D93-03 Predicting Outcomes in Immunocompromised Critical Illness: The Role of the Chelsea Critical Care Physical Assessment Compared With Conventional Clinical Scores

Rationale Several instruments have been developed to assess physical function in critically ill patients admitted to the intensive care unit (ICU). Although the association between functional status and survival is well established, few studies have directly compared the prognostic performance of the Chelsea Critical Care Physical Assessment (CPAx) with conventional clinical scores. We hypothesized that CPAx indices, compared with clinical severity scores, would effectively predict survival and the need for invasive mechanical ventilation (IMV) in immunocompromised critically ill patients. Methods A prospective cohort study was conducted between January 2024 and April 2025 at the National Institute of Infectious Diseases, Rio de Janeiro, Brazil. Consecutive adults (≥18 years-old) admitted to the ICU with a length of stay ≥48 hours were included. Demographic and clinical data were collected at ICU admission, including SOFA, SAPS-III, and CPAx scores. CPAx at ICU discharge, ΔCPAx (difference between discharge and admission), and the CPAx index (ΔCPAx divided by ICU length of stay) were calculated. The primary outcome was survival, and secondary outcome was the need for IMV. Discrimination was evaluated using receiver operating characteristic curve analysis, with the area under the curve (AUC, 95% CI) as a measure of predictive performance. The Youden criterion identified optimal cutoffs, and comparisons among AUCs were performed using DeLong’s method. Results Of 3,193 ICU admissions, 708 patients were screened and 540 met inclusion criteria. Among them, 320 survived and 220 died. The most common reasons for ICU admission were HIV infection (49%) and tuberculosis (24%). For survival prediction, AUCs for CPAx at ICU admission and SOFA were comparable (0.76; 95% CI, 0.72-0.80 vs 0.74; 95% CI, 0.70-0.78; p = 0.398). In contrast, CPAx at ICU discharge, ΔCPAx, and the CPAx index demonstrated significantly better discrimination than SOFA and SAPS-III (all p < 0.001), even after adjustment. For IMV, CPAx at ICU admission showed higher predictive accuracy than SOFA (AUC 0.80; 95% CI, 0.76-0.84 vs 0.71; 95% CI, 0.67-0.76; p < 0.001). A similar pattern was observed for CPAx at ICU discharge and the CPAx index when compared with SOFA and SAPS-III. Conclusion In immunocompromised ICU patients, the CPAx score at admission showed similar performance to SOFA and SAPS-III for survival prediction, whereas CPAx-derived measures demonstrated superior prognostic accuracy. For invasive mechanical ventilation, CPAx-derived indices also outperformed conventional clinical scores. Given its simplicity and low resource requirements, the CPAx functional scale may serve as a feasible prognostic tool in low- and middle-income settings. This abstract is funded by: FAPERJ, CNPq, CAPES