Evolving ICU strategies and improved survival in cardiogenic shock: a 13-year cohort study

Background Most studies on cardiogenic shock (CS) focus on selected subgroups (predominantly acute myocardial infarction) or single interventions. Comprehensive longitudinal analyses across heterogeneous CS populations over extended time periods remain scarce. Purpose To evaluate temporal changes in patient characteristics and to assess trends in mortality and morbidity in relation to evolving therapeutic strategies within a broad CS population. We hypothesized that structured adaptations in intensive care management contributed to improved clinical outcomes. Methods This retrospective dual-center cohort study included 2,029 CS patients treated between 2013 and 2025. Temporal trends were analyzed using descriptive statistics and multivariate logistic regression. Era-specific effects (2013–2018 vs. 2019–2025) were investigated using 1:1 propensity score matching. The primary endpoint was 90-day mortality; secondary endpoints comprised complications (bleeding, sepsis, thrombosis, and stroke), neurological outcome at ICU discharge, and shock progression. To quantify the contribution of evolving treatment strategies, domain-specific therapy-shift analyses using g-computation standardization were performed, with counterfactual mortality using Monte Carlo resampling and subclass-cluster bootstrapping. Results Despite stable baseline demographics (mean age 65.9 years; 72% male), admission profiles shifted toward less advanced shock and ischemic etiology declined over time. Ninety-day mortality decreased significantly (odds ratio per year, 0.95; p < 0.001). After matching, survival gains were most pronounced in patients with moderate shock severity, absence of cardiac arrest, and a non-ischemic cause. Shock progression and major bleeding – including severe bleeding – declined significantly, whereas rates of sepsis, thrombosis, and stroke remained stable. Neurological recovery improved modestly. Outcomes paralleled distinct therapeutic shifts: transition from midazolam- to propofol-based sedation, increased use of dobutamine combined with noradrenaline and reduced epinephrine exposure, more selective deployment of mechanical circulatory support, and expanded senior intensivist involvement. Counterfactual modeling suggested that optimization of sedation and catecholamine strategies accounted for a substantial proportion of the observed reduction in mortality. Conclusions Over 13 years, outcomes in cardiogenic shock improved substantially, with declining mortality, reduced shock progression, fewer bleeding complications, and better neurological recovery. These findings suggest that the evolution of contemporary ICU management and treatment strategies in CS has translated into meaningful clinical benefits for patients. Gains were most evident in moderate and non-ischemic phenotypes, whereas outcomes in refractory shock remain poor, underscoring the persistent need for novel therapeutic approaches beyond current standards of care.