AbstractBackground and Purpose—
Understanding physician decision making is increasingly recognized as an important topic of study, especially in stroke care. We sought to characterize the process of acute stroke decision making among neurologists in the United States and Canada from clinical and epistemological perspectives.Methods—
Using a factorial design online survey, respondents were presented with clinical data to mimic an acute stroke encounter. The history, examination, computed tomographic (CT) scan, CT angiogram, and CT perfusion were presented in sequence, and respondents rated their diagnostic confidence and likelihood of treatment with tissue-type plasminogen activator after each element. Patient age, race, sex, and CT perfusion imaging results were randomized, whereas the rest of the clinical presentation was held constant.Results—
We collected 715 responses, of which 473 (66%) were complete. Diagnostic certainty and likelihood of treatment with tissue-type plasminogen activator rose incrementally as additional clinical data were provided. Diagnostic certainty and treatment likelihood were strongly influenced by the clinical history and the CT scan. Other factors such as physicians’ personal beliefs or biases were not influential. Respondents’ accuracy in interpreting CT angiographic and CT perfusion images was variable and generally low.Conclusions—
Diagnostic certainty and likelihood of treatment with tissue-type plasminogen activator increase with additional clinical data, with the history being the most important factor for diagnostic and treatment decisions. Respondents had difficulty in interpreting the results of CT perfusion scans although they had little impact on treatment decisions. We did not identify treatment bias based on patient age, race, or sex.