Abstract
BACKGROUND: Hospitalized coronavirus disease 2019 (COVID-19) patients have higher mortality rates. Parameters to predict mortality are needed. Therefore, we investigated the power of procalcitonin/albumin ratio (PAR) and C-reactive protein/albumin ratio (CAR) to predict in-hospital mortality in hospitalized COVID-19 patients. METHODS: In this study, 855 patients were included. Patients' PAR and CAR values were recorded from the hospital information management system. The patients were evaluated in two groups according to their in-hospital mortality status. RESULTS: In-hospital mortality was observed in 163 patients (19.1%). The median PAR and CAR values of patients in the non-survivor group were statistically significantly higher than those of patients in the survivor group, PAR (median: 0.07, interquartile range [IQR]: 0.03-0.33 vs. median: 0.02, IQR: 0.01-0.04, respectively; p < 0.001); CAR (median: 27.60, IQR: 12.49-44.91 vs. median: 7.47, IQR: 2.66-18.93, respectively; p < 0.001). The area under the curve (AUC) and odds ratio (OR) values obtained by PAR to predict in-hospital mortality were higher than the values obtained by procalcitonin, CAR, albumin, and CRP (AUCs of PAR, procalcitonin, CAR, albumin, and CRP: 0.804, 0.792, 0.762, 0.755, and 0.748, respectively; OR: PAR > 0.04, procalcitonin > 0.14, CAR > 20.59, albumin < 4.02, and CRP > 63; 8.215, 7.134, 5.842, 6.073, and 5.07, respectively). Patients with concurrent PAR > 0.04 and CAR > 20.59 had an OR of 15.681 compared to patients with concurrent PAR < 0.04 and CAR < 20.59. CONCLUSIONS: In this study, PAR was found to be more valuable for predicting in-hospital COVID-19 mortality than all other parameters. In addition, concurrent high levels of PAR and CAR were found to be more valuable than a high level of PAR or CAR alone.