BACKGROUND: Population-based screening for fecal occult blood has been shown to reduce mortality from colorectal cancer. Unfortunately, low fecal occult blood testing (FOBT) participation rates limit the potential impact of this screening intervention. One reason that patients choose not to complete and return their FOBT cards may be that they have difficulty following the recommended pretesting dietary restrictions. Substances that patients are often instructed to avoid before FOBT include red meat, fresh fruits and vegetables, vitamin C, iron, and nonsteroidal anti-inflammatory drugs.
POPULATION STUDIED: Participants in the studies included American Association of Retired Persons members (n=3783), patients of 32 Canadian Family Physicians (n=5003), patients in a single British general practice (n=153), patients aged 40 to 74 years not otherwise specified (n=634), and Veterans Affairs hospital patients (n=786).
STUDY DESIGN AND VALIDITY: Five randomized controlled trials were included in the meta-analysis. These trials were identified by a structured MEDLINE search augmented by hand searching and by contacting experts. All 5 studies randomized participants to either a dietary restrictions or no dietary restrictions arm before FOBT and reported screening completion rates for each arm. The authors do not specifically mention searching for unpublished trials, an important consideration in meta-analytic studies. Published studies are more likely to have positive results and this publication bias makes it more likely that the meta-analysis results will show a positive effect.
OUTCOMES MEASURED: The primary outcome was the difference between the completion rates of FOBT in the dietary restriction arm and the no dietary restriction arm. As a secondary outcome the investigators looked at positivity rates or the percentage of positive FOBT results among completed tests in each arm. Positivity rates are used here as a proxy for test specificity, which we would expect to be improved by dietary restriction.
RESULTS: When patients were counseled to avoid certain foods before obtaining a fecal sample, completion rates of FOBT across the studies ranged from 18.1% to 80.4%. This wide range of completion rates suggests heterogeneity in the interventions. It also suggests that factors other than dietary restriction may account for most of the difference in completion rates. Only one study showed a significant difference between the dietary restriction arm and the no dietary restriction arm. This study was the smallest of the 5 trials, and it had relatively complex dietary restrictions. There was no significant difference between positivity rates in any of the individual trials or in the pooled difference.