Gastro-oesophageal reflux disease (GORD), including endoscopically confirmed reflux oesophagitis of grade II and above [termed severe erosive oesophagitis (SEO)], is a common but difficult to diagnose illness. Symptoms are often not closely correlated with pathological findings. A decision analysis found empirical therapy based on initial treatment with omeprazole to be a dominant strategy for SEO compared with ranitidine or phase I therapy. Phase I therapy, the most common treatment, includes antacids, cessation of smoking, reduction of alcohol consumption, weight loss, dietary changes, and elevating the head of the bed. A subsequent randomised controlled trial found that omeprazole was a more cost-effective strategy than ranitidine plus metoclopramide; clinical outcome was better in the omeprazole-treated group than in the ranitidine/metoclopramide group, while cost was similar in both groups. However, the changing of treatment strategies (e.g. modifying dosages and regimens), or new surgical interventions, require periodic re-evaluation of the cost-effectiveness of all alternative therapies for GORD.