The Bigger the Better? Center Volume Dependent Effects on Procedural and Functional Outcome in Established Endovascular Stroke Centers

Abstract

Background: Mechanical thrombectomy (MT) rates for the treatment of acute ischaemic stroke due to large vessel occlusion are steadily increasing, but are delivered in heterogenic settings. We aim to investigate effects of procedural load in centers with established MT-structures by comparing high- vs. low-volume centers with regard to procedural characteristics and functional outcomes.

Methods: Data from 5,379 patients enrolled in the German Stroke Registry Endovascular Treatment (GSR-ET) between June 2015 and December 2019 were compared between three groups: high volume: ≥180 MTs/year, 2,342 patients; medium volume: 135-179 MTs/year, 2,202 patients; low volume: <135 MTs/year, 835 patients. Univariate analysis and multiple linear and logistic regression analyses were performed to identify differences between high- and low-volume centers.

Results: We identified high- vs. low-volume centers to be an independent predictor of shorter intra-hospital (admission to groin puncture: 60 vs. 82 min, β = -26.458; p < 0.001) and procedural times (groin puncture to flow restoration: 36 vs. 46.5 min; β = -12.452; p < 0.001) after adjusting for clinically relevant factors. Moreover, high-volume centers predicted a shorter duration of hospital stay (8 vs. 9 days; β = -2.901; p < 0.001) and favorable medical facility at discharge [transfer to neurorehabilitation facility/home vs. hospital/nursing home/in-house fatality, odds ratio (OR) 1.340, p = 0.002]. Differences for functional outcome at 90-day follow-up were observed only on univariate level in the subgroup of primarily to MT center admitted patients (mRS 0-2 38.5 vs. 32.8%, p = 0.028), but did not persist in multivariate analyses.

Conclusion: Differences in efficiency measured by procedural times call for analysis and optimization of in-house procedural workflows at regularly used but comparatively low procedural volume MT centers.

Keywords: GSR-ET; center volume; endovascular stroke therapy; mechanical thrombectomy; procedural volume; stroke.

Figure 1

Patient selection and comparison grouping. MT, mechanical thrombectomy; GSR-ET, German Stroke Registry Endovascular Treatment. Colors: blue: GSR-ET enrolling centers and center-specific exclusion criteria; grey: resulting patient cohort and patient-specific exclusion criteria.

Figure 2

Procedural times depending on center volume tertile. (A) Study cohort including all patients. Longer pre-hospital and shorter intra-hospital procedural times in high- vs. low-volume centers. No difference in overall procedural time. (B) Subgroup analysis: only patients primarily admitted to MT center. Shorter pre-hospital, intra-hospital, and overall procedural times in high- vs. low-volume centers. Displayed: median with interquartile range (IQR). ***Bonferroni-corrected p ≤ 0.001 in univariate comparison high- vs. low-volume center resulting from the pairwise Kruskal–Wallis test. SO/TOR, symptom onset/time of recognition; ADM, admission; GRO, groin puncture; FLR, flow restoration.

Figure 3

Functional outcome at 90-day follow-up depending on center volume tertile. (A) Study cohort including all patients. No significant difference in good functional outcome in high- vs. low-volume center. (B) Subgroup analysis: only patients primarily admitted to MT center. Higher share of good functional outcome (mRS 0–2) at 90-day follow-up in high- vs. low-volume centers. Displayed: absolute share of mRS-based good functional outcome. Staded are p-values of univariate comparison high- vs. low-volume center resulting from the chi-square test. mRS, modified Rankin scale.