Development and validation of a 28-day mortality risk prediction model for patients with AIDS suffering from sepsis secondary to pulmonary infections in the ICU

Abstract Background Patients with AIDS Suffering from sepsis secondary to pulmonary infections admitted to intensive care units (ICUs) carry a high burden of morbidity and mortality. Therefore, rapid identification and timely clinical decision-making are essential to optimize patient outcomes. Methods Patients with AIDS and sepsis secondary to pulmonary infections admitted to the ICU of Chengdu Public Health Clinical Medical Center (January 2016–June 2025) were stratified into a development cohort (n = 385) and a temporal validation cohort (n = 86). Clinical and laboratory variables within the first 24 h of ICU admission were retrospectively extracted; for those with repeated measurements, the values reflecting the most severe clinical status were selected. Univariate and multivariate logistic regression analyses were used to identify independent risk factors for 28-day mortality, which were integrated into a nomogram prediction model. Discriminative ability, calibration, and clinical utility of the model were evaluated using the area under the receiver operating characteristic (ROC) curve (AUC), calibration curve, and decision curve analysis (DCA), respectively, and the model was compared with the APACHE II and SOFA scores. Results Three independent risk factors were identified: number of dysfunctional organs (OR = 2.438, 95% CI: 1.634–3.638), septic shock (OR = 2.233, 95% CI: 1.034–4.822), and blood lactate level (OR = 1.207, 95% CI: 1.066–1.367). The 28-day mortality risk prediction model for patients with AIDS suffering from sepsis secondary to pulmonary infections in the ICU (MPASP) constructed with these factors exhibited an AUC of 0.877 (95% CI: 0.843–0.911), a sensitivity of 85.7%, and a specificity of 76.6% in the development cohort. In the temporal validation cohort, the AUC was 0.861 (95% CI: 0.778–0.943), with a sensitivity of 80.0% and a specificity of 82.5%. Calibration curves and DCA demonstrated good calibration and clinical utility. The MPASP has exhibited improved discriminative ability compared with the APACHE II score (AUC = 0.746) and SOFA score (AUC = 0.801) (both P < 0.01). Conclusions The MPASP model, based on the number of dysfunctional organs, presence of septic shock, and blood lactate level reflecting the most severe early clinical status within the first 24 h of ICU admission, shows potential utility for predicting 28-day mortality in ICU-admitted AIDS patients with sepsis secondary to pulmonary infection. Simple and easy to apply, it may serve as a practical reference for clinical decision-making in managing this specific patient population.