This study will compare patient adherence to different colorectal cancer (CRC) screening
tests to identify the most cost-effective strategy to decrease mortality from CRC. We
hypothesize that different types of tests will have different adherence rates, that these
rates will alter the cost-effectiveness analysis, and that we can identify patient variables
associated with non-adherence to specific CRC screening strategies.
Colorectal cancer (CRC) is a significant and preventable disease, yet CRC screening rates
remain low. Previous investigators have identified barriers to adherence to CRC screening;
however, the majority of data have been retrospectively derived, and the limited data
produced from prospective assessment have been limited to fecal occult blood testing (FOBT).
Constructs based on the Health Belief Model have been proposed to identify items associated
with non-adherence to CRC screening, but prospective validation of this model is lacking.
The U.S. Preventive Services Task Force report cites a major cause of uncertainty for
calculation of the incremental cost-effectiveness of CRC screening is the deficit in primary
data regarding adherence to CRC screening tests, specifically whether heterogeneity exists
in screening rates of competing strategies. The Institute of Medicine confirmed the
importance of detecting heterogeneity in adherence between strategies, understanding that
some strategies currently recommended for CRC screening may be dominated by strategies that
achieve greater levels of adherence.
This study aims to determine whether adherence rates to CRC screening are heterogeneous
between competing strategies (FOBT and colonoscopy). This study will also prospectively
examine domains of the Health Belief Model to identify associations with non-adherence to
screening. Adherence rates specific to tested strategies will be incorporated in our
existing economic models to compare the incremental cost-effectiveness of competing CRC
screening strategies. These data will greatly impact policy decisions regarding resource
allocation for CRC screening. It is also expected that future research based on data
generated through this project will aim to develop and test interventions that optimize
adherence to screening strategies to decrease mortality from CRC.
Patients who are due for CRC screening and meet eligibility requirements are identified
through a query of the electronic medical record database at San Francisco General Hospital
(SFGH). A research assistant (RA) obtains the PCP's approval to attempt recruitment at the
patient's primary care appointment. The PCP discusses CRC screening with the participant
during their regularly scheduled appointment.
Availability of CRC screening tests: Because of capacity constraints in the endoscopy unit
at SFGH, the screening method for those at average risk of CRC had been limited to annual
fecal occult blood testing (FOBT). However, the gastroenterology department initiated a
pilot program which allows different primary care clinics to refer average-risk patients for
colonoscopy screening in rotating 3-month time-blocks. To ensure the endoscopy unit has
sufficient capacity to provide CRC screening via colonoscopy, providers in a given primary
care clinic are able to refer their patients for (a) colonoscopy screening, (b) FOBT
screening, or (c) a choice of either colonoscopy or FOBT screening, depending on the time
block. This is not a study intervention; providers simply recommend that their patients
complete a standard CRC screening test, and discuss the option or options available.
Patients who decline to participate in the study undergo colorectal cancer screening under
the guidance of their primary care provider; the same screening tests are available to those
who participate and those who do not.
After giving written informed consent, participants complete an RA-administered 20-minute
survey based on constructs of the Health Belief Model. Participants also grant us approval
to review their medical records in one year to determine if they complete screening; those
without a record of testing are contacted to determine if they completed testing outside of
- Average-risk subjects (no family history of CRC, no personal history of polyps or
- 50 years of age or greater, but less than 80 years old.
- Due for CRC screening.
- Upcoming appointment scheduled with primary care provider.
- Primary care provider has agreed to refer patients for consideration of enrollment in
- Family history of CRC in a first-degree relative.
- Personal history of colonic adenomatous polyps, CRC or inflammatory bowel disease.
- Symptoms for which colonoscopy or sigmoidoscopy would otherwise be performed
(hematochezia, new onset diarrhea or constipation, abdominal pain).
- CRC screening current (FOBT within preceding 12 months, flexible sigmoidoscopy or
double contrast barium enema within 5 years, or colonoscopy within 10 years).
- Comorbid illness precluding endoscopic evaluation (coronary artery disease with
myocardial infarction within 6 months, unstable angina or congestive heart failure,
chronic obstructive pulmonary disease requiring home oxygen, other diseases that
limit life expectancy to less than 10 years).
Hal F Yee, MD, PhD
University of California, San Francisco