Background: Treatment trials for irritable bowel syndrome (IBS) usually define a responder as a patient who reports satisfactory relief or adequate relief of symptoms at the end of the trial. However, these measures have not been adequately validated.
Aims: (1) Compare a binary satisfactory relief measure to alternative ways of defining a treatment responder. (2) Determine whether baseline IBS symptom severity or psychological distress influence the sensitivity of these outcome measures.
Methods: A total of 350 patients (81% females, average age 50 yr) who had a medical diagnosis of IBS and satisfied Rome II criteria, were recruited from Group Health Cooperative of Puget Sound. At baseline the Irritable Bowel Severity Scale (IBSS) was used to assess symptom severity and to classify patients as mild, moderate, or severe. Psychological distress and IBS-specific quality of life (IBS-QOL) were also assessed. After 6 months treatment with standard medical care, IBSS and IBS-QOL were reassessed, and patients were asked whether they had experienced satisfactory relief and whether they were somewhat or markedly better.
Results: Initial severity of IBS significantly affected the proportion who reported satisfactory relief (mild, 72%; moderate, 53%; severe, 44%) and the proportion who were somewhat or markedly better (mild, 62%; moderate, 44%; severe, 38%), but did not affect the proportion with a 50% reduction in symptoms (mild, 26%; moderate, 25%; severe, 23%). Although mild patients were the most likely to report satisfactory relief, they showed no average decrease in symptom severity or improvement in IBS-QOL. Conversely, severe patients, who were the least likely to report satisfactory relief, had the largest reductions in IBS symptom severity and the largest improvements in IBS-QOL. Psychological distress had no significant effect on the responder rate after adjusting for IBS symptom severity.
Conclusions: These data from a descriptive study suggest that satisfactory relief is confounded with initial IBS symptom severity and is poorly correlated with the amount of symptom improvement. Confirmation of these findings in a clinical trial is needed.