The predictors of COVID-19 mortality in a nationwide cohort of Turkish patients

Abstract

The COVID-19-related death rate varies between countries and is affected by various risk factors. This multi-center registry study was designed to evaluate the mortality rate and the related risk factors in Turkey. We retrospectively evaluated 1500 adults with COVID-19 from 26 centers who were hospitalized between March 11 and July 31, 2020. In the study group, 1041 and 459 cases were diagnosed as definite and highly probable cases, respectively. There were 993 PCR-positive cases (66.2%). Among all cases, 1144 (76.3%) were diagnosed with non-severe pneumonia, whereas 212 (14.1%) had severe pneumonia. Death occurred in 67 patients, corresponding to a mortality rate of 4.5% (95% CI:3.5-5.6). The univariate analysis demonstrated that various factors, including male sex, age >= 65 years and the presence of dyspnea or confusion, malignity, chronic obstructive lung disease, interstitial lung disease, immunosuppressive conditions, severe pneumonia, multiorgan dysfunction, and sepsis, were positively associated with mortality. Favipiravir, hydroxychloroquine and azithromycin were not associated with survival. Following multivariate analysis, male sex, severe pneumonia, multiorgan dysfunction, malignancy, sepsis and interstitial lung diseases were found to be independent risk factors for mortality. Among the biomarkers, procalcitonin levels on the 3rd-5th days of admission showed the strongest associations with mortality (OR: 6.18; 1.6-23.93). This study demonstrated that the mortality rate in hospitalized patients in the early phase of the COVID-19 pandemic was a serious threat and that those patients with male sex, severe pneumonia, multiorgan dysfunction, malignancy, sepsis and interstitial lung diseases were at increased risk of mortality; therefore, such patients should be closely monitored.