One in five screening colonoscopies performed in Medicare patients aged 70 was considered potentially inappropriate, but study design limitations might have introduced error.
Previous studies suggest that a significant number of physicians systematically perform screening colonoscopies at 5-year versus recommended 10-year intervals and that many colonoscopists recommend that surveillance colonoscopy after resection of polyps begin earlier than is accepted in guidelines.
Now, investigators have retrospectively assessed the receipt of inappropriate screening colonoscopy in Medicare beneficiaries who had received a recent colonoscopy (index colonoscopy), including a 100% sample in Texas and a 5% sample in the U.S. Screening colonoscopy was distinguished from diagnostic colonoscopy by the absence of a relevant diagnostic indication on the index colonoscopy claim or on any claim 3 months before the procedure. Early repeated colonoscopy was defined as an index colonoscopy with no diagnostic indications preceded by a colonoscopy within 10 years that had negative findings (based on the last colonoscopy if >1). An inappropriate colonoscopy was defined as an early repeated colonoscopy in patients aged 70 to 75, a routine screening colonoscopy in patients aged 76 to 85, or any screening colonoscopy in patients aged >85.
In Texas beneficiaries, 23.4% of colonoscopies were potentially inappropriate, with variation by age group (70–75, 10%; 76–85, 39%; >85, 25%). Procedure-level factors associated with increased risk for inappropriate colonoscopy were location of ambulatory surgery center or office setting (vs. a hospital) and performance by higher-volume colonoscopists (vs. lower-volume), generalists or surgeons (vs. gastroenterologists), or U.S.- trained physicians (vs. non–U.S.-trained physicians). Patient-level risk factors were male gender, white race, fewer comorbid conditions, lower educational level, and residence in an urban area. Six percent of the variance in whether a colonoscopy was potentially inappropriate was explained by the physician variable.
Comment: This study has several limitations. First, not all doctors accept USPSTF age recommendations for screening colonoscopy; many find them to be arbitrary and prefer use of the previous recommendation to stop screening when life expectancy is <10 years. Second, guidelines for postpolypectomy surveillance include consideration of findings not only from the last colonoscopy but from previous colonoscopies. For example, in patients with high-risk adenoma findings, surveillance colonoscopy is recommended at 5-year intervals even after a negative examination. These repeat examinations would have been considered inappropriate in this study. Finally, it is unclear whether the investigators were fully able to account for postpolypectomy surveillance colonoscopies, to which even the USPSTF did not apply their age recommendations.
This study will undoubtedly be cited as evidence of inappropriate colonoscopy. However, the use of claims data to identify indications for colonoscopy is fraught with problems. The absolute rates of inappropriate colonoscopy could be significantly lower than shown here, and not all of the observed trends may reflect real differences in inappropriate use.
Source: Journal Watch Gastroenterology