.
• This study investigated the association between time to hospitalization, healthcare site, and COVID-19 mortality risk in Mexican patients, developing prognostic models.

Approach:

• Two COVID-19 inpatient cohorts (prospective and retrospective) from Chiapas, Mexico, were analyzed.
• Demographic, clinical, and laboratory data were collected; SARS-CoV-2 infection was confirmed via RT-qPCR.
• Multivariate logistic regression analyzed 30-day mortality, focusing on clinical variables within the first 48 hours of hospitalization.

Key Points:

• Of 392 patients, 59.4% died within 30 days of symptom onset.
• Hospitalization delay and healthcare site did not significantly correlate with mortality.
• Three prognostic models were developed (AUC 0.726-0.807), incorporating age, LDH, AST, lymphocyte count, OSI-WHO Classification, leukocyte count, diabetes, and diarrhea.

Conclusions:

• The study population exhibited pre-existing health deterioration compared to the general population.
• While age, sex, and comorbidities are mortality factors in healthy populations, their significance shifts in populations with compromised health.