Research Objective: To evaluate two related but distinct questions about stroke care quality: first, whether CSC certification itself improves 30-day survival among older AIS patients; and second, whether EVT procedural volume predicts better outcomes and at what threshold a meaningful survival benefit emerges. Together, these questions provide causal evidence to inform ambulance bypass policies that currently route patients to CSCs on the assumption that certification, rather than procedural experience, is the operative driver of stroke care quality.

Study Design: We used 2016–2019 100% Medicare fee-for-service claims to identify acute ischemic stroke (AIS) hospitalizations at short-term acute care hospitals, excluding transfer cases. For the certification analysis, we applied a staggered difference-in-differences design (Callaway & Sant'Anna estimator) exploiting variation in Comprehensive Stroke Center (CSC) certification timing across seven states, controlling for patient demographics and NIHSS stroke severity scores. For the volume analysis, we applied propensity score overlap weighting and swept annual endovascular thrombectomy (EVT) thresholds from 15 to 80 procedures per year in increments of five, dropping intermediate-volume hospitals at each comparison to produce clean high-versus-low contrasts. Standard errors were clustered at the hospital level.

Population Studied: Medicare fee-for-service beneficiaries aged 65 and older hospitalized with a primary AIS diagnosis from 2016 to 2019, comprising approximately 260,000 admissions annually across non-rural acute care hospitals.

Principal Findings: CSC certification was not associated with lower 30-day mortality at either the patient or hospital level. Results were robust to NIHSS adjustment. For EVT procedural volume, no mortality benefit was detectable at thresholds at or below 50 procedures per year, including the current CSC certification threshold of 25 procedures, with odds ratios ranging from 1.05 to 1.10 and confidence intervals crossing 1.0 throughout. Compared to being treated at hospitals that conduct 15 or fewer EVT per year, we only observed statistically significant benefits in 30-day mortality when patients are treated at hospitals that conducted more than 65 EVT procedures per year (OR = 0.85, P=0.047), a volume more than double the current certification requirement, with consistent patterns across age subgroups. Descriptively, EVT volume varies widely even among CSCs, and a meaningful share of PSCs and non-certified hospitals regularly exceed the 25-procedure certification threshold.

Conclusions: CSC certification as currently structured does not seem to be associated with any detectable survival benefit for Medicare AIS patients. EVT procedural volume does predict better outcomes, but only at thresholds substantially exceeding what the current certification policy requires. The certification label is neither necessary nor sufficient for high-quality stroke care, what matters is sustained high-volume procedural experience.

Implications for Policy or Practice: Current stroke triage protocols in some areas direct patients to CSCs on the assumption that certification signals superior care. Our findings challenge this assumption. Policymakers and accreditation bodies should reconsider whether the 25-procedure EVT threshold meaningfully distinguishes high-quality stroke care, and whether ambulance bypass policies routing patients to low-volume CSCs past higher-volume alternatives may be causing net harm. Certification criteria that emphasize sustained procedural volume over administrative designation may better serve patient outcomes.