Rating scales for heartburn (HB) studies

Abstract

Categorical or VAS scales at fixed intervals after a provocative meal (PM) have been used to show H2-receptor antagonists (H2RAs) prevent meal-induced HB. We investigated measurement characteristics of a 10-point ordinal rating (1= mildest ever through 10= worst ever) versus validated 4-category Likert rating (none, mild, moderate, severe). Methods: Double-blind parallel study with 282 HB-sufferers randomized to take famotidine (fam) 20 or placebo before 4 self-selected PMs eaten at home. HB intensity rated on 4-point scale q30 minutes post-meal for 3 hours, then maximum HB during the 3 hours rated on 10-point scale (value of 0 assigned if no HB). Construct validity assessed using Spearman correlation (SC) between the scales. Discriminant validity assessed by comparing mean of 10-point scale scores among groups of meal sessions with none, mild, moderate, or severe peak HB. Responsiveness between scales compared using effect size. Results: Peak HB with fam 20 significantly less than placebo according to periodic 4-point and single 10-point ratings. SC coefficient between scales was 0.943. Mean scores on 10-point scale show clear distinction among categories for peak HB on 4-point scale. Effect size with 10-point scale (−0.249) was virtually identical to that with 4-point scale (−0.252). Conclusion: Ability of single ordinal rating of peak HB to differentiate between treatments is similar to that seen with 4-category Likert rating. Clinical Pharmacology & Therapeutics (2004) 75, P56–P56; doi: 10.1016/j.clpt.2003.11.213