The final rule is here, and with it comes great news for providers in 2017. CMS heard your concerns, and has made changes to make it easier for providers to get started in the transition year of 2017. If you had a chance to read our earlier blog on the MIPS quality category, “MIPS Quality Performance Category Measures: What you need to know,” this blog will update some of the information that changed in the final rule.
So what’s changed?
The quality category is now worth 60% of a provider’s final score. For the 2017 transition year, CMS has eliminated the requirement that providers submit a cross-cutting measure. They have also decreased the “data completeness” criteria from 90% down to 50% of patients for all payers.
With the “Pick Your Pace” options, providers can submit as little as one quality measure for one patient and receive enough points to avoid the penalty in 2019. In the transition year of 2017, measures will be awarded a minimum of three points to a maximum of ten points.
The final rule also eliminates two of the three CMS calculated measures, leaving only the All Cause Readmission measure, which CMS will only include for practices with more than 15 providers and more than 200 attributable cases. If submitting as individuals, individual providers will not be scored on this measure.
Scoring: The Basics
The final rule implemented the decile scoring system initially proposed, with a slight modification. The final rule increased the minimum number of points awarded each measure submitted from one to three.
The final rule also introduced the concept of Class 1 and Class 2 measures. Class 1 measures will be awarded three to ten points depending on where a provider’s performance falls within the decile ranking of all submitted performance scores. All measures failing to meet the criteria for Class 1 status are deemed Class 2 measures, including those failing to meet the data completeness criteria. Class 2 measures will be awarded three points regardless of performance.
Based on the source of the data and choice of measures, you can also receive bonus points as well. This blog will focus on deciles and the assignment of points to the provider-reported measures.
In the next Quality Scoring blog, we will review the CMS calculated ‘All Cause Readmission’ measure, bonus points and calculating the final score.
What’s a decile?
Historically, where sufficient data was available, CMS has published benchmarks for each measure. The benchmark included a mean (average) score and a standard deviation. A provider’s quality of care would be deemed low, average or high, based on whether a provider’s performance, for a given measure, fell above, below, or within one standard deviation. Measures were normalized by dividing the observed deviation from the benchmark average of the submitted performance by the calculated standard deviation.
So what is a decile? “Deciles” rank the submitted performance levels by the number of individual providers submitting that performance level or range. In the example, below, the top 10%, or Decile, of providers submitted performance levels between 85% and 100%. Ten percent of the performance values fall within each of the ranked decile ranges. For MIPS, CMS will predetermine the decile scores for performance rates based on data submitted in the year that is two years prior to the performance period. For the MIPS performance year 2017, CMS will take all the scores submitted for a given measure in 2015 and determine the decile rankings. CMS will determine decile ranking at a level of precision of tenths of deciles. That means there will be 100 distinct decile performance levels for each measure, decile 0.1 to decile 10.0.
During the transition year of 2017, each Class 1 measure will be awarded a score from three to ten points, based on where the performance score falls within the deciles for that measure.
Under MIPs, if your performance falls within the top decile range (top 10% of performance scores) you will be awarded 10 points for that measure and so on down. The exception is that providers with performance in the lowest two deciles will still receive three points.
Here is an example of how the points might be distributed for Measure 110, Flu Vaccine.
|Example of Measure 110: Flu Vaccine|
|Benchmark Decile 1||0 – 6.9%||3|
|Benchmark Decile 2||7.0 – 15.9%||3|
|Benchmark Decile 3||16.0 – 22.9%||3.0 – 3.9|
|Benchmark Decile 4||23.0 – 35.9%||4.0 – 4.9|
|Benchmark Decile 5||36.0 – 40.9%||5.0 – 5.9|
|Benchmark Decile 6||41.0 – 61.9%||6.0 – 6.9|
|Benchmark Decile 7||62.0 – 68.9%||7.0 – 7.9|
|Benchmark Decile 8||69.0 – 78.9%||8.0 – 8.9|
|Benchmark Decile 9||79.0 – 84.9%||9.0 – 9.9|
|Benchmark Decile 10||85.0 – 100%||10|
Using the example for Measure 110, if you submitted with a performance score of 72.7%, your score falls within Decile 8, which has a range of 69.0 to 78.9% and is eligible for 8.0 to 8.9 points. With the tenths of deciles broken out, it would likely fall on a score around 8.4. Since the decile system is a ranking method rather than a mathematical comparison, you can’t know what the specific ranking will be until the numbers have been collated in rank order.
Considerations when choosing Measures:
No benchmark data available
In order to be a Class 1 measure, and eligible for more than three points, a measure must have a benchmark. To calculate decile benchmarks, a measure must have been submitted a minimum of 20 times during the benchmarking period.
New measures or those with a limited number of submissions during the benchmarking period, benchmarks can’t accurately be determined prior to the close of the performance year and therefore will be Class 2 measures. If a minimum of 20 cases are submitted during the performance period, it is possible that CMS will be able to determine a benchmark based on performance year submissions. However, if your goal is to maximize your points and earn an incentive, using Class 2 measures is risky, because you won’t know ahead of time how those measures will be scored. This is something to keep in mind when choosing measures.
“Topped Out” measures
Several measures that have been used in PQRS, have very high benchmarks. So high, that scoring one standard deviation above the benchmark, means that the score is greater than 100%. This allows little room for improvement. These are described as “topped out” measures.
In the decile ranking system, the topped out phenomenon is manifest as multiple deciles clustered on a high performance rate. An example might be a measure where 50% of submitters achieve a 100% performance rate. For such a measure, there is no differentiation between the 5th and the 10th decile ranks.
Medicare considered dropping measures as they topped out. But dropping topped out measures will make it harder for many specialties that already have trouble finding applicable measures. And Medicare recognizes a potential bias that not everyone is participating yet in Medicare quality reporting. If the early adopters are higher performers, later adopters might increase the spread in the decile rankings.
The current plan is to keep topped out measures and to score them at the mid-point of the clustered deciles. That means that for a measure in which 50% of submitters achieve a 100% performance rate, all submitters at 100% percent performance will be scored midway between the decile 5 and the decile 10 – a score of 7.5. Furthermore, a topped out measure will not get this treatment until the year after it was identified as topped out. That means that no measure will get the mid-point of the cluster score in 2017. No measure will be scored using the topped out methodology until the 2018 performance year.
Medicare is not explicit in the final rules on their plans in 2017 for topped out measures. It would appear that the intention is, for our example of a measure where 50% of submitters achieve a 100% performance rate, all would be scored at a decile ranking of 10.
In the next Quality Scoring blog, we will pull it all together: scores for provider-submitted measures and population measures, bonus points, etc. I hope this review of “decile scoring” helped to demystify some of the new scoring methodology we will see 2017!
This eBook will help you prepare for the transition to MIPS and APMs.