Abstract
BACKGROUND: The prognostic significance of dynamic changes in the albumin-to-C-reactive protein ratio (ACR) in elderly patients with community-acquired pneumonia (CAP) has not been fully elucidated. This study aims to evaluate the utility of ACR as a dynamic biomarker for predicting 28-day mortality and enhancing risk stratification in this high-risk population. METHODS: A retrospective cohort study was conducted on 437 elderly CAP patients (≥65 years). Serum albumin and C-reactive protein (CRP) levels were measured at admission (T0), 24 hours (T1), and 3 days (T2) post-admission. ACR was calculated for each time point, and its prognostic value was assessed using advanced statistical methods. RESULTS: The 28-day mortality rate was 16.7%. ACR levels were consistently lower in non-survivors across all time points (P < 0.001). RCS analysis revealed a nonlinear relationship between ACR and mortality risk. Time-varying ROC analysis demonstrated that ACR consistently outperformed CRP in predicting mortality, with superior area under the curve (AUC) values at all time points. Random-effects modeling indicated minimal inter-individual variability in ACR (random effects variance: 0.030; standard deviation: 0.175). Time-varying Cox regression confirmed a strong negative association between dynamic ACR changes and mortality risk, with a C-statistic of 0.833 (P < 0.001). CONCLUSION: Dynamic monitoring of ACR is a robust and clinically applicable tool for predicting short-term mortality in elderly CAP patients. By integrating markers of inflammation and nutritional status, ACR facilitates early identification of high-risk patients and supports personalized treatment strategies. These findings highlight the potential of ACR as a novel biomarker for improving clinical outcomes in this vulnerable population.