In this week’s episode, Dr. Dan Mingle answers listener questions about the proposed increase to the MIPS Performance Threshold and the introduction of Medicare CQMs in the 2024 Proposed Rule. He also shares some additional context about data completeness rules in response to three recent questions, and explains how an Enterprise Master Patient Index (EMPI) can be helpful for APP reporting, but isn’t completely necessary.

Click play to listen to this episode right now:

Question One: Medicare CQMs

Arline asks: “Medicare has proposed a Medicare CQM to make compliance with the APP easier over the next few years. Should we use Medicare CQMs for our ACO if it is finalized?”

I think that Medicare CQMs will be a valid and useful tool. However, it will not offer a significant advantage over MIPS CQMs.

To review: Medicare proposes Medicare CQMs as an alternate quality data collection type for Medicare Shared Savings Plan (MSSP) ACOs to meet the APM Performance Pathway (APP) requirements.

The proposal is that Medicare CQMs will use the same measures with the same specifications as MIPS CQMs, but restrict the denominator patients to Medicare Part B beneficiaries, setting aside the all-patient, all-payer inclusion of MIPS CQMs.

To support this and provide a tool to help ACOs aggregate and deduplicate patients, Medicare proposes to provide a list of performance year Medicare Part B patients of the ACO to the ACO early in January.

Medicare’s patient list won’t be of much help to you. It can’t be the sole source of your patient list because:

1. It will only include patients through the 3rd quarter of the performance year.
2. If you’re working with us and have claims data from each of your practices, we have enough to deduplicate your patients. And by filtering on Medicare Part B claims, we can easily toggle between Medicare CQMs and MIPS CQMs submissions.
3. As we envision using Medicare’s patient list to assist the process, we don’t see a substantial additional effort without a distinct advantage.

But still, there are advantages to Medicare CQMs:

• Suppose you must manually abstract some charts to meet your data completeness requirements or uncover electronically inaccessible performance data. In that case, the Medicare CQM route gives you a far smaller set of charts to abstract than the corresponding MIPS CQM.
• It’s also credible to think that there are enough differences in the Medicare Part B patient and care dynamics that there might be a better performance score in your Medicare Part B limited patient population than in your general all-patient, all-payer population. However, the reverse is also possible.

Question Two: MIPS Performance Threshold

Carrin asks: “We have a group of close to 40 MDs, and our composite MIPS score was 80.2 in 2022. It was enough to earn a small positive adjustment. But we are really concerned about the proposal to increase the performance threshold to 82 for 2024. Do you have any advice for us?”

Reaching 82 points is going to be hard. I think many organizations will shift their strategy from earning a positive adjustment to minimizing the negative adjustment.

My first piece of advice is that you have to earn all of the “gimme” points.

There is no reason you can’t earn all of the available 15 points for Improvement Activities. Make sure you have this covered early in the year.

It may take a little more investment, but earning all 25 points for Promoting Interoperability is also a priority.

• The biggest challenge is likely to be the public health registry measure. In that measure, you depend on your vendor being willing to connect and a local public entity that has invested in the connectivity infrastructure. You’ll also depend on fitting the project into your budget.
• And the other interoperability measures, physician-to-physician each way and physician-to-patient, are not trivial.

With Improvement Activities and Promoting Interoperability in the bag, that’s 40 points of the required 82. You still need 42 of the possible 60 points in Cost and Quality to avoid a negative adjustment (30 points are available for each category).

If you have that perfect score in Improvement Activities and Promoting Interoperability, you must earn 70% of the available points in Cost and Quality to avoid a negative adjustment.

Earning the full 30 points in Quality is achievable but challenging. CMS publishes the score you need for each measure in advance, based on national scores from two years before the current performance year.

It will take a coordinated effort of your whole practice to achieve optimum results for each measure.

And finally, the Cost performance category may be your biggest challenge.

Of the four performance categories, Cost is the only category truly scored on percentile rank with the comparison made to your peers in the same performance year.

Cost is the only category to use the entire range of possible scores from the first to the hundredth percentile (or from 0 to 30 points). Of the 30 available points in Cost, half of practices will be below 15 points and half will be above. If you are at the average of 15 points, you will need 27 of 30 Quality points to reach the performance threshold.

Cost is hard to follow in the short term. And it’s hard to change. It’s mired in local habits, traditions, and expectations of patients, providers, and payers. Your best guidance to your Cost performance is previous scores.

And although there will be some random variation from year to year, without concerted effort, Cost won’t vary much.

It makes great sense to engage a reputable registry like Mingle Health. Get involved early. Submit Quality data regularly, at least quarterly, to monitor your quality scoring and take action early enough to improve your scores with enough time ahead of you to improve your final annual MIPS composite score.

A good registry like Mingle Health will help you build a competitive Quality score, and we can provide guidance, advice, and submissions, where indicated, for Promoting Interoperability, Improvement Activities, and Cost. We also monitor the rules for you to be sure you are aware, in a timely fashion, of changes significant to you.

Question Three: Data Completeness & Practices Joining Organization

Kenadi asks: “We have some practices joining our organization mid-year. We are trying to understand what this means to us for MIPS reporting for the year they join. The practices are small compared to our whole group and will represent less than 70% of our total patient volume. So even if we don’t get all of the data for the year they join us, we should be confident that they are not pulling us below the data completeness requirement. But we are confused by the apparent contradiction in Medicare’s statement that we must account for 100% of our eligible patients. Can you explain this?”

Most of you know that Medicare’s data completeness requirement is 70% for 2023, goes up to 75% for 2024 and 2025, and is proposed to go up to 80% for the 2026 performance year.

But, seeming to contradict the “leniency” in that requirement is Medicare’s assertion that you must account for 100% of your eligible patients.

I’ll try to help you make some sense of that.

This apparent contradiction is resolved with firm division, in your mind, of numerator versus denominator.

The denominator is the count of patients, visits, or episodes of care that are eligible for a measure. Since the denominator can be a count of different things, Medicare calls the things counted “eligible instances.”

A denominator is a count of eligible instances, such as patients, visits, events, etc.

These instances represent opportunities to provide care. It is easier to count the instances, or opportunities, than it is to identify, with unambiguous data, the care provided.

Determining the denominator – or the eligible instances – is the first step in any quality reporting effort.

Medicare requires you to know what this total number is. Only by having an accurate denominator count can you know that you have fallen below the data completeness threshold.

Denominator data primarily comes from claims records. The great thing about claims data is it is nearly universally available. It’s been around long enough that it is a mature and stable data set with near-universally accepted standards.

Every practice generates insurance claims. Claims are relatively simple and standardized. Electronic claims were all designed to emulate Medicare’s paper HCFA 1500 form. They all include universally recognized data standards of ICD10 diagnosis codes and CPT service codes.

A healthcare claim nicely defines the existence of a patient-doctor relationship. In the claims are:

• This doctor
• On this date
• Saw this patient
• Of gender
• And DOB
• Providing these services
• For these diagnoses

That data set is the foundation of the denominator for every measure.

Getting that data is not trivial but is generally achievable in every practice. There are at least three different ways to get the data, and part of our task at Mingle Health is to help you find the best way, given your unique strengths and weaknesses, to get that data from each of your practices.

It should also help you to know that claims data from your practice management system is adequately standardized so that there are rarely any barriers to combining that data from multiple practices, multiple practice management systems in one practice in the same year, or any other mixing or matching you need to do.

Now that you know that the denominator is a count of eligible instances, the gold standard for the count comes from claims data, and Medicare wants you to be 100% accurate on that count; let’s talk about numerator data and data completeness.

Numerator data is primarily clinical data.

It answers the specific question: “What happened to the patients, or in the visits or episodes in our denominator? Did they get what we wanted them to get? Did they avoid complications?”

Numerator data is less mature, standardized, and available than claims data. The three big problems with clinical data are:

1. Too much of our clinical data is in natural language, not in structured and discrete elements.
2. Numerator specifications are not always optimized for a structured data set.
3. And we don’t have solid standards for expressing commonly required quality measurement concepts.

The project and process changes when you change EHRs. Even if you’re switching to a new instance or version of the same EHR.

So, Medicare gives you a little more leeway with the numerator.

Regarding the requirement to report 100% of your patients, Medicare refers to the denominator, which counts the patients, visits, or episodes of care that a standard of care applies to. With the 100% requirement, they say you must know who your patients are, even if you don’t know what you did for them.

Regarding the 70%, 75%, or 80% data completeness requirement, Medicare refers to the numerator, the care provided, found in clinical data in the EHR.

Question Four: Penalties for Errors in Eligible Instance Counting

Greg asks: “Are we going to be penalized if we make an error in counting the eligible instances?”

Don’t sweat it too much.

You need to make a credible effort, but there is always error in any measurement.

Medicare knows that as well as anybody else and accounts for it. You don’t need to worry if your error rate is in the low single digits. Medicare seems to tolerate about a 3% error rate before they take corrective action.

Question Five: Claims Data & Denominator Criteria

Jean asks: “You have said that data completeness criteria apply to numerators and that Medicare wants us to report 100% of our denominators. And that the ‘gold standard’ for denominator data is claims data. But many measures have clinical data that define the denominator. Can you explain this further?”

I had hoped to hear that question.

In the early days of quality reporting, measure authors limited denominator specifications to claims-accessible data. That proved limiting and swelled the denominators for many measures to massive levels. Now we see a mix.

At Mingle, we take a stepwise approach to calculating denominators. Whenever we import claims data, we run the data set against the claims-accessible denominator specifications, producing “Denominator Candidates.”

We only get a final denominator count once we add clinical data.

But when you are calculating reporting rates and answering the question, “Am I meeting data completeness requirements?” we find that calculating from the denominator candidates yields a credible calculation.

If you examine the logic:

• True eligible instances can never be a higher number than the denominator candidates.
• All true eligible instances are contained within the denominator candidates.
• And the factors that make them different numbers are reasonably expected to be similar from practice to practice.

Despite not being the final denominator count, counts of denominator candidates reflect the actual distribution of eligible instances across your many practices and are the source of a credible calculation of data completeness.

Question Six: Enterprise Master Patient Index (EMPI) for the APM Performance Pathway (APP)

Georgina asks: “Medicare requires that the quality reports for our Medicare Shared Savings Plan ACO be deduplicated for patients amongst our many practices. Don’t we have to have an Enterprise Master Patient Index (“EMPI”) to accomplish the requirements of the APM Performance Pathway?”

The EMPI is a great tool and will give you more confidence in your patient identification. The level of accuracy and confidence the EMPI provides is a tremendous advantage to guide clinical care in a diverse community health care system.

But the EMPI is more discernment than you need for value-based care quality reporting.

Medicare measures and compares your quality with your peers, calculating the performance rate to four decimal places. It’s two decimal places if you are considering the percentage calculation.

We find that by combining simple, always-present data from multiple claims data sets, we achieve confidence in identification out to the limit on which you get reported and scored.

If you have an EMPI, so much the better. Providing us with your EMPI identifier makes deduplication easier.

Send us your value-based care questions!

If you’d like to ask a question about the APP transition, MIPS, ACO quality reporting, or any other Alternative Payment Model, you can reach out to us in three ways:

You can leave your questions in a YouTube comment under any episode of Ask Dr. Mingle.

On LinkedIn, leave your questions in a comment on any of our posts.

And you can reach out directly by sending an email to hello@minglehealth.com.