Vasoactive–Inotropic Score Reduction Rate Is Highly Associated With Prognosis for Critically Ill Patients With Cardiogenic Shock: Insights From the Real‐World Dynamic Data

TL;DR

This study shows that a faster decrease in vasoactive-inotropic support is linked to better survival in patients with cardiogenic shock.

Contribution

The study introduces the vasoactive–inotropic score reduction rate (VRR) as a novel dynamic prognostic indicator for cardiogenic shock patients.

Findings

Increasing vasoactive-inotropic scores over time are consistently associated with higher mortality in cardiogenic shock patients.

VRR was validated across two large ICU databases, showing robustness and generalizability.

A greater reduction in vasoactive-inotropic scores is strongly linked to lower in-hospital and ICU mortality.

Abstract

In cardiogenic shock (CS), titration of both the choice and dosage of inotropes and vasopressors is crucial. The vasoactive–inotropic score (VIS) quantifies hemodynamic support, while the effect of temporal VIS change on prognosis remains unclear. This study evaluated the association between the vasoactive–inotropic score reduction rate (VRR) to measure VIS changes over time and explore its association with mortality in CS patients.

We performed a retrospective observational study using two large intensive care databases (MIMIC‐IV and eICU). Adult patients with CS receiving vasoactive–inotropic agents within 24 h after ICU admission were included. VIS was calculated using the updated VIS 2020, and VRR was defined to capture the relative change in VIS over time. Patients were categorized into VIS‐decreasing and VIS‐increasing groups. The primary outcome was in‐hospital mortality; secondary outcomes were intensive care unit (ICU) and 28‐day mortality (28‐day mortality only for the MIMIC‐IV database). Associations between VRR and outcomes were examined using four models: (1) log‐rank analysis, (2) Cox model adjusted for all covariates, (3) Cox model adjusted for covariates selected by univariable analyses, and (4) Cox model adjusted for covariates selected by a random forest algorithm.

A total of 3170 adult CS patients were analyzed. All models in both MIMIC‐IV and eICU showed that an increasing VIS over time was consistently associated with higher in‐hospital, ICU, and 28‐day mortality compared with a decreasing VIS in the MIMIC‐IV cohort (p < 0.001). These findings were reproduced in the external eICU cohort for ICU and in‐hospital mortality (p < 0.001), supporting the robustness and generalizability of VRR as a prognostic indicator across heterogeneous ICUs.

In adult CS patients, a greater reduction in VIS over time is strongly associated with lower mortality. VRR provides a simple, dynamic summary of vasoactive trajectories and may serve as a useful adjunct for risk stratification alongside other clinical and biochemical markers, although prospective validation is warranted.

Figures