Independent mortality predictors in older COVID-19 patients with delayed hospitalization: A retrospective study from Eastern China, 2023

Delayed hospitalization of older patients with COVID-19 results in the loss of the early antiviral treatment window, as well as the progression to critical illness involving shock, hypoxemia, and corticosteroid therapy. Nevertheless, the hierarchy of independent mortality predictors in this specific context remains to be fully established. The study aimed to identify and rank these predictors to inform clinical management. This retrospective observational study was conducted at the First People's Hospital of Xiaoshan District, China (January-February 2023). The present study included hospitalized patients aged ≥ 60 years with first-time severe acute respiratory syndrome coronavirus-2 infection and delayed hospitalization (>5 days from symptom onset). The data were predominantly collected electronically. Univariate and multivariate logistic regression were performed to identify independent risk factors for mortality. A total of 300 older hospitalized COVID-19 patients (median age 82 years; 57.3% male) were included. Overall mortality was 17.3% (52/300). In multivariable logistic regression, independent factors associated with mortality included older age (OR: 1.11, 95% CI: 1.03-1.19), elevated cardiac troponin I (OR: 1.01, 95% CI: 1.0-1.02), presence of shock (OR: 26.72, 95% CI: 7.27-98.18), prolonged corticosteroid use (OR: 8.19, 95% CI: 2.39-28.05), and requirement of oxygen therapy for hypoxia (OR: 2.98, 95% CI: 1.70-5.23). Shock was the strongest independent predictor of mortality in older COVID-19 patients with delayed hospitalization, outweighing prolonged corticosteroid use and hypoxemia. These findings emphasize the critical importance of early shock recognition in this vulnerable population, while also highlighting the need for judicious corticosteroid use. The identified risk hierarchy may serve as a framework for clinical decision-making in resource-constrained settings.