Abstract / Summary
The estimated pulse wave velocity (ePWV) is a rapidly acquirable tool for assessing arterial stiffness and vascular elasticity. However, the relationship between ePWV and in-hospital mortality in patients with sepsis-associated encephalopathy (SAE) has not been elucidated. We conducted a cohort study involving 6571 participants diagnosed with SAE admitted to the ICU from the MIMIC-IV (v3.1) database and the eICU Collaborative Research Database (eICU-CRD). The calculation of ePWV relied on a validated formula dependent upon age and mean blood pressure. Considering ePWV's role in assessing cumulative vascular load, we classified the cohort according to ePWV levels (high/low) using a threshold of 10 m/s. Adjusted for clinical and laboratory variables, Cox proportional hazards models evaluated the relationship of ePWV to in-hospital mortality. We employed restricted cubic spline (RCS) analysis to examine potential non-linear effects. Model prediction was evaluated in terms of discrimination (AUC), calibration (calibration plot/Brier score), and overall performance (accuracy and precision). Furthermore, analyses of subgroups were performed according to the demographic and clinical characteristics. There were 3623 and 2948 patients in the MIMIC-IV and eICU-CRD cohorts, respectively. Higher ePWV was independently associated with increased in-hospital mortality, with adjusted hazard ratios of 1.12 (95%CI 1.07–1.18; P < 0.001) in MIMIC-IV cohort, 1.04 (95%CI 1.01–1.06; P = 0.001) in eICU-CRD cohort. In both cohorts, survival was significantly poorer in the high ePWV cohort according to Kaplan-Meier analysis ( P < 0.05). The RCS model suggested a positive linear link between ePWV and in-hospital mortality. Combining ePWV with other established risk scores improved discrimination, calibration, and clinical utility, surpassing the combined severity score model in both cohorts. Elevated ePWV is independently associated with increased risk of in-hospital mortality in patients with SAE. The application of ePWV can help clinicians achieve early warning of high-risk mortality cohort.