Related Experiment Video
Updated: Jul 1, 2026
Murine Isolated Heart Model of Myocardial Stunning Associated with Cardioplegic Arrest
Published on: August 6, 2015
Mortality Benefit of Microaxial Flow Pump Use in Infarct-Related Cardiogenic Shock: a Bayesian Reanalysis of the
Marco Tomasino1,2,3, Garazi Urio-Garmendia1,2,3, Eduard Ródenas-Alesina1,2,3
1Medicine Department, Universitat Autònoma de Barcelona (UAB), Barcelona, Spain.
Background:
The Danish-German Cardiogenic Shock (DanGer Shock) trial was the first randomized study to suggest a survival benefit of microaxial flow pump use in patients with ST-elevation myocardial infarction complicated by cardiogenic shock. The frequentist analysis reported borderline significance, leaving uncertainty regarding the robustness and magnitude of benefit. Bayesian methods offer a complementary framework by quantifying the probability of clinically meaningful treatment effects.
Methods:
We conducted a Bayesian reanalysis of the DanGer Shock trial using aggregate, published 180-day mortality data. Priors reflecting a spectrum of clinical beliefs (from strongly skeptical to strongly enthusiastic) were employed. Posterior probabilities of benefit were estimated for relative risk (RR) thresholds < 1, < 0.90, < 0.80, and < 0.70, and for absolute risk reduction (ARR) thresholds ≥ 2%, ≥ 4%, ≥ 6%, ≥ 8%, ≥ 10%, and ≥ 18%, assuming a baseline 180-day mortality risk of 58.5%.
Results:
With a minimally informative prior, the median posterior RR was 0.78 (95% credible interval, 0.64-0.95). The posterior probability of any mortality benefit (RR < 1) was 99%, whereas the probability of a large effect (RR < 0.70) was 15%. The posterior median ARR was 12.8% (95% credible interval, 3.0%-21.6%), with a 98% probability of ARR ≥ 2%, but only 17% probability of ARR ≥ 18%. Across all priors, the probability of RR < 1 ranged from 98% to 100%.
Conclusions:
This Bayesian reanalysis of the DanGer Shock trial supports a high probability that microaxial flow pump use reduces mortality in ST-elevation myocardial infarction complicated by cardiogenic shock, although large treatment effects appear unlikely. These findings complement the original trial and provide a more intuitive probabilistic interpretation for clinicians.
