Trends in stroke-related mortality in California hospitals from 2010 to 2020: Have the large core stroke trials made a difference?

Abstract

Background

Publicly reported hospital risk-adjusted mortality rates (RAMRs) reflect real-world outcomes and may be used to understand the impact of advances in medical evidence. Our study presents an analysis of RAMRs in California hospitals across the time period of publication of major trials in stroke intervention, to interrogate the effect of these trials upon population-level mortality from stroke.

Methods

Stroke (total acute, ischemic, hemorrhagic, subarachnoid hemorrhage) RAMR data from 2010 to 2020 was extracted from the California Hospital Inpatient Mortality Rates and Quality Ratings dataset. Hospitals were categorized by county population, size and type (academic/community). ANOVA with Tukey–Kramer and Bonferroni-corrected t-tests, and independent t-tests were used for statistical comparison of RAMRs across different population groups and hospital types.

Results

There was a statewide decline in acute stroke mortality from 11.4 % to 8.6 %, with ischemic stroke mortality decreasing from 24.9 % to 21.6 %. RAMRs decreased from 5.7 % to 5.0 % in community hospitals (p = 0.006), a trend not mirrored in academic settings. Hemorrhagic stroke RAMRs fluctuated, while subarachnoid hemorrhage RAMRs increased, except in academic institutions. Hospitals in the >2M population group had significantly lower RAMRs (p < 0.005) than the 0-500k group. There were no significant RAMR differences between academic and community hospitals across all stroke types.

Conclusions

Despite the publication of paradigm-shifting trials, California in-patient stroke mortality only modestly changed, reflecting the complexity of replicating clinical trial outcomes in real-world data. Consistent, longitudinal quality and outcome metrics at state and national levels remain essential for understanding the impact of clinical research and innovation.