Association Between Pre-Existing Conditions and COVID-19 Hospitalization, Intensive Care Services, and Mortality: A Cross-Sectional Analysis of an International Global Health Data Repository

Abstract

Background: The use of globally shared individual-level data in answering epidemiological questions during health emergencies of international concern is still debatable. In this study, we investigated the association between pre-existing conditions and clinical outcomes of COVID-19 using data from a global data sharing repository. Methods: We used data of all cases recorded in the Global Health Data repository up to the 10th of March 2021 to carry out a cross-sectional analysis of associations between cardiovascular diseases (CVD), hypertension, diabetes, obesity, lung diseases, and kidney disease and hospitalization, ICU admission, and mortality due to COVID-19. The Global Health repository reported data from 137 countries, but only Brazil, Mexico, and Cuba reported more than 10 COVID-19 cases in participants with preexisting conditions. We used multivariable logistic regression to compute adjusted odds ratios (aOR) of the three outcomes for each pre-existing condition in ten-year age groups from 0 to 9 years and up to 110–120 years. Findings: As of March 10, the Global Health repository contained 25,900,000 records of confirmed cases of COVID-19, of which 2,900,000 cases from Brazil, Mexico, and Cuba had recorded data on pre-existing conditions. The overall aOR of ICU admission for each pre-existing condition were; CVD (aOR 2.1, 95%CI 1.8–2.4), hypertension (aOR 1.3, 95%CI 1.2–1.4), diabetes (OR 1.7, 95%CI 1.5–1.8), obesity (OR 2.2, 95%%CI 2.1–2.4), kidney disease (OR 1.4, 95%CI 1.2–1.7) and lung disease (aOR 1.1, 95%CI 0.9–1.3). Overall aORs of mortality for each pre-existing condition were: CVD (aOR 1.7, 95%CI 1.6–1.7), hypertension (aOR 1.3, 95%CI 1.3–1.4), diabetes (aOR 2.0, 95%CI 1.9–2.0), obesity (aOR 1.9, 95%CI 1.8–2.0), kidney disease (aOR 2.7, 95%CI 2.6–2.9), and lung disease (aOR 1.6, 95%CI 1.5–1.7). The odds of each adverse outcome were considerably larger in children and young adults with these preexisting conditions than for adults, especially for kidney disease, CVD, and diabetes. Conclusion: This analysis of a global health repository confirms associations between pre-existing diseases and clinical outcomes of COVID-19, and the odds of these outcomes were especially elevated in children and young adults with these preexisting conditions. This study shows that global data sharing can unlock answers to many epidemiological questions efficiently especially during the early stages of global health emergencies.