DIAGNOSTIC STRATEGIES IN THE MANAGEMENT OF ACUTE UPPER GASTROINTESTINAL-BLEEDING - PATIENT AND PHYSICIAN PREFERENCES

Background: Routine diagnostic endoscopy is commonly obtained for low-risk patients with acute upper gastrointestinal bleeding despite evidence from controlled trials that it does not improve patient outcome. Objective: To determine whether endoscopy is being overutilized for these patients or whether considerations not addressed in the trials could justify its widespread use. Methods: Twenty-five patients recovering from a recent hemorrhage and 22 primary care physicians used the analytic hierarchy process to perform an individualized analysis regarding the diagnostic management of acute upper gastrointestinal bleeding. The results were used to determine whether routine endoscopy was consistent with each subject's interpretation of relevant data and judgments regarding the relative importance of five management goals. The authors then compared the proportion of subjects in each group who preferred endoscopy over two non-endoscopy strategies (upper gastrointestinal series and no routine diagnostic test) with 85%, the current rate of endoscopy at the authors' hospital. Results: Endoscopy was preferred by 92% (97.5% CI: 70%-98%) of the patients and 55% (97.5% CI: 31%-77%) of the physicians. The patients ranked identifying the cause of bleeding the second most important management goal after avoiding a poor outcome from the acute bleeding episode. Conclusions: The current rate of diagnostic endoscopy is higher than would be expected based on physicians' preferences but quite consistent with patients' preferences. Patients regard knowledge of the bleeding site as important, even if this information will not affect management or prognosis. Attempts to assess the use of diagnostic endoscopy and other diagnostic tests should take both patient preferences and the pure value of diagnostic information into account.