WASHINGTON—Screening for colorectal cancer traditionally has been done on a per-visit basis at the VA. A patient comes in and, based on the physician’s recommendation, undergoes a test for the disease.
Researchers at the Albuquerque, NM, VAMC suggested there was a better way and set out to prove it.
“For years, we’ve been obtaining population-based registries,” said Richard Hoffman, MD, a VA physician and researcher, as he presented his recent research into CRC screening at the annual VA Health Services R&D conference. “We always have current information on demographics, as well as files on problem-lists, labs, pathology and patient procedures. At any point, we could identify patients who need screening.”
Could a population-based screening method be viable? This was one question that Hoffman and his colleagues wanted to answer. They also wanted to know if the fecal immunochemical test (FIT) was as efficacious as the standard guaiacum fecal occult blood test (gFOBT). FIT has been found to be more sensitive than gFOBT, requires no medication or dietary restrictions and involves fewer tests.
Between 2008 and 2009, researchers looked at patients in the New Mexico VA Healthcare System—which had the lowest CRC screening rate in its Veterans Integrated Service Network. Participants were all between 50 and 80 years old, were due for a CRC screening according to VA guidelines and had no history of CRC. The researchers used the VA registry to identify all patients that fit those criteria as of May 2008. They then mailed invitation letters with return postcards, inviting patients to enter the study.
Those that replied saying they wanted to participate were randomly assigned test kits—202 were given gFOBT kits and 202 were given FIT kits. There were also three control groups. The first control group was made up of 3,184 patients not invited to be part of the test group. The second group consisted of 2,525 patients who were sent invitations but did not respond. The third group was made up of 255 individuals who could not be contacted.
The gFOBT test involves smearing feces on an absorbent paper treated with a chemical. Patients are required to repeat this process over three consecutive days. They are also asked to avoid ingesting vitamin C supplements, citrus fruit and citrus juices for three days before the test and during the stool sample collection period. The samples are sealed and mailed back to the VA.
When the test goes to the lab, hydrogen peroxide is dropped on the paper. The heme component of hemoglobin breaks down the hydrogen peroxide, so if trace amounts of blood are present, the paper will quickly change color.
The mechanics of the FIT test are similar, insofar as the patient smears feces on a piece of paper. However, the FIT test uses antibodies to detect the presence of globin. Patients are given no dietary restrictions and only have to complete two tests, rather than three with the gFOBT kit.
One complication in the study, but one that would eventually provide more data, was that VA did not accept FIT as an acceptable colorectal screening test at the time. “We were forced to send anyone who sent back a negative FIT test a set of gFOBT cards and have them repeat the process. Anyone with a positive FIT [or gFOBT] was sent for a colonoscopy,” Hoffman said.
Kit Vs. Kit
Adherence was significantly higher in the FIT group compared to gFOBT —62% compared to 55%. And of those patients who took both (78 in all), 62% said they preferred FIT compared to the 12% that preferred gFOBT. The reasons patients gave for preferring FIT were the lack of dietary restrictions and easier-to-follow instructions.
The control groups—those who were never sent test kits and who only were screened if they visited a VA facility during the period of the study and were offered screening—had a significantly lower number of patients who received screening than the by-mail group. This led Hoffman and his colleagues to conclude that a population-based strategy appears more effective than visit-based for achieving screening.
“Using the population registry for screening fits better with our veteran-centered philosophy of care,” Hoffman said.
Following the study, the Albuquerque VAMC adopted FIT as an acceptable CRC screening tool. However, the decision between which screening tool is used more frequently may end up being based on cost.
A gFOBT test kit costs $1, while a FIT test kit costs $9. Even though widespread use of FIT may result in discounts for VA facilities buying in bulk, it is unlikely that the cost would drop to the level of the gFOBT test. Also, the positive predictive value of both tests was about the same, Hoffman said.