Dyspepsia in the community: value of a community-based mailed survey to identify potential participants for a randomized clinical trial

Abstract

Objective To assess the usefulness of a community-based mailed survey to identify participants with functional dyspepsia (FD) for a clinical trial. Material and Methods In 2008, a valid self-report questionnaire of gastrointestinal (GI) symptoms required for diagnosis of FD was mailed to randomly selected cohorts of Olmsted County, Minnesota residents. From survey responses (54%), FD cases and controls were identified. Phone calls were completed in 2010 and 2011 to 54% of respondents offering participation to those meeting criteria. Results Of 937 people identified from the survey, 189 cases and 265 controls were contacted by phone using four questions similar to the written survey resulting in a moderate level of agreement (Kappa 0.43, 95% CI: 0.35 to 0.51; p =0.11). The proportion reporting FD symptoms by survey was 42% while the proportion by phone was 38%. Comparing classification of cases and controls, 118 (62%) survey cases had dyspepsia symptoms on phone screening while 53 (20%) of the survey controls reported FD symptoms. Of 171 who had symptoms, 60 (35%) declined, 33 (19%) were over study age limit, 24 (14%) had inadequate symptom levels, and 36 (21%) had co-morbidities. Of survey respondents contacted, 6 (3%) people were enrolled with 2 screen fails resulting in 4 (1%) randomized. Conclusions Agreement between survey and phone questions was modest. Classifications between case and control changed. People eligible and willing to participate were a fraction of people reporting symptoms. People participating in clinical trials do not broadly represent those in the population.