Background: Symptoms of gastro-oesophageal reflux are common, and currently available methods for diagnosing reflux disease are expensive and uncomfortable for the patient. The diagnostic value of a treatment test with omeprazole is unclear.
Methods: Patients with dyspepsia including heartburn admitted for upper gastrointestinal endoscopy were studied in a prospective, randomized, double-blind Scandinavian multicentre study. Before entry 188 patients were enrolled, and 160 were randomized to 1-week treatment with 20 mg omeprazole twice daily or placebo. Gastro-oesophageal reflux disease (GERD) was defined as reflux oesophagitis Savary-Miller grades II-III at endoscopy or pH < 4 exceeding 4% of the total time at 24-h oesophageal pH-monitoring and was found in 135 patients. The treatment test was considered positive when the patient's symptoms improved during the treatment week compared with the pretreatment day.
Results: The sensitivity in diagnosing reflux disease was 71-81% with omeprazole as a diagnostic test, compared with 36-47% for placebo during treatment days 3-7. The specificity was similar for the two treatment arms during the first days of the study. During the end of the week a larger proportion of the patients with normal endoscopy and pH test responded to omeprazole treatment, giving omeprazole lower specificity than placebo. The investigators' overall evaluation of whether the patient was a responder to the test had a sensitivity of 75% and a specificity of 55% in the omeprazole-treated patients. The corresponding figures in the placebo group were 17% and 92%, respectively.
Conclusion: One week of omeprazole treatment is a simple diagnostic test with a fairly high sensitivity. The specificity is poor owing to the placebo effect and to the lack of a gold standard in diagnosing reflux disease.