Using the National Inpatient Sample for 2016-2022, hospitals were graded into five tiers of cardiac capability and in-hospital mortality after cardiogenic shock was modelled with a latent construct for acute physiological severity (cardiac arrest, acute kidney and liver injury, ventilation). Crude mortality declined from 64.4% at non-PCI hospitals to 36.5% at transplant/LVAD centres, and adjusted odds of death were roughly a third of tier 1 at higher tiers (tier 4 OR 0.34, 95% CI 0.30-0.38). Transfer-in status carried higher mortality (OR 1.36) but the penalty was attenuated at surgical and transplant centres.
Why it is interesting: Puts a size on the institutional gradient in cardiogenic shock survival, though administrative data cannot separate capability from case selection and transfer decisions.