2019 MIPS Quality Historic Benchmarks

See attachments for 2019 Quality Benchmark Excel

Quality Measure Benchmarks Overview

What Are Quality Measure Benchmarks?
When a clinician or group submits measures for the Merit-based Incentive Payment System (MIPS) quality performance category, each measure is assessed against its benchmark to determine how many points the measure earns. In program year 2019, a clinician or group can receive anywhere from 3 to 10 points for each MIPS measure (not including any bonus points) that meets the data completeness standards and case minimum requirements by comparing measure performance to established Benchmarks. Benchmarks are specific to the collection type: Qualified Clinical Data Registry (QCDR) measures, MIPS Clinical Quality Measures (MIPS CQMs), eCQMs, CMS Web Interface measures, the Consumer Assessment of Healthcare Providers and Systems (CAHPS) for MIPS survey, and Part B Claims measures. In order to measure performance that is comparable across the spectrum of performance, benchmarks are established using historical data. Historical benchmarks are based on actual performance data from 2017 that was submitted to the Quality Payment Program (QPP) in 2017, except for the CAHPS surveys. For 2019 CAHPS for MIPS, we have not yet established benchmarks. The CMS Web Interface uses benchmarks from the Medicare Shared Savings Program.

How Are Benchmarks Displayed?
Each benchmark is presented in terms of deciles. Points will be awarded within each decile (see Table 1). Clinicians who receive a score in the first or second decile will receive 3 points. Clinicians who are in the 3rd decile will receive somewhere between 3 and 3.9 points depending on their exact position in the decile, and clinicians in higher deciles will receive a corresponding number of points. For example, if a clinician submits performance data of 83% on a non-inverse measure, and the 5th decile begins at 72% and the 6th decile begins at 85%, then the clinician will receive between 5 and 5.9 points because 83% is in the 5th decile. For inverse measures where a higher performance is seen by a lower number on the performance score, the scores are reversed in the benchmark deciles.

Historical Benchmark Inclusion Criteria
Benchmarks are established using historical data, if available, which is the performance data submitted (during applicable submission period) by individuals and groups two years prior the start of the applicable performance year for which historical benchmarks are established. Data reported in the historical Benchmarks reflect submissions by the following:

• Individual clinicians who:
  • Are identified as a clinician type under the definition of a MIPS eligible clinician;
  • Are not newly Medicare-enrolled; and
  • Exceed the low-volume threshold established for the 2019 performance year (have more than $90,000 in Medicare Part B allowed charges and provides care to more than 200 Part B-enrolled Medicare beneficiaries and have more than 200 Part B Medicare services) as an individual.
• • Groups that:
  • Have at least one clinician who is identified as a clinician type under the definition of a MIPS eligible clinician; and
  • Exceed the low-volume threshold established for the 2019 performance period (have more than $90,000 in Medicare Part B allowed charges and provides care to more than 200 Part B-enrolled Medicare beneficiaries and have more than 200 Part B Medicare services) as a collective group.

Comparable Alternative Payment Model (APM) data was included when possible. Benchmarks were created if there were at least 20 reporting clinicians or groups that met the criteria for contributing to the benchmark, including meeting the minimum case size (which is generally 20 patients), meeting the data completeness criteria (60% reporting rate), and having performance greater than 0 percent (less than 100 percent for inverse measures).

What If A Quality Measure Does Not Have A Historical Benchmark?

For measures with no historical benchmark, MIPS will attempt to calculate benchmarks based on 2019 performance period data. The same MIPS eligibility criteria listed above will be applied prior to establishing performance period benchmarks for measures. If no historical benchmark exists and no benchmark can be calculated, then the measure will receive 3 points as long as data completeness and case minimums have been met. In the list of measure benchmarks, measures without historical benchmarks have no data provided in the decile fields.

Benchmark Descriptions

Each benchmark has the following information:

• Measure name and ID

• Collection type (eCQMs, QCDR measures, MIPS CQMs, Medicare Part B claims, CMS Web Interface, the CAHPS for MIPS survey, and administrative claims measures)

• Measure type (e.g., outcome, process,)

• Whether or not a benchmark could be calculated for that measure/submission type

• Range of performance rates for each decile to help identify how many points the clinician earns for that measure

• Whether the benchmark is topped out (topped out means the measure is not showing much variability and may have different scoring in future years)

• Whether the measure will receive special scoring in a topped out status (specially scored measures will receive 7 points)

• Whether the measure was topped out in program Year 2018

• Whether the measure was topped out in program Year 2019

Table 1: Using Data Benchmarks to Determine Points (Non-Inverse Measures)*

*For inverse measures, the order would be reversed. Where Decile 1 starts with the highest value and Decile 10 has the lowest value.

Historical Benchmarks with Less Than Ten Deciles

By using historical measure performance data, some benchmarks across one or more submission types were identified with maximum rates (i.e. 100%) without utilizing all ten deciles. These benchmarks are identifiable when the deciles from three to nine are not populated while the tenth decile is identified at 100%. This is evident in inverse measures as well. Deciles that are not populated indicate that the historical benchmark analysis identified that between 10% to 60% or more of the clinicians performed at the maximum achievable performance rate. For example, in the benchmark for Measure #117 presented below, historical benchmarking identified that the top 40% of clinicians performed at the maximum rate. Therefore, clinicians using this submission type that performed above the 6th decile would receive a maximum performance score of 10 points. T
Special Considerations

Historical Benchmarks for CMS Web Interface Reporters

For the CMS Web Interface, the benchmarks are the same as the 2019 Medicare Shared Savings Program performance benchmarks. While the benchmarks are the same, the scoring will be adjusted to be consistent with other MIPS measures. In order to align with the Medicare Shared Savings Program not including benchmarks below the 30th percentile (which is the start of the 4th decile), any value below the 30th percentile will receive 3 points. However, if performance is above the 30th percentile, then scoring will be the same as other measures. For additional guidance on the 2019 MSSP Benchmarks, click here. https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/sharedsavingsprogram/Downloads/2018-and-2019-quality-benchmarks-guidance.pdf

Benchmarks for Consumer Assessment of Healthcare Providers & Systems (CAHPS) Reporters

For CAHPS for MIPS, 2019 benchmarks will be available for each summary survey measure (SSM). However, the CAHPS for MIPS benchmarks have not yet been established because a revised survey was implemented in 2018. Therefore, we will calculate benchmarks for 2019 based on 2019 performance data. A range of 3 to 10 points are assigned to each SSM by comparing performance to the benchmark (similar to other measures). The final CAHPS for MIPS score will be the average number of points across all scored SSMs. The CAHPS for MIPS benchmark file will be provided when it is available at the end of the performance period.

Historical Benchmarks for the All-Cause Hospital Readmission Measure

The percentile- and decile- level benchmarks for the all-cause hospital readmission (ACR) measure will be created using the 2017 MIPS data. The 2019 ACR MIPS benchmarks will be provided in early 2019.

Historical Benchmarks for Topped-Out Measures

For each process measure, a measure is topped out if the median performance rate is 95% or higher (non-inverse measure) or is 5% or lower (inverse measures). For each non-process measure, a measure is topped out if the truncated coefficient of variation (TCV) is less than 0.10 and the 75th and 95th percentiles are within 2 standard errors. The status of topped-out is identified in the benchmark file that accompanies this fact sheet. For further understanding on how topped out measures are scored, please review the MIPS Scoring Guide.

Benchmarks for Multi-Strata Measures

Some measures have more than one numerator and denominator, or stratum, used to calculate overall performance. These multi-strata measures usually employ the average or weighted average of each numerator and denominator combination (i.e. the stratums are combined). However, in some measures there are specified stratum identified as the primary stratum for a performance rate to use in calculating benchmarks. The list of multi-strata measures and performance processes are provided as a tab in the measure benchmark file package and will be updated and posted in 2019.