A 4-Domain Composite PCMR Score for Risk Stratification in Community-Acquired Pneumonia: A Single-Center Pilot Study

Objective: Established community-acquired pneumonia (CAP) severity scores, such as CURB-65, rely on physiological derangement. We explored whether an admission score combining inflammatory, metabolic, nutritional, and radiographic data could stratify risk. Methods: We retrospectively analyzed 74 of 96 adults hospitalized with CAP at a Turkish state hospital between January 2024 and June 2025. The PCMR score assigns 1 point each for neutrophil-to-lymphocyte ratio ≥7.85, triglyceride-glucose index ≥8.89, Geriatric Nutritional Risk Index ≤92, and multilobar infiltrate. The laboratory cutoffs were derived via Youden index within this cohort. The primary outcome was intensive care admission, invasive mechanical ventilation, or in-hospital death. Comparators included CURB-65, complete SMART-COP, and Pneumonia Severity Index class. Internal validation used repeated cross-validation and bootstrapping. Results: The outcome occurred in 10 of 74 patients (13.5%), all requiring intensive care. The apparent area under the receiver operating characteristic curve (AUC) of the PCMR score was 0.90 (95% CI, 0.81–0.97), decreasing to 0.86 after bootstrap optimism correction and 0.81 in cross-validation. The apparent AUC was 0.92 for SMART-COP, 0.90 for Pneumonia Severity Index, 0.81 for CURB-65, and 0.77 for neutrophil-to-lymphocyte ratio alone. No paired difference was statistically significant. No patient scoring 0 or 1 experienced an event. Conclusions: In this pilot cohort, the PCMR score showed good apparent discrimination, which decreased after internal validation and was not superior to SMART-COP or Pneumonia Severity Index. This 4-domain concept requires prospective validation with fixed cutoffs in larger cohorts before clinical use.