Patient’s Life Expectancy and the Need for Fecal Occult Blood Testing

A retrospective study of elderly US veterans who had a positive fecal occult blood test (FOBT) result found that only 56% underwent subsequent colonoscopy. Practice guidelines leave decisions regarding follow-up colonoscopy in patients aged >75 years to the discretion of the physician and patient following a positive FOBT result. The study findings lend support to this advice, indicating that follow-up colonoscopy tends to confer more harm than benefit in elderly patients with limited life expectancies. Just as follow-up colonoscopy is targeted to healthier older adults, the authors suggested a similar targeted approach for FOBT screening might also be appropriate [Arch Intern Med. 2011;171(15):1344-1351].

The researchers examined medical claims at 4 Veterans Administration (VA) facilities to identify patients aged ≥70 years who received a positive FOBT result in 2001. The FOBT date served as the index date, and patients experiencing colorectal cancer, polyps, inflammatory bowel disease, colectomy, or colostomy in the 6 months prior to this date were excluded, as were patients whose claims suggested the FOBT was not administered for screening purposes. The final cohort included 211 men and 1 woman with a mean age of 76.4 years, of whom 84.9% were white.

Combining the Charlson-Deyo Comorbidity Index (CCI) with age, investigators stratified the patients according to life expectancy. The best life expectancy group comprised 51 patients (24.1%) aged 70 to 79 years with a CCI score of 0. The 131 (61.8%) patients in the average life expectancy group were 70 to 84 years with a CCI of 1 to 3 or ≥85 years with a CCI of 0. The worst life expectancy group included 30 (14.2%) patients aged >70 years with a CCI ≥4 and those aged >85 years with a CCI ≥1.

The patients were tracked for 7 years, during which 118 of them underwent follow-up colonoscopy. Results were negative for 59% (n=70) of patients, whereas 5.1% (n=6) had cancer and 29% (n=34) had significant adenoma. The remaining 8 patients received a colonoscopy outside the VA institution, and their results were unavailable, but nothing in the subsequent claims for these patients indicated a colorectal cancer diagnosis.

Of the patients having colonoscopy, 10% (n=12) developed complications, such as bleeding, pain, falls, hospitalization, and death, within 3 days of an undetermined cause. The 15.6% (n=33) of patients found to have a major adenoma and/or colorectal cancer who survived at least 5 years were considered to have received a net benefit. The 22 patients with a small adenoma (<1 cm) who survived ≥5 years and the 8 patients whose colonoscopy findings were absent were categorized as receiving an indeterminate benefit. The other 149 patients—including 45 patients whose FOBT result was found via colonoscopy to be a false positive, 94 patients who had no follow-up colonoscopy, and 10 who died within 5 years of adenoma removal—were considered to have received a net burden.

In the group with the worst life expectancy, 87% of patients experienced net burden versus 70% of patients with an average life expectancy and 65% of patients with the best life expectancy. Patients with the best life expectancy were more likely than those with average and worst life expectancies to receive a net benefit (20% vs 15.3% vs 10%, respectively).

Nearly half the patients (n=43) who did not undergo colonoscopy after a positive FOBT result died within 5 years of causes other than colorectal cancer. The authors said, “[This] suggests that the decision to forego follow-up colonoscopy was appropriate for those patients.” They added that these patients could have avoided FOBT if clinicians had determined they were not suitable candidates for follow-up colonoscopy prior to the FOBT.

“Our study supports guidelines that recommend using life expectancy to guide colorectal cancer screening decisions in older adults and argues against one-size-fits-all interventions that simply aim to increase overall screenings and follow-up rates,” the authors concluded. They cautioned that the study was limited because it was not a randomized trial and used a small sample size. In addition, most patients were white male veterans, which the authors said might limit applicability of the findings to other patient populations.