Between-Center Variation in Outcome After Endovascular Treatment of Acute Stroke: Analysis of Two Nationwide Registries

Background: Insight in differences in patient outcomes between endovascular thrombectomy (EVT) centers can help to improve stroke care. We assessed between-center variation in functional outcome of patients with acute ischemic stroke who were treated with EVT. We analyzed to what extent this variation may be explained by modifiable center characteristics.

Methods: We used nationwide registry data of patients with stroke treated with EVT in the Netherlands and in the Czech Republic. Primary outcome was modified Rankin Scale score at 90 days as an indicator of disability. We used multilevel ordinal logistic regression to quantify the between-center variation in outcomes and the impact of patient and center characteristics. Between-center variation was expressed as the relative difference in odds of a more favorable modified Rankin Scale score between a relatively better performing center (75th percentile) and a relatively worse performing center (25th percentile).

Results: We included a total of 4518 patients treated in 33 centers. Adjusted for patient characteristics, the odds of a more favorable outcome in a center at the 75th percentile of the outcome distribution were 1.46 times higher (95% CI, 1.31-1.70) than the odds in a center at the 25th percentile. Adjustment for center characteristics, including the median time between stroke onset and reperfusion per center, decreased this relative difference in odds to 1.30 (95% CI, 1.18-1.50, P=0.01). This translates into an absolute difference in likelihood of good functional outcome of 8% after adjustment for patient characteristics and to 5% after further adjustment for modifiable center characteristics.

Conclusions: The considerable between-center variation in patient outcomes after EVT for acute ischemic stroke could be largely explained by center-specific characteristics, such as time to reperfusion. Improvement of these parameters may likely result in a decrease in center-specific differences, and an overall improvement in outcome of patients with acute ischemic stroke.