Purpose: Studies suggest dyspepsia is common in the U.S. with as many as one in four reporting symptoms. Most of these people are thought to have functional dyspepsia (FD). While reported as common, recruitment of patients with FD to a clinical trial has been challenging. This study assessed the usefulness of identifying potential participants for a clinical trial using reported symptoms of dyspepsia from a community-based mail survey.
Methods: A survey mailed to 6481 randomly selected individuals from Olmsted County was used to identify people with symptoms of dyspepsia, gastroesophageal reflux disease (GERD), and irritable bowel syndrome (IBS) based on Rome III criteria. People were able to actively decline phone contact via return letter. Calls were made to survey respondents reporting ≥1 of four dyspepsia symptoms at least one day/week to identify potential subjects for recruitment to a clinical trial comparing antidepressant use in dyspepsia. A randomly selected subset of survey respondents reporting no symptoms or symptoms less than ≥1 day/week were also contacted by phone (“controls”). The Kappa coefficient (95% CI) for agreement (survey vs. phone follow-up) was used to estimate the reliability of identifying people meeting criteria for dyspepsia.
Results: A total of 3282 people (51%) responded to the survey. Phone contact was made with 454 (of a total 937 selected to be called) persons. Of the 937, 50 (5%) had requested no call, 8 (1%) had died, and 52 (6%) had no active phone number yielding a contact rate of 55%. The mean (±SD) age of the total survey respondents was 63 (±15) and 54% were female compared to 63 (±13) in those contacted by phone with 63% female. Agreement between survey response and phone interview was modest (kappa=0.42 [0.33, 0.50]) with the proportion implied by the survey (46%) higher than those identified by phone interview (38%) (p<0.01, McNemar's test). Of the 454 contacted, 375 (83%) were ineligible, 67 (15%) declined/were not interested, and 12 (3%) were eligible and interested (6 screened and 4 randomized). Primary reasons for ineligibility were insufficient symptom severity (58%), age >75 (26%), and GERD or IBS overlap (6%).
Conclusion: While symptoms were commonly reported on a mailed survey, they may overestimate people with dyspepsia compared to screening by phone. The generalizability of the clinical trial results may be limited since few in a community-based sample with dyspepsia symptoms actually qualified for the study. This study was partially funded by NIDDK U01DK 065713.