1. Variables influencing ADR. CMS explained that the measure does not account for variables that may influence the ADR such as geographic location, socioeconomic status of patient population, community compliance of screening, etc. The agency further states that according to the risk factors outlined by the American Cancer Society, African Americans have the highest colorectal cancer incidence and mortality rates of all racial groups in the United States. “In addition, dietary factors, such as consumption of highly processed meats will contribute to an increased risk of colorectal cancer. This diet is more prevalent in lower socioeconomic areas, which could influence the outcome of the measure. There are other patient factors like education, health literacy, etc. that might also affect things like the adequacy of bowel preparation, which in turn could affect performance.”
The societies advised CMS that this rationale reflects a misunderstanding of the definition of ADR, which includes all average-risk patients in whom the physician finds at least one adenoma. Further, ADR only includes colonoscopies with adequate bowel preparation and complete examinations. Studies demonstrate that ADR is not influenced by socioeconomic status and sex mix of the provider’s patient population, or by the rate of screening in the community.
Socioeconomics, ethnicity, and diet are not relevant factors of ADR. That said, our societies welcome the opportunity to work with CMS on creating age and sex standardized ADRs for the U.S. population, if feasible, in order to capture information that CMS deems important.
2. Failure to detect all adenomas. CMS stated that the measure does not account for MIPS eligible clinicians that fail to detect adenomas but may score higher based on the patient population.
The societies pushed back with CMS explaining that this rationale again reflects a misunderstanding of the definition of ADR, which includes average-risk patients for whom the physician finds at least one adenoma. Colonoscopy is heavily operator dependent. In an average-risk, mixed population, the variability in ADR reflects quality of the provider’s endoscopic skills and pathology recognition, rather than the risk of the underlying population.
3. Incidence measure. CMS concluded that Measure 343: Screening Colonoscopy Adenoma Detection Rate is considered an “incidence measure” that does not assess the quality of the care provided. In essence, according to CMS, the measure is based on happenstance rather than the eligible clinician providing a thorough examination.
The societies strongly disagreed with this characterization of ADR. Measure 343: Screening Colonoscopy Adenoma Detection Rate is the only measure that assess the quality of the exam performed by the physician in an average-risk patient with an adequate bowel preparation. Physicians can improve their adenoma detection rate by paying attention to detail, spending more time looking for adenomas, and learning better techniques.
4. Benchmarking. CMS stated that because of the measure construct, benchmarks calculated from this measure are misrepresented and do not align with the MIPS scoring methodology where 100% indicates better clinical care or control. Guidelines and supplemental literature support a performance target for adenoma detection rate of 25% for a mixed sex population (20% in women and 30% in men). CMS determined that Measure 343: Screening Colonoscopy Adenoma Detection Rate may be appropriate for other programs but does not align with the scoring logic within MIPS. When this measure was introduced, according to the agency, it was under the legacy program, Physician Quality Reporting System (PQRS), a pay-for-reporting program that does not have the same scoring implications as MIPS.
The societies reminded CMS that the 25% is the minimum requirement for performance and is not a benchmark. This minimum requirement continues to increase as well. With 25% being the threshold, for every 1% increase in ADR the risk of fatal interval colon cancer decreases by 3%. In one important study by Corley et al, the lowest quintile of ADR was 19% or lower, and was associated with the highest risk of interval colon cancer.4
CMS must begin to move beyond traditional approaches toward benchmarking performance where 100% compliance is expected. It was encouraging to see CMS acknowledge that nuances to evaluating scores are needed based on the ability of a measure to accurately identify and capture performance based on the patient population and measure specifications. For example, these adjustments were finalized for the blood pressure and diabetes HbA1c measures, where the highest number of points will be achieved for anyone scoring 90% or higher. This modification was based on the knowledge that it is not realistic nor in the interest of patients to assume that clinicians will be able to achieve the desired targeted outcome for every patient. The potential for unintended consequences was factored into an assessment of what performance could be considered achievable.
In our view, ADR is a similar example where 100% performance across a clinician’s population of patients is biologically impossible since not every individual who receives a screening colonoscopy will have an adenoma detected. ADR is the best-established colorectal neoplasia-related quality indicator and research demonstrates that high rates are associated with significant reductions in colorectal cancer risk.
CMS must continue to explore alternative strategies toward benchmarking in MIPS to ensure that achievement is fairly assessed, and top performance scores are determined not solely based on peer performance but also based on clinical evidence balanced with minimizing unintended consequences. The MIPS program and its benchmarking and scoring methodologies must continue to innovate to ensure that physicians provide the best possible care to their patients while also accurately and fairly representing and rewarding clinicians’ performance. Continuing to promote a siloed view toward quality will only reduce the relevance of the MIPS program and lead our members to question the integrity and validity of the program.
