Diffuse alveolar hemorrhage (DAH) is a life‑threatening pulmonary complication associated with various autoimmune diseases, vasculitides, and other systemic disorders. Despite advances in immunosuppressive and supportive therapies, the short‑term mortality of DAH remains high, ranging from 20% to 50% across different cohorts. Early identification of patients at highest risk of in‑hospital death is crucial for guiding intensive monitoring and treatment decisions. However, existing risk stratification tools are often based on disease‑specific cohorts or require complex scoring systems, limiting their generalizability in real‑world clinical practice. The C‑reactive protein to albumin ratio (CAR) has emerged as a simple, inexpensive, and readily available inflammatory‑nutritional biomarker that integrates both acute inflammation (CRP) and nutritional status (albumin). Although CAR has been shown to predict prognosis in various critical illnesses such as sepsis, acute respiratory distress syndrome, and certain cancers, its prognostic value specifically in DAH patients remains largely unexplored. Therefore, this study aimed to: (1) describe the clinical features of a cohort of patients with DAH confirmed by bronchoscopy or clinical criteria; (2) identify independent risk factors for short‑term mortality (in‑hospital death) using univariate and multivariate logistic regression; and (3) evaluate the predictive performance of baseline CAR, including determination of its optimal cutoff value, sensitivity, and specificity for risk stratification.

This retrospective study analyzed data from 62 patients with DAH at Nanjing Drum Tower Hospital between January 2015 and December 2023, using both inpatient records and outpatient follow-up data. Based on their discharge outcome, patients were divided into survival and non-survival groups. Demographic characteristics, clinical symptoms, and laboratory indices were compared between the two groups. Independent risk factors for short-term mortality were identified using univariate and multivariate logistic regression analyses.

A total of 62 patients with DAH were included in the study. Among them, 27 (43.5%) died during hospitalization. In multivariate analysis, an elevated C-reactive protein to albumin ratio (CAR) was identified as an independent predictor of short-term mortality (odds ratio [OR] = 11.78, 95% confidence interval [CI]: 1.84 - 75.25; P = 0.009). The predictive performance of CAR was evaluated using receiver operating characteristic (ROC) curve analysis. The area under the curve (AUC) was 0.754 (95% CI: 0.60 - 0.87, P = 0.002). At the optimal cutoff value of 2.53, CAR yielded a sensitivity of 70% and a specificity of 87.5%.

In this study, an elevated baseline CAR at admission was an independent predictor of short-term mortality in patients with DAH. This readily available biomarker emerges as a potential tool for risk stratification in clinical practice. Early identification of high-risk patients using CAR may aid in guiding more intensive monitoring and treatment.