Objective: Determine the CALL SCORE's diagnostic accuracy for the prediction of ICU admission and death in patients hospitalized for COVID-19 pneumonia in a reference hospital in Peru.

Methods: We performed an analytical cross-sectional observational study. We included patients with COVID-19 pneumonia treated at the "Dos de Mayo" National Hospital. Patients over 18 years old with a diagnosis confirmed by rapid or molecular testing were included. Those with an incomplete, illegible, or missing medical history and/or bacterial or fungal pneumonia were excluded. Data were extracted from medical records. The primary outcomes were mortality and admission to the ICU. The Call Score was calculated for each patient (4 to 13 points) and classified into three risk groups. Summary measures were presented for qualitative and quantitative variables. The area under the model curve and the operational characteristics (sensitivity, specificity) were calculated for the best cut-off point. 

Results: The Call Score reported an area under the curve of 0.59 (IC95%: 0.3 to 0.07), p = 0.43 for predicting death. However, for a cut-off point of 5.5, a sensitivity of 87% and a specificity of 65% were obtained. The area under the curve for ICU admission was 0.67 (95%CI: 0.3 to 0.07), p = 0.43; the 5.5 cut-off point showed a sensitivity of 82% and a specificity of 51%.

Conclusions: The Call Score shows a low performance for predicting mortality and admission to the ICU in Peruvian patients.

Keyword: (MESH) Mortality; Intensive Care Units; Clinical Decision Rules