APP Plus Measure Deep Dive: Diabetes / Hemoglobin A1c Poor Control (CMS 122v13)
Diabetes remains one of the costliest and most complex chronic conditions to manage. In a MIPS world, Diabetes: Hemoglobin A1c Poor Control (CMS 122v13) is definitely a measure you should be reporting. But just reporting it is not enough. I want to show you where you can gain extra points, avoid common pitfalls, and make it a lever for clinical and operational improvement. In this post, I’ll strike a balance between what clinicians need to know and what practice managers must execute. You’ll get an insider’s view of how to turn this measure into a performance advantage.
What CMS 122v13 Measures, and Why It Changed
The evolution from v12 to v13 (Courtesy of eCQI Resource Center)
In prior versions, the measure was strictly about HbA1c > 9.0 % (i.e., “poor control”).
In v13, CMS broadens that to a glycemic status assessment. That means either HbA1c or GMI (Glucose Management Indicator, often derived from CGM data) can count.
The rule is that if you have multiple results on the same date, you take the lowest (i.e., the “best” for the patient) number.
This shift enables practices that adopt CGM / remote monitoring tools to rely on that data. However, this also introduces complexity; not every vendor supports GMI, and not every registry knows how to consume it. You must check your EHR or registry setup before relying on GMI.
Core logic: denominator, numerator, exclusions
Denominator (who counts)
Patients aged 18 to 75 by the end of the measurement period.
A diagnosis of diabetes.
At least one qualifying visit (office/outpatient/telehealth) in that period.
Exclusions/denominator exceptions
You can exclude patients under certain conditions:
Hospice care at any time in the measurement period
Patients 66 or older who are long-term institutionalized (nursing home)
Patients 66 or older with frailty + advanced illness or on dementia medications
Patients receiving palliative care during the period
The key to your practice is mapping who in your panel is truly “exception-eligible” to avoid undercounting or overcounting exclusions.
Numerator (“bad outcomes”)
A patient is in the numerator (i.e., negative performance) if:
Their most recent glycemic status (HbA1c or GMI) is > 9.0 %
OR the glycemic status assessment is missing/was not done in that period
If their result is ≤ 9.0 %, they are not in the numerator, which is good. Lowering that numerator percentage is your goal.
Because missing or undocumented data counts against you, your workflows must ensure tests are ordered, results captured, and documentation is solid.
Important nuance: reverse measure
CMS treats this as a reverse measure (i.e., you want the numerator as low as possible). That means many strategies focus on preventing patients from entering the numerator, rather than just reacting after they do.
Where Practices Trip Up: and How to Get Ahead
Here’s where many practices lose easy points. I’ll pair each pitfall with a tip.
Pitfall 1: relying solely on HbA1c, ignoring GMI
If your practice uses CGM devices but your registry or EHR does not accept GMI, you lose an edge. GMI can rescue borderline cases in patients who otherwise might have high HbA1c values.
Insider tip:
Ask your EHR vendor or registry (or IT team) to map GMI data fields into the eCQM feed. Even if your practice is small, a one-time configuration to accept both HbA1c and GMI can yield measure gains year after year.
Pitfall 2: letting “no lab” = data missing
Many practices only count those tested; they think that if a patient was lost to follow-up, that patient falls out of consideration. Not true. Missing labs often default to the numerator.
Tip:
Set up an “A1c missing” flag in your care management dashboard. Each quarter, generate a list of patients who did not get a glycemic test. Reach out (via phone, portal, or outreach) to get the lab done. Even a home A1c kit can help document the result.
Pitfall 3: poor visit or ordering workflows
Clinicians sometimes forget to order A1c because it’s not included in the visit template. Some labs aren’t interfaced, so results don’t flow in.
Tip:
Embed an “order A1c / GMI” reminder in every diabetes visit template. Use soft alerts in the EHR (but don’t burnout clinicians with alerts). Prepopulate lab order sets.
Also, ensure lab interfaces are working so results auto-flow into discrete fields, not as attachments or scanned PDFs.
Pitfall 4: exclusion mapping mistakes
If you fail to tag legitimate hospice, frail, or palliative care exclusions, you might count patients who should be excluded. That drags your denominator and performance.
Tip:
At the start of the performance year, run a query to flag patients over 66 who might be “frail/advanced illness” and tag them in your registry. Keep a manual review process. If you receive a new diagnosis or hospice referral midyear, check immediately to see if exclusion should apply.
Pitfall 5: clustering poor control patients for passive “hope”
Some practices let poorly controlled patients linger without active intervention because the measure is retrospective. That is a missed opportunity.
Tip:
Split your poorly controlled cohort into tiers (e.g.,> 11%, 9.1–11%, uncontrolled but improving). Build short-term outreach cycles (phone, telehealth, med titration) every 3 months. Use care coordinators or pharmacists to own that list.
Pitfall 6: underutilizing patient self-monitoring data
Many patients now use CGM, home glucose logs, or digital apps. However, practices don’t integrate this into their workflows because they fear “non-standardized data.”
Tip:
Decide on a “trusted data protocol,” such as accepting 14 days of CGM or home logs with physician review, and then convert those into a documented GMI or average. Capture that in discrete fields. Over time, that can reduce the “missing data” penalty.
Step-by-Step Workflow You Can Adopt
Here’s a blueprint your practice can follow. Use it to draft your local version based on your EHR/registry.
Baseline audit
Export your diabetes panel, filter the results for individuals aged 18–75, and then separate those with no recorded HbA1c or GMI in the past year.
Identify patients already with >9.0%.
Identify patients near threshold (e.g., 8.5–9.0) for early intervention.
Template & order setup
Modify your diabetes visit template to auto-include the A1c / GMI order unless a valid exclusion applies.
Build lab panels that include A1c by default.
Ensure the lab interface maps discrete result fields (not just attachments or scanned PDFs).
Exclusion tracking
Develop a flagging system (manual or algorithmic) to identify hospice, frailty, and palliative care patients.
Review yearly and midyear for new eligible exclusions.
Missing data outreach
Each quarter, run a “no test” list.
Assign either a care coordinator or nurse to outreach (call, portal, mail).
Consider sending home A1c kits or arranging local lab draws.
Record results back into discrete fields.
Stratified intervention for poor control
Divide “>9.0%” patients into categories (e.g., 9.1–10.5, 10.6–12, >12).
Assign care managers, pharmacists, or clinicians to own each group.
Trigger frequent touchpoints: monthly check-ins, medication review, behavioral support, and telehealth.
Use CGM / patient data.
If the patient uses CGM or digital logs, integrate that into the chart.
Calculate or import GMI where possible.
Use that data to adjust therapy between visits.
Ongoing monitoring and reporting
Each month, check your numerator and denominator trends.
Watch for “leakage” (patients added or lost).
Before submission, export the measure file and test it locally where possible (e.g., in a sandbox) to spot missing patients or data gaps.
Unique, Harder-to-Find Tips
These are edge strategies that many websites won’t talk about:
“Smart exclusion override review”: Periodically review patients who were excluded for frailty or advanced illness. Some may have “reversed” status (improved) and should return to measure inclusion. That can boost your denominator and dilute the numerator.
“Shadow GMI import”: Even if your registry rejects GMI, import CGM / home log data into a private field for clinician review. Use it clinically to adjust therapy, even if it doesn’t hit the measure. Over time, you’ll build internal confidence in that data stream.
“Lab prompt via patient portal”: Send patient portal reminders with a message like, “It’s time for your A1c check — labs open at your convenience.” Some patients will come in without prompting from staff.
“Pharmacy lab sync”: Collaborate with local pharmacies offering walk-in lab draws to share lab data. Some pharmacies already do basic lab panels. Get them to send you the A1c result.
“Quarterly “nudge” campaigns: In months with historically low visit volume (e.g., November), run mini campaigns encouraging diabetic patients to come for lab checks. These can catch last-minute lab completions before the year’s end.
“Measure discrepancy audits”: After you generate your final export, randomly pick 10 patients in the numerator and check their charts manually. You may find minor documentation errors or missing tag usage that, when corrected, improve your score.
How This Helps Clinicians and Practice Managers
Clinicians
You know which patients need active, frequent attention.
You can intervene earlier, not just after A1c runs skyward.
You gain confidence in CGM/home data by setting protocols.
Practice Managers / Administrators
You gain control of your quality score, not just “hoping clinicians remember.”
You can distribute tasks (data capture, outreach, exclusion review) to non-physician roles.
You can track trends and catch “data gaps” before submission.
When both sides work together, the measure becomes a tool, not a burden.
Here’s the thing: CMS 122v13 (Diabetes: Glycemic Status Assessment > 9%) offers more flexibility compared to its older versions. But that flexibility comes with more moving parts. Don’t let that trip you. Use the steps, tips, and insider strategies above to turn it into a source of advantage.
Resources for this post:
