Overuse of screening colonoscopy in the Medicare population

Abstract

Background: All relevant authorities recommend an interval of 10 years between normal screening colonoscopies. We assessed the timing of repeated colonoscopies after a negative screening colonoscopy finding in a population-based sample of Medicare patients.

Methods: A 5% national sample of Medicare enrollees from 2000 through 2008 was used to identify average-risk patients undergoing screening colonoscopy between 2001 and 2003. Colonoscopy was classified as a negative screening examination finding if no indication other than screening were in the claims and if no biopsy, fulguration, or polypectomy was performed. Time to repeated colonoscopy was calculated.

Results: Among 24,071 Medicare patients who had a negative screening colonoscopy finding in 2001 through 2003, 46.2% underwent a repeated examination in fewer than 7 years. In 42.5% of these patients (23.5% of the overall sample), there was no clear indication for the early repeated examination. In patients aged 75 to 79 years or 80 years or older at the time of the initial negative screening colonoscopy result, 45.6% and 32.9%, respectively, received a repeated examination within 7 years. In multivariable analyses, male sex, more comorbidities, and colonoscopy by a high-volume colonoscopist or in an office setting were associated with higher rates of early repeated colonoscopy without clear indication, while those 80 years or older had a reduced risk. There were also marked geographic variations, from less than 5% in some health referral regions to greater than 50% in others.

Conclusions: A large proportion of Medicare patients who undergo screening colonoscopy do so more frequently than recommended. Current Medicare regulations intending to limit reimbursement for screening colonoscopy to every 10 years would not appear to be effective.

Figure 1

Schema for selection of the study cohorts. The numbers for the various indications in the right hand boxes add up to more than the total because some patients had multiple indications. GI indicates gastrointestinal tract; DCBE/CT, double-contrast barium enema/computed tomography.

Figure 2

Cumulative percentage of repeated colonoscopies for patients 66 years or older who underwent a complete a colonoscopy between 2001 and 2003. The blue line is for all patients with a complete colonoscopy between 2001 and 2003 (N=236 145). The green line is for patients who had a negative colonoscopy result (n=114 468). The black line is for patients who had a negative colonoscopy finding with no indication other than screening (n=24 071).

Figure 3

Percentage of Medicare fee-for-service enrollees who underwent early repeated colonoscopy with no clear indication, by health referral region. All patients underwent a negative baseline screening colonoscopy between 2001 and 2003.