Avoid Single Fecal Occult Blood Testing When Screening for Colorectal Cancer
By Bruce Sylvester
Recent evidence from 2 separate sources published simultaneously in the Annals of Internal Medicine have uniformly suggested the inaccuracy of the single, precolonoscopic and in-office fecal occult blood test (FOBT) as a more convenient substitute for the 6-sample series.
A prospective study (2005;142:81-85) of 3121 symptomatic patients at 13 VA centers nationwide compared the sensitivity of FOBT obtained during digital rectal exam (DRE) with that of the 6-sample testing for advanced neoplasia and its specificity for no neoplasia. “We calculated predictive values and likelihood ratios for advanced neoplasia, defined as tubular adenomas 10 mm or more, adenomas with villous histology or high-grade dysphasia, or invasive cancer,” Judith F. Collins, MD, of the Portland Medical Affairs Medical Center, Portland, OR, and colleagues, wrote.
For patients who were asymptomatic of colorectal cancer, the single, office-based FOBT failed to detect advanced precancerous lesions in 95% of the time. “Single digital FOBT is a poor screening method for colorectal neoplasia,” they concluded, “and cannot be recommended as the only test.”
An accompanying survey (2005;142:86-94) of 1147 primary care physicians and 11,365 patients aged ³50 who were screened for colorectal cancer showed that 90% of physicians surveyed used FOBT obtained by DRE at least once monthly and 32.5% of them used the single FOBT as the exclusive diagnostic tool for colorectal cancer. Moreover, about 30% of these physicians followed a positive FOBT with yet another FOBT rather than with an immediate colonoscopy.
“The most important new finding of this article is that testing for blood in stool obtained during a digital rectal examination is a very poor screening test for colon cancer,” Harold Sox, MD, editor of the Annals of Internal Medicine and author of the editorial (2005;147:146-148), told IMWR. “Unfortunately, the [survey] shows that many physicians use this method. The clinical message is to stop using the single sample as a screening test for colon cancer and start sending the patient home with instructions on how to obtain 6 samples of stool, a proven method that reduces the death rate from colon cancer,” Dr Sox advised.
Dr Sox also added that the study and survey findings, “identify another potential culprit in the surprisingly low decline in colorectal cancer mortality over the recent decade: clinicians appear to be substituting a poor test for the tests that the evidence supports….These 2 studies discredit the office-based, single-sample screening test for occult blood….Taken together, they send a strong message to primary care physicians to reexamine their colorectal screening practices. Perhaps we need to put the guaiac cards in a locked drawer labeled ‘use only in case of emergency.’ ”
In an earlier study (N Engl J Med. 2004;351:2704-2714), fecal DNA analysis was shown to detect a greater proportion of important colorectal neoplasia than does FOBT. Stool sample testing was obtained from 2507 average-risk patients aged ³50 years who had undergone colonoscopy. Of 31 invasive cancers, 16 (51.6%) were detected by DNA testing compared with only 4 (12.9%) by FOBT (P = .003). DNA and FOBT detected 40.8% and 14.2%, respectively, of invasive cancers plus adenomas (P <.001). Among patients with advanced neoplasia, DNA testing detected 18.2% and occult blood testing detected 10.8%. Specificity in patients with negative findings on colonoscopy was 94.4% for DNA testing and 95.2% for occult blood testing