Association between baseline pulse pressure and prognosis in critically ill patients: a retrospective cohort study

Abstract

Abstract To investigate the association between baseline pulse pressure (PP) and mortality in critically ill patients. This cohort study analyzed 22,035 ICU patients from the MIMIC-IV database. Baseline PP was defined as the first recorded value after admission. Outcomes included in-hospital, 180-day, and 365-day all-cause mortality. Feature importance was assessed using XGBoost. The relationship between PP and mortality was modeled with smooth curve fitting and threshold analysis. Multivariable logistic regression and survival analyses were performed to compare mortality risk across PP groups. Subgroup analyses examined effect modifications. The restricted mean survival time (RMST) analysis validated the robustness of the results. PP holds certain clinical significance in predicting ICU outcomes. A U-shaped relationship was identified, with an inflection point at 49 mmHg (95% CI 46–51). Patients were categorized into three groups: < 46, 46–51, and ≥ 51 mmHg. Using the 46–51 mmHg group as reference, both lower and higher PP groups showed significantly increased risks of in-hospital ( P  < 0.05) and long-term mortality ( P  < 0.05). Significant interactions were observed in the elderly and septic subgroups. RMST analysis further confirmed the stability of the results. Baseline PP nonlinearly predicts mortality risk in ICU patients, with the optimal range being 46–51 mmHg. This association is stronger in elderly and septic patients.