Wen Li, Zhanfang Zhu, Jinxia Hao, Yunhui Zheng, Ruochen Zhang, Xiaoxiang Liu, Ying Lv, Fuqiang Liu, Na Zhu
RAR has been confirmed as the most reliable indicator for predicting 30-day mortality in RHD patients. It offers a simple, cost-effective tool for early risk stratification, particularly in resource-limited settings where RHD burden remains highest.
BACKGROUND: Rheumatic heart disease (RHD) is a chronic valvular condition characterized by persistent autoimmune inflammation and nutritional deficiencies. Albumin not only reflects the body's nutritional status but also eliminates pro-inflammatory stimuli and mitigates inflammatory responses. This study aimed to evaluate the prognostic value of novel albumin-related inflammatory biomarkers for short-term mortality in patients with RHD.
METHODS: This retrospective dual-cohort study included 496 RHD patients from the MIMIC-IV database and 268 patients from Shaanxi Provincial People's Hospital. The primary outcome was all-cause mortality within 30 days. Cox proportional hazards regression, restricted cubic spline analysis, time-dependent ROC curves, and integrated discrimination improvement (IDI) were performed to assess the independent associations, dose-response relationships, discriminative ability, and incremental predictive value of each biomarker.
RESULTS: RAR consistently demonstrated the strongest association with 30-day mortality across both cohorts. In fully adjusted Cox models, RAR remained significantly associated with mortality (MIMIC: HR 4.12, 95% CI 1.79-9.49; external validation: HR 3.20, 95% CI 1.25-8.23). RAR also showed significant positive linear dose-response relationships, achieved the highest time-dependent AUC, and provided significant incremental predictive value beyond SOFA, SAPS II, and OASIS in both cohorts (all IDI p < 0.05). LAR was protective in the validation cohort (HR 0.61, 95% CI 0.43-0.88), while NAR and PAR showed cohort-specific associations. MAR demonstrated limited prognostic value.
CONCLUSION: RAR has been confirmed as the most reliable indicator for predicting 30-day mortality in RHD patients. It offers a simple, cost-effective tool for early risk stratification, particularly in resource-limited settings where RHD burden remains highest.