Determining intravenous rt-PA eligibility in the Emergency Department

The purpose of this study was to assess the agreement of Emergency Department (ED) attendings, ED residents, and neurology residents compared with stroke neurologists in the assessment of intravenous rt-PA eligibility. A convenience sample of patients presenting with possible stroke symptoms to the University of Michigan Hospital ED from June 2003 to July 2004 was identified. A physician from each of four groups: ED attending, ED resident, neurology resident, and stroke neurology attending independently evaluated each patient for eligibility for intravenous (i.v.) rt-PA. Accuracy, sensitivity, and positive predictive value (PPV) with 95% confidence intervals (CI) were calculated by physician type, compared with the stroke neurologist, for eligibility for i.v. rt-PA. Exactly 36 (49%) out of the 73 evaluated patients were diagnosed with acute ischemic stroke and 11 were deemed eligible for treatment with i.v. tPA by the stroke neurologist. Agreement with the stroke neurologist for rt-PA eligibility was 93% [95% CI: 84%, 98%] (sensitivity = 82% [48%, 98%], PPV = 82% [48%, 99%]) for the ED attendings, 79% [65%, 90%] (sensitivity = 75% [35%, 97%], PPV = 43% [18% 71%]) for the ED residents, and 84% [73%, 92%] (sensitivity = 100% [74%, 100%], PPV = 52% [31%, 73%]) for the neurology residents. There were two false positive cases identified by ED attendings, eight, by ED residents, and 11 by neurology residents. This study suggests that the agreement between ED attendings and stroke neurologists for determination of rt-PA eligibility is good. There is room for improvement, however, in the determination of acute stroke therapy eligibility in the ED setting especially among trainees.