The dQM roadmap: building a foundation for 2030
Subscribe to our newsletter
SubscribeThe Centers for Medicare & Medicaid Services (CMS) has set a target of moving all its quality programs to digital quality measures and digital data collection by 2030.
The National Committee for Quality Assurance (NCQA) is heading the same way, shifting its Healthcare Effectiveness Data and Information Set (HEDIS®) program from traditional to digital quality measurement. NCQA calls 2030 its “north star” for going fully digital.
So how do you get there? Earlier this year, we covered what’s changing in the shift from eCQMs to dQMs and why the economics work in your favor, not against you. The question I hear most now is harder to answer: how do we start?
The 18 to 24 month reality
Building a foundation for digital quality measures takes time. Some of it is the regulatory calendar, worth a direct look first. Some of it is work you control, which is what we’ll explore more below.
The table below covers NCQA’s HEDIS path, since that’s where the dates are most concrete. CMS has set the same 2030 target for its own programs, like hospital and clinician reporting, but most of its milestones are still at the comment and request-for-information stage. CMS asked for feedback on draft FHIR measures in early 2026.
NCQA describes its own shift in four phases. The years below come from NCQA’s 2023 roadmap and are approximate, since its current pages list the phases without dates:
| Milestone | What’s happening |
|---|---|
| 2023 | Digital Introduction. NCQA launches Digital Content Services; early adopters get reference CQL engine access. |
| 2024 to 2026 | Digital Enabled. Administrative portions of measures go digital through Digital Content Services; plans can use any supported CQL engine. |
| Any time you start | Best practice: comparative testing (formerly called parallel testing). As of March 2026 it’s optional, and NCQA leaves the timing and scope to you. Digital results still go through the normal HEDIS audit. |
| Through MY2029 | Fully Digital. Hybrid measures retire one at a time. In January 2026, NCQA refined some measure-specific pathways and interim dates, though the MY2029 endpoint hasn’t moved. |
| Around 2030 | Digital Only. NCQA plans to retire paper specifications and traditional measure certification for vendor logic. HEDIS audits continue. |
Traditional HEDIS measurement is largely claims-based and backward-looking by design. Every team that codes measures from written specifications has to interpret them, which is where the variation we’ve written about before comes from. Digital measures remove that step, because NCQA ships the logic itself.
Most organizations underestimate the data layer, not the engine. Mapping your clinical data to FHIR, aligning terminology, and validating that mapping against real patient records is where most of those 18 to 24 months are spent, in our experience. Measure execution itself, once the data is ready, moves a lot faster.
That validation does not stop once the mapping is done. Every FHIR Measure spells out the data requirements its logic needs to run (which resources, profiles, and value sets), and the engine can only work with what it is given. Feed it incomplete or incorrectly mapped data and you get an incomplete or incorrect result, no matter how good the engine is.
The friction is rarely the standard itself. In practice, it’s resource allocation (your HEDIS team is already running this year’s cycle), provider documentation (different providers record the same event differently, and that inconsistency carries straight into your FHIR data), and internal coordination (the team that owns the data usually isn’t the team that owns the measure, or the budget).
The 90-day pilot
Most organizations don’t think in terms of a pilot. It looks like an all-or-nothing choice: stay on the current system, or commit to the full transition. A 90-day pilot is the option in between: test the model on one or two measures, against your own data and alongside your existing reporting process, before you commit to anything bigger.
It doesn’t have to wait for a complete data foundation either. That’s a common sequencing mistake. The two can move in parallel.
- Pick a measure with real stakes and a manageable scope. Colorectal cancer screening is a strong, non-hypothetical starting point: ECDS-only since MY2024, already used in NCQA’s own Health Plan Ratings, listed by NCQA as a suggested measure for comparative testing, and complex enough to be a genuine test.
- Prove the execution trace works against your actual data. See exactly which logic branch produced which result, not just a final number you have to trust.
- Produce something you can validate against what you already report. If the pilot doesn’t match what you currently report, you want to know why during the test, not during a submission cycle.
- Build the internal case before you build the whole thing. A working pilot does more for a budget conversation than any roadmap slide.
The engine work itself is the easy part. With Firely dQM, loading a new measure release takes an afternoon, which leaves the rest of the 90 days for validation rather than waiting on the engine.
If you want to see what that execution trace is built from, I walked through the mechanics in What’s inside a FHIR Measure.
Your roadmap
NCQA’s phases above, and CMS’s own timeline, describe where the industry has to be. This is how your own organization gets there:
- Stage 1: Readiness check. FHIR foundation and data layer work: initial FHIR mapping, slow but everything else depends on it.
- Stage 2: Comparative testing. One or two measures running alongside your existing reporting, proving the model on real data before you commit further.
- Stage 3: Scale and report. The measure portfolio expands and legacy processes begin retiring, measure by measure rather than all at once.
- Stage 4: Operationalize. Custom measures, on demand execution, and a genuinely operational dQM environment instead of an annual reporting exercise.
These are your own stages, separate from NCQA’s regulatory phases above, and not every organization starts at the same one. The most common surprise is in Stage 1: a number of payers have told us their quality team assumed it had no FHIR data at all, when its interoperability team already had some flowing for CMS’s Patient Access API, with more on the way for the CMS-0057-F APIs due January 1, 2027.
The two teams simply weren’t talking, and a FHIR-enabled endpoint doesn’t mean data arrives in FHIR format automatically, so there’s still mapping work ahead. But knowing that data already exists changes who you need to talk to first.
If you’re starting this quarter, I’d recommend three things:
- Confirm who runs your CQL engine, in writing. Your submission vendor may run it, your analytics vendor only for testing, or you may need to run it yourself. That answer shapes most of what follows.
- Benchmark your FHIR data against NCQA’s HEDIS Core Implementation Guide before committing to a pilot measure. In our experience, most of the gap between your current results and your digital results traces back to data elements, not measure logic. Every FHIR Measure spells out its data requirements, so follow them closely: how a given piece of clinical information should be expressed in FHIR is not always obvious, even when the underlying data already exists somewhere in your systems.
- Start comparative testing early. NCQA no longer requires it, but still recommends it as best practice. It’s also the best way to catch discrepancies before they reach a submission cycle. NCQA also suggests starting with a small set of measures and scaling up, so the sooner your pilot starts, the better.
The argument for starting now
While there is still time before 2030, it’s worth factoring in how long things take to go into production.
Organizations starting in 2028 will have no room to learn, no room to fix the data quality problems every organization eventually finds. A team that starts now can run at least one full measurement cycle before hybrid reporting ends in MY2029. A team that waits until 2028 runs its first cycle at the same time everyone else does.
Starting now buys something beyond deadline compliance: a buffer that late starters won’t have time to build, and the earlier that work happens, the more time it saves down the line. Every measure you validate this year is one you aren’t validating under pressure in 2029.
What’s next
If you want to go deeper, we’ll soon be sharing a practical, step-by-step playbook covering everything you need to know about digital quality measures, and how to get ready for what’s next.
If you are somewhere on this path already and want to talk through where you actually stand, get in touch.
