Am Fam Physician. 2005;72(8):1600-1601
Colorectal cancer screening options endorsed by major organizations include fecal occult blood testing (FOBT). This is based on clinical trials showing that a protocol of testing six samples obtained by a patient at home reduced morbidity and mortality. In practice, however, physicians often substitute FOBT of a single specimen obtained during a digital rectal examination in the office. Collins and colleagues conducted a prospective cohort study to compare the sensitivity and specificity of office FOBT with home FOBT.
The study population consisted of randomly recruited patients 50 to 75 years of age at primary care clinics in 13 Veterans Affairs medical centers. Nearly 97 percent of the patients were men. Those with pre-existing lower gastrointestinal tract disease and abdominal symptoms that required a medical evaluation were excluded. Enrolled patients underwent a complete physical examination, including digital rectal examination. Office FOBT was performed on one stool sample. Patients then completed home FOBT and underwent colonoscopy of the cecum. Visible polypoid lesions were removed or biopsied and sent to pathology. Results were interpreted independently by three pathologists who were blinded to FOBT results. Of the 3,121 original patients, 2,665 completed all parts of the study and were included in the analysis.
More than one half of all patients had polypoid lesions identified on colonoscopy examination, and 284 patients (10.7 percent) were found to have pathology consistent with advanced neoplasia (i.e., tubular adenoma of at least 1 cm, villous adenoma, high-grade dysplasia, or cancer). A positive result on home FOBT had 23.9 percent sensitivity and 93.8 percent specificity for advanced neoplasia; in contrast, office FOBT had 4.9 percent sensitivity and 97.1 percent specificity. Although a positive result on home FOBT indicated that a patient was nearly four times as likely to have advanced neoplasia, a positive result on office FOBT was statistically insignificant; a negative result did not affect likelihood of neoplasia. Finally, adding the office test to the home test did not improve the sensitivity of home FOBT.
The authors conclude that their results do not support the practice of office FOBT of a single stool sample for colorectal cancer screening. Based on a trend toward increased likelihood of advanced neoplasia in patients with positive results on office FOBT, they recommend a follow-up colonoscopy (rather than repeat testing) for these patients. Because a negative result on office FOBT is essentially meaningless, they recommend follow-up home FOBT or lower endoscopy.
editor’s note: The significance of this study is underlined in an accompanying report by Nadel and colleagues,1 who surveyed random samples of primary care physicians and patients nationwide regarding FOBT. Nearly 30 percent of patients who reported having received screening had undergone only office FOBT. This result could not be attributed to poor patient adherence because almost one third of physicians reported using office FOBT only, which fails to detect more than 95 percent of advanced neoplasia. Nearly the same number of physicians said that they would follow up a positive office FOBT result with a second office FOBT rather than proceed to colonoscopy. In an accompanying editorial, Sox2 observes that physicians’ failure to screen large numbers of patients for colorectal cancer has been compounded by a disturbing tendency to screen with a test that is practically useless. He calls on physicians to abandon office FOBT as a screening practice and redouble efforts to get patients to comply with home FOBT.2—k.w.l.