Assessment tools are necessary for the adequate stratification of patients with AIDS-related pneumocystis pneumonia (PCP). The aim of this study was to evaluate the ability of severity assessment scores and inflammation- and nutrition-based parameters for predicting the 180-day mortality of AIDS-related PCP. This was a retrospective cohort study of patients with AIDS-related PCP admitted at the Beijing Di-Tan Hospital. The CURB-65 score, Pneumonia Severity Index (PSI) score, Acute Physiology And Chronic Health Evaluation II (APACHE II) score, C-reactive protein-to-albumin ratio (CAR), procalcitonin, neutrophil-to-lymphocyte ratio, lymphocyte-to-monocyte ratio, and platelet-to-lymphocyte ratio during the first 24 h of intensive care unit admission were analyzed. The prognostic values of the severity assessment scores and biomarkers for 180-day mortality were evaluated using receiver operating characteristic (ROC) curves and integrated discrimination improvement (IDI) indexes. A total of 123 patients with AIDS-related PCP were included. Fifty-five patients were dead, and 68 were still alive at 180 days after admission. CAR, CURB-65, PSI, and APACHE II were independent predictors of 180-day mortality. The optimal cutoff value of CAR was 2.0 mg/g [area under the ROC curve = 0.844, 95% credential interval (CI) = 0.776-0.913], and CAR >2.0 mg/g increased the prognostic value of all three severity assessment scores, with an IDI index of 5.1% for the CURB-65 score, 8.1% for the PSI score, and 4.1% for the APACHE II score (all p < .05). Combining CAR >2.0 mg/g enhanced the capability of CURB-65, APACHE II, and PSI in predicting the 180-day mortality of patients with AIDS-related PCP.
Keywords: AIDS; C-reactive protein; Pneumocystis; pneumonia; predictive value of tests; serum albumin.