Abstract T P193: Impact of Observing TIA Patients in a Clinical Decision Unit

Abstract

Background: TIA carries a high risk of early stroke. National guidelines recommend that TIA cases should have rapid work up, ideally within 24 hours. Hospitalization is recommended when rapid outpatient work-up is not possible. As a result, most TIA patients are managed in the hospital. However, recent literature supports the use of observation in the Emergency Department for testing and monitoring of TIA patients. Here, we review the effects of our protocol for the observation of TIA cases in a Clinical Decision Unit (CDU) at our Comprehensive Stroke Center (CSC). Methods: Retrospectively, we reviewed the charts of all TIA cases, from one year prior to the use of a CDU protocol, to one year after its use. Patients were identified by billing codes, using our informatics system. Data on location of care, final diagnosis, cost of care, contribution margin, admission rate, and imaging test was collected and analyzed. Descriptive and analytical statistics were used to compare groups. Results: From May 2011 to May 2013, 400 suspected TIA cases were seen. Prior to the initiation of our protocol, 209 TIA patients were hospitalized. After the protocol, 191 patients were seen, with 85 patients seen in the CDU and 106 patients being admitted. The inpatients were mostly cases transferred to our CSC from regional facilities. Of the 85 CDU cases, 51 had a final diagnosis of TIA, 7 had confirmed strokes, 25 were non-cerebrovascular, and 2 were transient global amnesia. From the CDU, 20 patients were admitted, due to confirmation of stroke, recurrence of symptoms, or need for urgent intervention. Over two thirds of all patients underwent MRI/MRA for neuroimaging; the rest had non-contrast CT and carotid ultrasound. Since the start of the TIA protocol, the direct cost of care per case was $453.65 less (p=0.02) in the CDU cases. During the same time period, there was no significant difference in the contribution margin between the CDU and inpatient groups. Conclusion: Given the high risk for early stroke after TIA, early evaluation and monitoring remains imperative. Our results are similar to recent literature, showing that similar management can be provided for less cost and similar clinical outcomes. Larger studies are necessary to determine the overall financial impact of such programs.