Association between international normalized ratio-to-albumin ratio and mortality in critically ill patients with gastrointestinal bleeding: a retrospective MIMIC-IV database study

BMC Gastroenterol. 2025 Aug 11;25(1):574. doi: 10.1186/s12876-025-04179-1.

ABSTRACT

BACKGROUND: While the international normalized ratio-to-albumin ratio (PTAR) is an established independent prognostic indicator for various diseases, its predictive value for clinical outcomes in critically ill patients with acute gastrointestinal bleeding (GIB) has not been systematically evaluated. The present study aims to examine the correlation between PTAR levels and clinical outcomes in critically ill patients with GIB.

METHODS: To evaluate mortality risk, we conducted multivariable Cox proportional hazards regression analyses for both short-term and long-term survival, supplemented by Kaplan-Meier (K-M) curve analysis to compare survival trends among subgroups. The discriminatory performance of PTAR was further assessed using receiver operating characteristic (ROC) curve analysis, while restricted cubic splines (RCS) modeled nonlinear associations. Subgroup analyses were performed to examine potential interaction effects, and feature selection was executed via the Boruta algorithm. Finally, a prognostic nomogram integrating the key predictors was developed.

RESULTS: To explore the prognostic value of PTAR in GIB, this retrospective cohort study analyzed 1,557 patients. X-tile analysis determined an optimal cutoff value of 0.73 for PTAR in predicting 28-day mortality. Subsequent multivariate Cox regression analysis demonstrated that PTAR was an independent predictor of all-cause mortality (P < 0.001). This finding was corroborated by significant differences in KM survival curves (log-rank P < 0.001). Further assessment via ROC analysis indicated that PTAR exhibited moderate prognostic discrimination. Additionally, RCS analysis revealed a linear association between PTAR and mortality (P-nonlinear > 0.05). Subgroup analyses were also performed, yet no significant interactions were observed (all P-interaction values > 0.005). Finally, a nomogram developed using the Boruta algorithm, which incorporated PTAR as a key variable, achieved AUC values of 0.804 for 28-day mortality and 0.791 for 365-day mortality, respectively.

CONCLUSION: The PTAR is significantly associated with increased short- and long-term mortality risk in critically ill patients with GIB.

PMID:40790158 | PMC:PMC12337426 | DOI:10.1186/s12876-025-04179-1