Adnan I Qureshi, Hassan Raza, Akash Roy, Christy N Cassarly, M Fareed K Suri
In this small cohort, qualifying recurrences occurred predominantly early, but the estimates were imprecise. Without a stent placement comparator, the findings are hypothesis-generating and cannot establish comparative effectiveness or guide treatment selection.
BACKGROUND: Symptomatic extracranial vertebral artery stenosis (EVAS) is associated with recurrent vertebrobasilar arterial ischemic stroke, but recurrence under medical management remains unclear. We describe the 4-year frequency and timing of recurrent non-fatal vertebrobasilar arterial ischemic stroke among patients with EVAS who did not undergo vertebral artery stent placement.
METHODS: We conducted a retrospective cohort study using the HealthVerity Launch Sample, an administrative claims database. Eligible patients had an institutional claim from January 2020 through June 2021 containing both diagnoses and continuous enrollment for 1,461 days. The primary outcome was a new vertebrobasilar arterial ischemic stroke hospitalization as the primary or admitting diagnosis at least 30 days after index and within 4 years. Analyses were descriptive and exploratory and included Wilson confidence intervals (CIs), person-time incidence, Kaplan-Meier estimation, post hoc beta-binomial Monte Carlo simulation, and sensitivity analyses.
RESULTS: Among 43 patients, mean age was 71.5 ± 15.2 years, and 33 (76.7%) were men. Seven patients experienced recurrence within 4 years, corresponding to an event rate of 16.3% (95% CI, 8.1%-30.0%) and 4.71 events per 100 person-years. The posterior mean recurrence probability was 17.0% (95% uncertainty interval, 7.6%-29.3%). Cumulative incidence was 11.6% at 1 year, 14.0% at 2 years, and 16.3% at 3 and 4 years. Four recurrences occurred between days 33 and 45; none occurred after day 732.
CONCLUSIONS: In this small cohort, qualifying recurrences occurred predominantly early, but the estimates were imprecise. Without a stent placement comparator, the findings are hypothesis-generating and cannot establish comparative effectiveness or guide treatment selection.