Two Estonian Stroke Centers, Same Tools, Sharply Different Treatment Rates

When two hospitals serve similar populations with comparable stroke-care infrastructure, you might expect them to treat patients in much the same way. A retrospective study from Estonia shows that this is not necessarily the case. The study, conducted by a team from the University of Tartu and affiliated hospitals, included all 635 patients admitted with acute ischemic stroke between January and June 2021 at the country’s two largest comprehensive stroke centers.

At Tartu University Hospital, 45.2 percent of these patients received recanalization therapy, compared with 30.7 percent at North Estonia Medical Centre. After adjusting for other factors in multivariable logistic regression, patients at North Estonia Medical Centre had roughly half the odds of receiving the therapy, with an adjusted odds ratio of 0.49 and a 95 percent confidence interval of 0.33 to 0.74. The key to this gap, the researchers found, lay in a specific group of patients: those with mild strokes, defined in the study as a score of 5 or lower on the National Institutes of Health Stroke Scale.

Mild strokes accounted for 301 of the 635 patients, or 47.5 percent of the total. Among these mild-stroke patients, the difference between the centers was striking. Tartu University Hospital treated 33.1 percent of them with recanalization therapy, while North Estonia Medical Centre treated only 10.0 percent, a difference that was highly statistically significant.

The Estonian data suggest that decisions about this group, rather than differences in how more severe strokes are handled, were the main driver of the overall inter-center variation in treatment rates. To assess whether these different treatment patterns translated into different results, the team examined both neurological and functional short-term outcomes. Neurological improvement was measured as the reduction in NIHSS score, while functional outcome was assessed with the modified Rankin Scale.

In a descriptive analysis, mild-stroke patients who received recanalization therapy showed greater neurological improvement than those who did not: a median NIHSS reduction of 2.0 points, with an interquartile range of 0.8 to 3.0, versus a median reduction of 1.0 point, with an interquartile range of 0.0 to 2.0, among untreated patients. The Wilcoxon rank-sum test gave a p value below 0.001 for this comparison. However, the picture changed once the researchers adjusted for other variables.

In the multivariable analysis, neurological improvement was independently associated with the treatment center and with the baseline NIHSS category, but not with recanalization therapy itself, which had an adjusted odds ratio of 0.94 with a 95 percent confidence interval of 0.43 to 2.06. Recanalization therapy was likewise not independently associated with good functional outcome. In other words, the apparent advantage seen in the raw data did not survive statistical adjustment, and the treated and untreated mild-stroke patients could not be shown to differ in outcome once other factors were accounted for.

The authors’ central conclusion is sobering for anyone who assumes that standardized infrastructure guarantees standardized care. Comparable stroke-care infrastructure, they report, did not ensure uniform recanalization therapy utilization. The variation between the two centers was driven primarily by how mild stroke patients were managed, highlighting a zone of clinical uncertainty where specialists at different hospitals can reach very different decisions about similar patients.

As a retrospective two-center study, the analysis has inherent limitations. It captures a six-month window at two hospitals in a single small country. The researchers used nonparametric and categorical tests for demographic, treatment, NIHSS and modified Rankin Scale data, and multivariable logistic regression to identify predictors of both recanalization therapy use and stroke outcome.

The findings carry a broader message about how stroke-care quality should be judged. A center with a high recanalization rate is not automatically delivering better outcomes, and a center with a lower rate is not necessarily withholding effective care, at least on the short-term measures examined here. Understanding why centers differ, the study suggests, requires looking beyond the headline numbers to the specific patient subgroups where treatment decisions are most contested.

Subject of Research: Inter-center variation in recanalization therapy use for acute ischemic stroke in Estonia

Article Title: Beyond the numbers: understanding inter-center variation in recanalization therapy rates

Article References: Quzeih, Y., Kolde, A., Vibo, R., Tuularu, T., Sikk, K., & Kõrv, J. (2026). Beyond the numbers: understanding inter-center variation in recanalization therapy rates. Frontiers in Neurology, 17, Article 1898540. https://doi.org/10.3389/fneur.2026.1898540

Image Credits: AI Generated

DOI: 10.3389/fneur.2026.1898540

Keywords: stroke, recanalization therapy, mild stroke, thrombolysis, NIHSS, modified Rankin Scale, practice variation, Estonia, comprehensive stroke centers, acute ischemic stroke, stroke outcomes, Beyond