Colon Cancer Screening among Patients Receiving Dialysis in the United States: Are We Choosing Wisely?

Abstract

The American Society of Nephrology recommends against routine cancer screening among asymptomatic patients receiving maintenance dialysis on the basis of limited survival benefit. To determine the frequency of colorectal cancer screening among patients on dialysis and the extent to which screening tests were targeted toward patients at lower risk of death and higher likelihood of receiving a kidney transplant, we performed a cohort study of 469,574 Medicare beneficiaries ages ≥50 years old who received dialysis between January 1, 2007 and September 30, 2012. We examined colorectal cancer screening tests according to quartiles of risk of mortality and kidney transplant on the basis of multivariable Cox modeling. Over a median follow-up of 1.5 years, 11.6% of patients received a colon cancer screening test (57.9 tests per 1000 person-years). Incidence rates of colonoscopy, flexible sigmoidoscopy, and fecal occult blood test were 27.9, 0.6, and 29.5 per 1000 person-years, respectively. Patients in the lowest quartile of mortality risk were more likely to be screened than those in the highest quartile (hazard ratio, 1.53; 95% confidence interval, 1.49 to 1.57; 65.1 versus 46.4 tests per 1000 person-years, respectively), amounting to a 33% higher rate of testing. Additionally, compared with patients least likely to receive a transplant, patients most likely to receive a transplant were more likely to be screened (hazard ratio, 1.68; 95% confidence interval, 1.64 to 1.73). Colon cancer screening is being targeted toward patients on dialysis at lowest risk of mortality and highest likelihood of transplantation, but absolute rates are high, suggesting overscreening.

Keywords: colon cancer screening; dialysis; hemodialysis.

Figure 1.

Study population and procedure flowchart. (A) shows the cohort assembly. (B) shows the number of patients with screening and nonscreening procedures.

Figure 2.

Kaplan–Meier curves by quartiles of risk of mortality. In (A), the quartile of patients with the highest risk of mortality had the shortest median time to death at 1.6 years compared with those with the lowest risk of mortality at over 5 years. In (B), those patients with the highest risk of mortality had the greatest probability of remaining free from testing compared with those with the lowest risk of mortality. Kaplan–Meier curves by quartiles of risk of transplant can be found in Supplemental Figure 1. Q, quartile.

Figure 3.

Numbers of patients with colon cancer screening tests. (A) Rates by quartiles of mortality risk. (B) Rates by quartiles of likelihood of transplant.