Should TIA patients be hospitalized or referred to a same-day clinic?: a decision analysis.

Abstract

For patients presenting with TIA, a previous study concluded that hospitalization is cost-effective compared to discharge without treatment from the emergency department. We performed a cost-effectiveness analysis of hospitalization vs urgent clinic evaluation following TIA. Among a cohort of TIA patients, we created a decision tree model to compare the decision to hospitalize or refer to urgent-access specialty clinic. We estimated probabilities, utilities, and direct costs from the available literature and calculated incremental cost-effectiveness ratio (ICER). We assumed equal access to standard medical treatments between the 2 approaches; however, we estimated higher tissue plasminogen activator (tPA) utilization among hospitalized patients. We performed sensitivity analyses to assess all assumptions in our model. In patients with TIA aged 65-74 years, hospitalization yielded additional 0.00026 quality-adjusted life-years (QALYs) at 1 year, but at an additional cost of $5,573 per patient compared to urgent clinic evaluation (ICER = $21,434,615/QALY). Over 30 years, the ICER was $3,473,125/QALY. These results were not sensitive to varying 48-hour stroke risk, length of stay, tPA utilization rate, QALYs saved per tPA treatment, and hospitalization and clinic costs, and cost saved per tPA treatment. Despite increased access to tPA in the hospital, we found that hospitalization is not cost-effective compared to same-day clinic evaluation following TIA. A very small fraction of patients benefits from hospitalization if urgent-access TIA clinics are available. The widespread development of urgent-access TIA clinics is warranted.