Appropriate use of upper gastrointestinal endoscopy--a prospective audit. Steering Group of the Upper Gastrointestinal Endoscopy Audit Committee.

Work by this group has shown that there is a wide range of opinion as to patients' suitability for endoscopy. In a recent study, 1297 questionnaires were sent to a random selection of doctors, including 350 general physicians, 400 surgeons, 477 gastroenterologists, and 70 general practitioners. The respondent was asked to indicate whether or not he would refer the patient described by each case vignette for endoscopy. Depending on the indication, the positive referral rate varied from 4.5% to 99% overall, and from 4.5% to 63.8% for all those clinical situations that the working party felt to be inappropriate. A second study examined the appropriateness of 400 consecutive cases referred from four units within one health region; these cases were judged independently, and without conferring, by a panel of seven gastroenterologists. The same cases were rated by software that incorporated American opinion (the Rand criteria). Although only 45 (11%) of the cases were classed as inappropriate by the British panel, 120 cases (31%) assessed by the American software were rated inappropriate. These differences occurred largely because in the USA it is recommended that one month's antiulcer treatment be tried before considering endoscopy for dyspepsia and thus many referrals were seen as inappropriate by the American database. Of the 45 cases found to be inappropriate by the British doctors no important abnormality was found at endoscopy; whereas of 120 cases judged inappropriate by the Rand criteria, three duodenal and two gastric ulcers, and one gastric cancer were diagnosed at gastroscopy. This study attempts a quantitative assessment of inappropriate use and serves to encourage further work to define appropriateness.