We’re focusing on additional APM Performance Pathway (APP) related questions for ACOs in this episode of Ask Dr. Mingle.

You’ll hear Dr. Dan Mingle explain:

• Beneficiary eligibility for Medicare CQMs
• CEHRT requirements for ACO member practices, and if paper chart practices can participate in MSSP ACOs after the end of the Web Interface
• How quality data is deduplicated for APM Performance Pathway (APP) reporting
• Types of data collected for APP reporting and a look at how to manage your data infrastructure to ensure your data is complete and accurate

Click play on the video below (or scroll down to read the summary) to access this episode now.

Question One: Medicare CQMs for APM Performance Pathway (APP) Reporting

Erika asks: “We have a client asking if Medicare CQMs are to only include Medicare Part B claims, or if they are to include Medicare Advantage claims as well. We’ve heard you say so. But what is your source?”

It’s one of those subjects that creates the foundation of MIPS knowledge and structure, so it is hard to find explicitly in all the provider-facing content that Medicare produces. However, there are credible sources, such as the Electronic Code of Federal Regulation. This is where all federal rules are collated and tracked as they grow and change through the annual rule-making process.

For the specific source for your question, you can find it here: eCFR: Medicare CQMs

It’s the 13th defined term in this table of definitions, titled: “Beneficiary eligible for Medicare CQMs”

• “means a beneficiary identified for purposes of reporting Medicare CQMs for ACOs participating in the Medicare Shared Savings Program (Medicare CQMs), who is either of the following:
  • (1) A Medicare fee-for-service beneficiary (as defined at § 425.20) who –
    • (i) Meets the criteria for a beneficiary to be assigned to an ACO described at § 425.401(a); and
    • (ii) Had at least one claim with a date of service during the measurement period from an ACO professional who is a primary care physician or who has one of the specialty designations included in § 425.402(c), or who is a physician assistant, nurse practitioner, or clinical nurse specialist.
  • (2) A Medicare fee-for-service beneficiary who is assigned to an ACO in accordance with § 425.402(e) because the beneficiary designated an ACO professional participating in an ACO as responsible for coordinating their overall care.”

Question Two: CEHRT Requirements for ACOs

Belinda asks: “I’m trying to determine why an ACO should even entertain keeping a ‘paper chart’ provider on their MSSP roster for 2025 knowing CQMs and eCQMs are the only reporting options. Some things I read say you simply CAN’T report paper charts to CMS after 2024, which makes total sense for eCQMs but CQM seems grey to me because it lists ‘various documentation’ as a collection method. Regardless, I wouldn’t know how to get that into a QRDA for submission. Am I wrong about this? Thinking paper charts CAN’T be in the MSSP in 2025? Or, do you have any suggestions or links for more information? I’d be so appreciative.”

Great question, Belinda!

You’re justifiably confused. You’ve been misled by sources saying you can’t report paper charts after 2024.

But you’ve stumbled onto the correct answer. MIPS CQMs and the new Medicare CQMs collection methods are highly flexible and, for the foreseeable future, are both valid ways of generating complete and accurate submissions to Medicare from a mixed data environment.

And you don’t need QRDA for either MIPS CQMs or Medicare CQMs. A few words about QRDA:

• QRDA3 is the automated output of CEHRT that a practice or provider can submit themselves to Medicare to report eCQMs.
• You can’t add manually abstracted chart data to a QRDA and keep it valid and usable.
• Vendors who report eCQMs on your behalf generally do not use QRDA3. They insert the metrics into a special Medicare API (Application Program Interface) to post the numbers directly into Medicare’s management system.
• No path merges manually abstracted data with electronic data to make an eCQM.
• When manually abstracted data touches an eCQM, it no longer qualifies as an eCQM.
• QRDA3 has no applicability to MIPS CQM or Medicare CQM submissions.
• Qualified Registries who build quality submissions on your behalf also submit directly through the API.

You need a registry with capabilities like ours to bring electronic and manual data sources together into one submission.

MIPS CQMs and Medicare CQMs are flexible:

• You can build them all electronically with end-to-end certified data sources, just like eCQMs.
• Or you can build them with data that comes from uncertified electronic sources
• You can also build them with all manually abstracted data
• Or any mix or combination of the above.

If you’ve locked yourself into eCQMs, you are more right than wrong. eCQMs are inflexible. All data must come from and through CEHRT, end-to-end certified. There’s one exception in the flow. A non-certified entity, typically the practice or a qualified registry vendor, can make the final submission to Medicare. But they must not alter the contents of the QRDA3 which was calculated from and by certified sources.

Even with eCQMs there is a path to a valid submission for an ACO with participants on paper: the tactical use of the data completeness requirement.

For the 2024 and 2025 performance years, the data completeness requirement is at 75%. That means your submission to Medicare needs to include only 75% of the eligible patients or visits.

You can thrive in the APP reporting process if your clinical volume on paper charts does not exceed 25% of your total volume. Don’t expect the data completeness requirement to stay at 75%. Expect it to go up gradually, though I don’t expect it ever to be at 100%.

Of course it is not a trivial matter to measure your data completeness. The eCQM process assumes 100% data completeness and simply doesn’t have access to the data to calculate data completeness. A good registry can and will calculate data completeness with a methodology that will survive Medicare audit.

Both MIPS CQMs and the new Medicare CQMs are fully flexible. The data has to be correct, but it can come from any conceivable source.

Some Qualified Registries may restrict their data sources to electronic only. That might lead to a mistaken view that Registries can’t use manual data.

At Mingle, we have been able to use data from nearly any form and format our clients can generate since the beginning. Like eCQMs, we can accept and use electronic data like FHIR, QRDA1, other HL7 clinical file architectures, SQL flat file data extracts, etc.

But we can also ingest your chart abstraction data delivered in an Excel spreadsheet or other flat file format. Most importantly, we can combine electronically extracted with manually abstracted data, matching and deduplicating patient identities and calculating data completeness rates for valid complex submissions.

You don’t have to jettison your practices that are still on paper charts.

And I don’t see that day coming any time soon.

Some critical elements of your local healthcare system are likely still on paper, and it’s still possible to provide excellent care documented on paper. Also, measuring quality without the global use of CEHRT is possible.

The three measures in the APP can be abstracted, especially in the Medicare CQM style. The depression measure is, by far, the most difficult to abstract, both for the reason of the global patient eligibility for the measure and the broad and somewhat subjective category of addressing depression when screening is positive.

There is also an issue that Medicare has eliminated, starting in the 2025 performance year, the 50 or 75% threshold of CEHRT use in the ACO. That number goes to 100% beginning in 2025. But it still has all the forgiveness built into the Promoting Interoperability performance category. Small practices, non-patient-facing practices, and facility-based practices are still eligible for automatic reweighting of the PI category and are exceptions to the 100% CEHRT requirement for ACOs.

It will progressively get harder to be without CEHRT and comply with quality measurement requirements. I think you should have your own digital strategy that maps a route for all of your practices to be on CEHRT.

The data completeness threshold will go up.

The three measures currently required for the APM Performance Pathway will change. There will undoubtedly be different measures in the future, and possibly more than three will be required.

There may be a local need for other measurement to meet your clinical objectives.

Today, you’d be wise to contractually require data compliance from every practice, whether manually abstracted or electronically extracted. Include consequences, up to expulsion from the group, for failure.

And you’d be wise to have a digital strategy for every practice. Set a path and possibly provide resources to help your practices migrate to CEHRT.

Question Three: ACO APM Performance Pathway (APP) Reporting Process

Jason asks: “How do we deduplicate the data for APM Performance Pathway (APP) reporting? Do we aim for a certain percent accuracy? If so, what is that?”

In any quality measurement and reporting program, the first challenge is patient identification, and its flip side is deduplication.

It would be great to get a reliable Medical Record Number from all of our clients about each of their patients. But you’d be surprised at how difficult that task proves to be. Even in relatively simple, unified local health systems, multiple codes exist that uniquely identify patients. Without a deep, intimate knowledge of the various information systems in use, it’s hard to settle on and deliver a unifying identifier.

In the simple environment of MIPS reporting, where submissions are on behalf of individual providers or single-entity group practices, only about half of our clients can send us a reliable identifier.

In the much more complex ACO environment, it’s nearly impossible, short of enormous investments of time, effort, and money, to implement and maintain an Enterprise Master Patient Index.

Ultimately, you should consider using the same process that measures your quality to give you decision support. And when you are at the point of using the data clinically, you should build a highly reliable system. It’s possible to construct a 6-sigma patient identification system. And we’d be delighted to build that with you.

But our healthcare system has a roughly 2.7 sigma error rate overall. That means there are about 111,000 errors in every 1,000,000 clinical actions. And Medicare’s QPP quality measurement tolerates about the same error level.

We have found that First Name, Last Name, and Date Of Birth (DOB) almost universally give us the precision in patient identification and deduplication that Medicare requires.

If you need higher precision than that, we can customize your system to use other available identifiers. I’d recommend insurance numbers as a next step.

If you install and maintain an Enterprise Master Patient Index, we’d love to build that into your data flow.

But you don’t need that expense and maintenance effort to satisfy Medicare. We’ve found that with simple manipulations of First Name, Last Name, and DOB, we get to a level of precision that defines your performance out to the second decimal place in percentile rankings and passes audit, both those that Medicare requires we perform on your submissions and those that Medicare requires directly from you.

Question Four: ACO APP Reporting Process Continued

Jason asks: “What types of data are collected for each of the three APP clinical measures? For example, what is being extracted for the follow-up part of depression screening?”

There are, of course, three measures currently required for the APM Performance Pathway (APP).

We typically build our submissions from separate analyses of denominator and numerator data.

There are other ways to build these, and we may find, with your ACO, that there is a better way. But it typically works well with this approach.

We determine denominator eligibility from claims data. Claims data always contains name, DOB, diagnosis, and service codes. That data set nicely and accurately defines denominators for each of the three APP measures and pretty much every other measure we may need or want to do with you.

For the diabetes HbA1c measure, we need A1c data, which nearly always exists as a number in your EHR, accessible to simple data query tools.

For the HTN, BP control measure, blood pressure nearly always exists as two numbers, systolic and diastolic, recorded in your EHR database and accessible to simple data query tools.

Data for the depression measure is a bit more of a challenge. We may need to get creative with you to satisfy the depression measure.

If you are setting up your system, I’d set up whatever you use for depression screening as an orderable and resultable test. That means we can recognize its existence and its results electronically.

I would set up the variety of actions you might take on a positive screen as orderable items. The existence of those orders, then, satisfies that second part of the depression measure requiring action on positive screens.

Failing either or both of those configurations requires chart abstraction to characterize your performance correctly. Or a more complex electronic search can satisfy “Action on positive.” We can search medications for an antidepressant, referral orders for a behavioral health referral, next-visit orders for a re-evaluation visit, or anything else that you might choose to do to manage that positive screen.

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.