Association of the admission albumin–bilirubin score with long-term major adverse cardiovascular events among patients who survived to discharge after percutaneous coronary intervention for non-ST-segment elevation myocardial infarction: a single-center cohort study

Objective To investigate the association of admission albumin–bilirubin (ALBI) score with long-term major adverse cardiovascular events (MACEs) among patients who survived to discharge after percutaneous coronary intervention (PCI) for non-ST-segment elevation myocardial infarction (NSTEMI), and to assess whether ALBI provides prognostic information beyond the admission GRACE score. Methods We retrospectively included 681 patients hospitalized between January 1, 2023, and September 1, 2025, who survived to discharge after PCI for NSTEMI. Patients were categorized into ALBI tertiles (T1, T2 and T3). Kaplan–Meier analysis, multivariable Cox proportional hazards regression and restricted cubic spline analysis were used to evaluate the association between ALBI and long-term MACEs. Missing data were handled using predictive mean matching, generating 50 imputed datasets. Subgroup and sensitivity analyses were performed. The discriminatory performance of the admission GRACE-based model was compared before and after ALBI was added. Follow-up continued through September 6, 2026. Results Among 681 patients, the median follow-up duration was 585 days, during which 186 MACEs occurred, representing 27.3% of the cohort. The cumulative proportions of MACEs in T1, T2 and T3 were 18.1%, 29.1% and 34.8%, respectively, differing significantly among groups (all P < 0.001). After adjustment, each 1-standard-deviation (1-SD) increase in ALBI was associated with a 25% higher risk of long-term MACEs (HR = 1.25, 95% CI: 1.06–1.46, P = 0.007). Compared with T1, the corresponding HRs were 1.59 for T2 and 1.60 for T3. Restricted cubic spline analysis showed no evidence of a significant nonlinear association. The results were therefore compatible with an approximately linear association within the observed ALBI range. The direction of association remained positive in sensitivity analyses, although the complete-case estimate was less precise and was not statistically significant. Adding ALBI produced small, statistically nonsignificant increases in the GRACE model’s C-index and 365-day time-dependent AUC. Conclusion Among patients who survived to discharge after PCI for NSTEMI, a higher admission ALBI score was independently associated with increased risk of long-term MACEs, with an approximately linear pattern across the observed range. ALBI may serve as a simple marker of long-term risk. Its incremental predictive value beyond the GRACE score requires validation in multicenter prospective studies.