The epidemiology of the gastrointestinal randomized clinical trial

Has the randomized clinical trial (RCT), generally accepted as the method of choice for evaluation of most treatments, obtained a footing in gastroenterology? Among 35,228 citations on gastroenterologic therapy indexed in MEDLARS 1964-1974 306 (0.9 per cent) were RCT's. During the decade their frequency rose significantly (P less than 0.05) from 0.3 per cent in 1964 to 1.7 per cent in 1973. The "typical" RCT was a double-blind two-group comparison of a new and an established drug on the symptoms of peptic ulcer; 50 patients were followed for six weeks, and the number of dropouts was unknown. The new drug was found to be more effective. It is postulated that the RCT's in gastroenterology are quantitatively and qualitatively insufficient and that co-ordinated planning of RCT's and uninhibited publication of statistically proved negative results may help to ensure that the patient with gastrointestinal disease receives the best available treatment.