Loss of predictive value of gastric ulcer symptoms in a randomized treatment trial.

Analysis of clinical data obtained in a double-blind randomized study, which compared liquid antacid (neutralizing capacity 120 mmol per day) with 1 g cimetidine in the treatment of 125 patients with gastric ulcer, revealed that, before starting treatment, 71% of the patients complained of epigastric pain, approximately 50% of bloating, and approximately 30% of nausea, heartburn, constipation or vomiting. Epigastric pain before treatment was significantly more frequent in patients with large ulcers (P less than 0.05) and in patients with ulcers unhealed after 4 weeks of therapy (P less than 0.05). This finding was the result of a highly significant correlation between diurnal epigastric pain and ulcer size and delayed healing (P less than 0.005). Nocturnal pain did not correlate with prognosis. In contrast to this correlation between pain before therapy and healing, the disappearance of epigastric pain with therapy did not signify ulcer healing. Only 14 (38%) of the 37 patients with healed ulcer were free from pain after the 4 weeks of therapy, whereas 25 (49%) of the 52 patients with persistent ulcers had no pain at this time. Placebo pain tablets relieved ulcer pain effectively in more than 85% of the patients, irrespective of whether the ulcer was healing or not. The other symptoms (bloating, nausea, heartburn, constipation or vomiting) were also alleviated by 4 weeks of therapy but no correlation was found with ulcer size or prognosis. The loss of the prognostic significance of ulcer pain is probably due to a complex interaction of the trial schedule on the patient's level of consciousness.(ABSTRACT TRUNCATED AT 250 WORDS)