Skip to content
1-866-359-4458 Log In
Get Started

Quality Reporting Data Sources: Denominators, Numerators, and FHIR | Ask Dr. Mingle

We’re continuing our conversation about the best data sources for calculating healthcare quality measures in this week’s episode. Click play below to hear Dr. Dan Mingle explain data sources for denominators, numerators, and how the FHIR standard fits into quality reporting.

Question One: Quality Reporting Data Sources – Denominators

Kyle asks: “What about practice management system data makes it great for denominators?”

There are several reasons, including:

  • Availability
  • Standardization
  • Perspective

Practice management systems were the first health information systems to go electronic. They’ve had a lot of time to mature and standardize. Medicare long ago drove standardization of the health claims system around the HCFA 1500 claim form. It’s a reliable and pervasive standard.

And the perspective is nearly ideal. There can be a little more to it than this, but the best documentation that a patient-doctor relationship exists is its tracks through the billing system and a record of encounters. There is no more reliable record of the patient-doctor relationship.

Included with each claim is the foundational data for all denominators:

  • Identity of the practice and the provider
  • Identity of the patient, including gender, date of birth, address
  • Identity of the insurer
  • Diagnoses addressed that visit
  • Services provided that visit

It’s a solid foundation.

Claims give us a first-pass identity and count of denominator-eligible patients. We call these denominator candidates.

It sometimes requires additional clinical data to complete the denominator eligibility determination and the final denominator count.

For instance, we can know from claims that a patient is hypertensive. But we need clinical data to know they are on a diuretic and should have a periodic potassium determination.

With claims data, we build the foundation of your quality reporting system. We identify:

  • Your patients.
  • Their diagnoses.
  • The services you provide.
  • And a preliminary count of your eligible instances.
    • Clinical data can decrease this count, but your actual eligible instances will never exceed this number.

Question Two: Quality Reporting Data Sources – Numerators

Kyle asks: “You’ve told us a lot about claims data and the value of claims in determining quality measure denominators. What about numerators?”

Numerator data is, typically, clinical data.

A numerator is typically answering a question about the care of those patients or visits populating the denominator:

  • Did the provider prescribe an indicated medication?
  • Was an indicated test ordered?
  • Did treatment lead to control of the blood pressure, HbA1c, depression, etc.?

Clinical numerator data can come from a variety of sources. There is a lot more variation in clinical data than in claims data. The variation relates to the variety of possible sources and differences in the documentation structure within each source.

Numerators can come from claims data. We must realize that claims data has two principal sources: practice-sourced and insurer-sourced. There are different strengths and weaknesses of each.

Claims data for numerators:

  • You can find things like flu shots, pneumonia vaccines, mammography, and screening colonoscopies in claims data.
  • This type of clinical data is reliable. It’s rare that these kinds of interventions are provided but not billed or billed but not provided.
    • An insurer can have a complete record of claims for flu shots in its population of beneficiaries, but a practice will only have claims evidence for the flu shots provided in the practice. Flu shots are available and supplied from a wide variety of locations.
    • Mammography is more extreme. The provider ordering the mammogram is rarely the source of the mammogram and its billing.
  • For interventions with longer appropriate intervals, claims become less and less helpful.
    • Changes in providers and changes in insurers make the data less accessible and less reliable.
      • If I measure something that happened in your practice, it’s relatively easy to get all your claims.
      • It’s also relatively easy to get insurer sourced claims from one insurer to learn what has happened outside the practice.
        • But suppose I am trying to measure specific things that happened to any of your patients through any insurer. In that case, it’s challenging to get timely and reliable claims data from all of the 900 health insurers in the United States, most of whom stratify their operations regionally.

Most often, numerators come from clinical data:

  • It can be the legal medical record, either a paper chart or an electronic record.
  • Or it can be a derivation of the medical record like a tumor registry, a diabetes registry, or any other managed list of patients, interventions, and/or outcomes.
  • An understanding of the strengths and weaknesses of the source and its potential errors guides how that data is treated, particularly what we conclude when data is absent.
  • For paper charts, quality measurement usually involves manual chart abstraction.
    • Guided by electronic assessment of denominators, we can reduce the effort otherwise required to abstract the charts.
  • Manual abstraction may also be the best way to get specific data out of electronic charts, mainly when the needed documentation is unstructured and text-based.

Question Three: Automating Quality Measurement with FHIR

Kyle asks: “If our clients want to automate their quality measurement, do they need to use the Fast Healthcare Interoperability Resources (FHIR) standard?”

We should all be advocating for the FHIR standard. But it has less immediate impact on your choices than you might think.

FHIR is a means to an end. You should invest your time, money, and energy into getting to that end. That end is high-value, highly reliable, highly effective healthcare. FHIR is one step in one credible path to that universally desirable end.

You should advocate for FHIR politically. You should pressure your vendors to provide it by asking them how much they support it.

But you should not include it today in your buying decision. You need to measure the quality of your care today. It can make a big difference in your payments from insurers, particularly Medicare.

Quality scores are increasingly visible to your patients and affect their provider choices. You need a scalable and sustainable measurement system to help you enter our high-value healthcare future.

Buying FHIR today is like being the first to own a fax machine. Who can fax to you? Who can you send faxes to?

There’s no problem really with the FHIR standard. It’s a good standard. It is usable. It is better than most (maybe all) alternatives. More vendors are capable of it than there are using it. We’ll find holes and weaknesses as it gets into broader use. Once found, we’ll fix those problems and eventually evolve the standard into better and better versions and, perhaps, a new name.

It will not be FHIR as we now know it that you will integrate into your systems. It may not be called Fast Healthcare Interoperability Resources. It will evolve as HL7 Clinical Data Architecture was the basis of, and evolved into, FHIR.

At Mingle Health, we can read and incorporate FHIR-sourced data into quality assessments.

But there are still three big problems with the FHIR standard:

  1. It’s not in universal use.
  2. When used, there is no universal population of precise data.
  3. And the biggest problem is that documentation relies too heavily on free text without enough structured documentation to populate a FHIR message. Until data sources are more machine-usable, FHIR can’t take hold.

If your data is structured, our job of quality reporting becomes more manageable, and we can provide more and better insight for you to improve and prosper.

It makes a more significant difference to you and us to migrate off text toward structured data than to migrate to any specific standard.

If you can provide it, we can read and translate between FHIR, CDA, CCD, QRDA, RxNorm, SnoMed, LOINC, CPT, ICD, etc.

There is a more meaningful difference in function and utility moving from unstructured textual entries to any structure than there is migrating between structures.

The biggest barrier between the current state and adopting the FHIR standard is the need for more structured data documenting the patient’s typical annual healthcare experience.

Let me go even a step further. FHIR is an intermediate outcome. It is not a desirable end in itself. FHIR is one of many potential intermediate goals to get to a fully digital system. And even that fully digital system is intermediate to the overarching goal of the high-value, highly reliable, highly effective healthcare system.

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

Want to learn more about the APM Performance Pathway (APP)?

For many ACOs, the transition to all-patient, all-payer quality reporting is a significant challenge and often feels like an unnecessary burden imposed by CMS.
We've created our latest PDF guide to help ACOs orient themselves to the challenge of all-patient, all-payer reporting while finding opportunities to thrive in the future.

Access the Guide
Get Helpful News & Resources
  • This field is for validation purposes and should be left unchanged.