Esophagitis healing proportions are often incorrectly called the healing rate. The aim of this study was to compare different drug classes by expressing the speed of healing and symptom relief through a new approach. A fully recursive literature search to July 1996 identified 43 articles on gastroesophageal reflux disease (GERD) (7635 patients) meeting strict inclusion criteria: single- or double-blind randomized studies in adults with endoscopically proven erosive or ulcerative esophagitis. For each drug class, linear regression analysis estimated the average percentage of patients who were healed and heartburn free per week. Mean overall healing proportion irrespective of drug dose or treatment duration (< or =12 weeks) was highest with proton pump inhibitors (PPIs; 83.6% +/- 11.4%) vs. H2-receptor antagonists (H2RAs; 51.9% +/- 17.1%), sucralfate (39.2% +/- 22.4%), or placebo (28.2% +/- 15.6%). Correcting for patients without baseline heartburn, the mean heartburn-free proportion was highest with PPIs (77.4% +/- 10.4%) vs. H2RAs (47.6% +/- 15.5%). PPIs showed a significantly faster healing rate (11.7%/wk) vs. H2RAs (5.9%/wk) and placebo (2.9%/wk). PPIs provided faster, more complete heartburn relief (11.5%/wk) vs. H2RAs (6.4%/wk). More complete esophagitis healing and heartburn relief is observed with PPIs vs. H2RAs and occurs nearly twice as fast. This semiquantitative expression of speed of healing and symptom relief permits comparisons for future economic evaluation and quality-of-life assessments.