Insights from One of the Largest COVID-19 Cohorts: A Detailed Analysis of Risk Factors, Mortality Predictors, and Vaccine Efficacy in Istanbul

This retrospective study of 2,624 COVID-19 inpatients treated at Koç University and American Hospital in Istanbul, between March 2020 and April 2022 provides critical insights into factors influencing patient outcomes. The extensive follow-up captured detailed disease progression, essential for optimizing regional treatment strategies and improving patient care during ongoing and future pandemics. Using SPSS and Python for data analysis, we identified risk factors for mortality, including hypertension (p=0.003), coronary artery disease (p<0.001), diabetes mellitus (p=0.013), COPD (p<0.001), chronic kidney disease (p<0.001), active hematological malignancy (p<0.001), active solid cancer (p<0.001), and immunosuppressive treatment (p<0.001). We also assessed markers such as IL-6, procalcitonin, C-reactive protein, and the CRP/PCT ratio for predicting mortality. Procalcitonin emerged as the most reliable predictor of mortality within the first two weeks (AUC: 0.870) and sepsis in the first week (AUC:0.853). Unvaccinated patients had a 2.646-fold higher risk of in-hospital death compared to those with at least one vaccine dose (CI: 1.800-3.890, p<0.001). Partial vaccination was associated with more favorable outcomes (OR: 0.346, CI: 0.191-0.626, p<0.001). Multiple doses of the same vaccine showed the best outcomes (p<0.001), with +3 doses of mixed vaccines being the next best (p=0.048). Notably, 23.8% of patients with +3 Biontech doses had severe/critical disease compared to 50% with +3 Sinovac doses, potentially due to selection bias. These findings underscore the importance of personalized care and strategic vaccination in managing COVID-19. By identifying early mortality predictors and analyzing vaccination effects, this study aids in improving outcomes for COVID-19 patients in Turkey and beyond.