A Systematic Overview of the Use of Diary Cards, Quality-of-Life Questionnaires, and Psychometric Tests in Treatment Trials of Helicobacter pylori-Positive and -Negative Non-Ulcer Dyspepsia

Abstract

Background: Our aim was to evaluate the use of diary cards, quality-of-life questionnaires, and psychometric tests in treatment trials of non-ulcer dyspepsia. Methods: Data sources were a Medline search (up to 1966) and a manual search of five gastrointestinal journals (up to 1980) for original, randomized, double-blind, placebo-controlled trials with at least 20 patients which evaluated treatment regimens for non-ulcer dyspepsia. Results: Of the 67 eligible studies, 31 used diary cards. Diary cards were used alone in 15 of the 31 studies (48%), whereas the others (52%) also used a physician assessment. The symptoms assessed by diary cards were epigastric pain (100%), nausea/vomiting (65%), heartburn (52%), belching (39%), regurgitation (29%), fullness (29%), and bloating (23%). Forty-five per cent also recorded antacid use. Severity of outcome measures was assessed by a visual analogue scale in 5 of the 31 studies (16%), Likert scales in 17 studies (55%), and unclear methods in 3 studies (10%). For statistical analysis daily averages of symptoms were used in 5 of the 31 studies (16%), weekly averages in 11 studies (35%), and 2-week intervals during the treatment period in the rest, with some studies using a combination (such as daily and weekly averages). Only 3 of the 31 studies (8%) checked for compliance with diary card data. None of the studies mention anything about missing data and how this was handled. One study evaluated quality of life questionnaires and one evaluated a psychometric test. Conclusions: Non-ulcer dyspepsia treatment trials frequently use diary cards but need to be much clearer about how information was obtained and how it was used in the statistical analysis. Not much information is available to comment on the use of quality-of-life questionnaires or psychometric tests for evaluation of outcome measures.