Effects of a Province-wide Triaging System for TIA: The ASPIRE Intervention.

BACKGROUND AND OBJECTIVES: Urgent transient ischemic attack (TIA) management to reduce stroke recurrence is challenging, particularly in rural and remote areas. In Alberta, Canada, despite an organized stroke system, data from 1999 to 2000 suggested that stroke recurrence after TIA was as high as 9.5% at 90 days. Our objective was to determine whether a multifaceted population-based intervention resulted in a reduction in recurrent stroke after TIA. METHODS: In this quasi-experimental health services research intervention study, we implemented a TIA management algorithm across the entire province, centered around a 24-hour physician's TIA hotline and public and health provider education on TIA. From administrative databases, we linked emergency department discharge abstracts to hospital discharge abstracts to identify incident TIAs and recurrent strokes at 90 days across a single payer system with validation of recurrent stroke events. The primary outcome was recurrent stroke; with a secondary composite outcome of recurrent stroke, acute coronary syndrome, and all-cause death. We used an interrupted time series regression analysis of age-adjusted and sex-adjusted stroke recurrence rates after TIA, incorporating a 2-year preimplementation period (2007-2009), a 15-month implementation period, and a 2-year postimplementation period (2010-2012). Logistic regression was used to examine outcomes that did not fit the time series model. RESULTS: We assessed 6,715 patients preimplementation and 6,956 patients postimplementation. The 90-day stroke recurrence rate in the pre-Alberta Stroke Prevention in TIA and mild Strokes (ASPIRE) period was 4.5% compared with 5.3% during the post-ASPIRE period. There was neither a step change (estimate 0.38; p = 0.65) nor slope change (parameter estimate 0.30; p = 0.12) in recurrent stroke rates associated with the ASPIRE intervention implementation period. Adjusted all-cause mortality (odds ratio 0.71, 95% CI 0.56-0.89) was significantly lower after the ASPIRE intervention. DISCUSSION: The ASPIRE TIA triaging and management interventions did not further reduce stroke recurrence in the context of an organized stroke system. The apparent lower mortality postintervention may be related to improved surveillance after events identified as TIAs, but secular trends cannot be excluded. CLASSIFICATION OF EVIDENCE: This study provides Class III evidence that a standardized population-wide algorithmic triage system for patients with TIA did not reduce recurrent stroke rate.

The Alberta Stroke Prevention in TIAs and mild strokes (ASPIRE) intervention: rationale and design for evaluating the implementation of a province-wide TIA triaging system.

RATIONALE: Stroke risk after transient ischaemic attack is high and, it is a challenge worldwide to provide urgent assessment and preventive services to entire populations. AIMS: To determine whether a province-wide transient ischaemic attack Triaging algorithm and transient ischaemic attack hotline (the Alberta Stroke Prevention in transient ischaemic attacks and mild strokes intervention) can reduce the rate of stroke recurrence following transient ischaemic attack across the population of Alberta, Canada (population 3·7 million, 90-day rate of post-stroke transient ischaemic attack currently 9·5%). It also seeks to improve upon current transient ischaemic attack triaging rules by incorporating time from symptom onset as a predictive variable. DESIGN: The transient ischaemic attack algorithm and hotline were developed with a broad consensus of clinicians, patients, policy-makers, and researchers and based on local adaptation of the work of others and research and insights developed within the province. Because neither patient-level nor region-level randomization was possible, we conducted a quasi-experimental design examining changes in the post-transient ischaemic attack rate of stroke recurrence before and after the 15-month implementation period using an interrupted time-series regression analysis. The design controls for changes in case-mix, co-interventions, and secular trends. A prospective transient ischaemic attack cohort will also be concurrently created with telephone follow-up at seven-days and 90 days as well as passive follow-up over the longer term using linkages to provincial healthcare administrative databases. STUDY OUTCOMES: The primary outcome measure is the change in recurrence rate of stroke following transient ischaemic attack at seven-days and 90 days, comparing a period of two-years before vs. two-years after the intervention is implemented. All cases of recurrent stroke will be validated. Secondary outcomes include functional status, hospitalizations, morbidity, and mortality. CONCLUSIONS: We are undertaking a rigorous evaluation of a population-based approach to improving quality of transient ischaemic attack care. Whether positive or negative, our work should provide important insights for all potential stakeholders.