Colorectal cancer screening is cost-effective and “should be considered well beyond age 75 years” for individuals not previously screened, according to a computer simulation study published in the Annals of Internal Medicine.
The researchers noted that while the U.S. Preventive Services Task Force recommends against routine screening for colorectal cancer in adequately screened persons older than 75 years, the recommendation does not address the appropriateness of screening in elderly persons without previous screening, and this “has led many members of the medical community to believe that no one older than 75 years should be screened for [colorectal cancer]. However, because unscreened elderly persons are at greater risk for [colorectal cancer] than adequately screened elderly persons, screening them is likely to be effective and cost-effective up to a more advanced age.”
Once individuals reach their late 80s or early 90s, however, “screening is not likely to be cost-effective, even in those without previous screening,” the researchers added, because of risks of death from competing disease and from screening-induced harms, such as colonoscopy-related complications and overdiagnosis and overtreatment.
Researchers used the Microsimulation Analysis Colon (MISCAN-Colon) model to quantify the effectiveness and costs of screening and determine at what ages colorectal cancer screening should be considered in unscreened elderly persons. The model was developed at Erasmus University Medical Center, Rotterdam, the Netherlands, and the study was conducted by researchers there and at Memorial Sloan Kettering Cancer Center, New York. The study was funded by the National Cancer Institute grant as part of the Cancer Intervention and Surveillance Modelling Network.
“For each age between 76 and 90 years, we simulated a cohort of 10 million elderly persons without previous screening with no, moderate, and severe comorbid conditions (a total of 45 cohorts),” the authors explained. “Compared with cohorts of adequately screened elderly persons, the risk for [colorectal cancer] in these cohorts was substantially greater.”
For example, among simulated patients aged 80 years who had negative screening colonoscopies at ages 50, 60, and 70 years, colorectal cancer was prevalent in 0.3% and adenomas in 14.1%, but in simulated patients aged 80 years without previous screening, colorectal cancer was prevalent in 2.6% and adenomas in 44.9%. Within each cohort, researchers simulated a one-time screening with colonoscopy, sigmoidoscopy, and fecal immunochemical test and assessed cost-effectiveness for each method.
The researchers concluded that “[colorectal cancer] screening should be considered well beyond age 75 years” for elderly persons in the United States without previous screening—a group, the researchers noted, that represents 23% of all U.S. persons older than 75 years. “In unscreened elderly persons with no comorbid conditions, [colorectal cancer] screening should be considered up to age 86 years (up to age 83 years for those with moderate comorbid conditions and up to age 80 years for those with severe comorbid conditions).” The cost-effectiveness of screening with sigmoidoscopy and [fecal immunochemical testing] depended on age and comorbid conditions, but “colonoscopy is indicated at most ages.”
The model used in the study “takes into account a wide range of relevant variables to estimate the quality-adjusted life-years gained or lost with screening,” according to an accompanying editorial. “Thus, the model is sensitive both to the benefits of screening when life expectancy is long and deaths from cancer are prevented and to the harms of screening when life expectancy is short and screening leads to detection and treatment that do not prolong life.”
The study “has important implications,” the editorialists added. “Colonoscopy should be considered in every American aged 75 years (and in many who are older) who has not had colorectal cancer screening, and those who do not have a fatal illness will likely benefit from screening.” ■
van Hees F, et al: Ann Intern Med 160:750-759, 2014.
Clark AV, Landefeld CS: Ann Intern Med 160:804-805, 2014.