A six-clinician family practice showed us the quality report it had submitted for PCMH annual reporting. It was 41 pages. It contained twenty-two measures, each with a rate, most with a goal of "improve," and a narrative for each that read like a job description. When we asked the lead physician which measure the practice had actually worked on that year, she thought for a moment and named one: hypertension control. When we asked what the rate was, nobody in the room knew without opening the report.
That is the failure mode for the quality improvement (QI) concept in the NCQA Patient-Centered Medical Home program. Practices treat the QI criteria as a document to produce rather than a system to run. The criteria are actually modest: monitor a small set of measures across a few required categories, set goals on some of them, act on the goals, and report the results to the people who can change them. A one-page dashboard that a practice looks at every month satisfies the standard better than a 41-page report nobody reads, and it is far less work to maintain.
This article describes how we build PCMH quality measure dashboards with the practices we support: how to choose measures that meet the QI 01 and QI 02 categories and also matter to your patients, how to set a goal you can defend, how to report performance to clinicians in a way that changes behavior, and what evidence to keep for annual reporting.
Key takeaways
- The core QI criteria ask a practice to monitor at least five clinical quality measures across four categories (immunizations, other preventive care, chronic or acute care, behavioral health) plus at least two resource stewardship measures, and since the 2024 reporting year NCQA requires standardized electronic clinical quality measure (eCQM) specifications for those measures.
- A dashboard needs six columns per measure: numerator, denominator, rate, goal, trend and owner; anything more is a report, not a dashboard.
- Goals are set from your own baseline and a benchmark, written as a number and a date ("controlling high blood pressure from 61 percent to 70 percent by June 30"), and tied to a specific change the practice will make.
- Reporting to the team means clinician-level rates shared monthly, which is also what the QI criterion on reporting performance asks for.
- The evidence NCQA wants (the measure list, the dated dashboards, the goal statements, the description of what you changed) is produced automatically by running the dashboard; it does not need to be written separately.
What the QI criteria actually require
Under the current PCMH standards, the QI concept begins with measurement. QI 01 requires the practice to monitor at least five clinical quality measures spread across four categories: immunization measures, other preventive care measures, chronic or acute care measures, and behavioral health measures. QI 02 requires at least two resource stewardship measures, meaning measures related to care coordination or health care costs, such as emergency department visits, hospital admissions, generic prescribing or high-cost imaging. Other criteria cover appointment availability and patient experience. The improvement criteria then ask the practice to set goals and act to improve on several of those measures, and the reporting criterion asks it to share performance with clinicians and staff.
Beginning with the 2024 reporting year, NCQA requires practices to use standardized eCQM specifications for the QI 01 and QI 02 measures and their annual reporting counterparts. The practical effect is that "our own version of the diabetes measure" no longer counts. The measure logic (who is in the denominator, what counts in the numerator, which exclusions apply) comes from the published eCQM specification, and the EHR's quality module usually implements those specifications already. That is a gift for a small practice: the hard part of measurement, defining the measure, has been done for you.
Choosing measures that meet the categories and matter
We start with the practice's payer quality programs and its patient population, then map to the eCQM list. A measure you are already paid on, or already being asked about by a Medicare Advantage plan's care gap list, is a measure with data flowing and a reason to improve it. Below is a set we see work for a typical adult and family practice; a pediatric practice swaps in childhood immunization status, well-child visits and adolescent depression screening.
| QI category | eCQM | What it measures | Why it is a good dashboard choice |
|---|---|---|---|
| Immunization | CMS147 Influenza Immunization | Patients 6 months and older who received a flu vaccine during the season | Seasonal, high volume, front desk and nursing can move it |
| Immunization | CMS117 Childhood Immunization Status | Children who turned two and completed the series | Required category coverage for practices with pediatrics |
| Other preventive care | CMS130 Colorectal Cancer Screening | Patients 45 to 75 with appropriate screening | Shared with HEDIS and most payer programs |
| Other preventive care | CMS125 Breast Cancer Screening | Women 50 to 74 with a mammogram in 27 months | Straightforward outreach measure |
| Chronic or acute care | CMS165 Controlling High Blood Pressure | Patients 18 to 85 with hypertension whose last BP was below 140/90 | Outcome measure, large denominator, clinician behavior matters |
| Chronic or acute care | CMS122 Diabetes: Hemoglobin A1c Poor Control | Patients with diabetes whose most recent A1c was above 9 percent (lower is better) | Ties to care management work |
| Behavioral health | CMS2 Screening for Depression and Follow-Up Plan | Patients 12 and older screened with a follow-up plan when positive | Workflow measure the rooming staff can own |
| Resource stewardship | ED visits per 1,000 patients; generic dispensing rate | From payer reports or claims data | Meets QI 02 with data payers already send |
Eight measures on the dashboard is enough. The practice will set goals on three or four of them; the rest are monitored. Resist the urge to add every measure the EHR can compute. A dashboard with twenty-two rows is where the 41-page report came from.
Building the dashboard itself
The dashboard is one page, updated monthly, with one row per measure and six columns: numerator, denominator, rate, goal, trend and owner. The numerator and denominator are there because a rate without them hides small-number noise (a 100 percent flu rate on three patients is not information). The trend is the last twelve monthly rates, ideally as a small run chart. The owner is a named person, not a role, who brings that row to the monthly meeting.
Data comes from the EHR quality module using the eCQM logic, exported on the same day each month, with the report parameters saved so the run is repeatable. Before trusting a rate, validate it once: pull twenty patients from the denominator and twenty from the numerator and confirm the chart agrees. In our experience, the first validation almost always finds a mapping problem, most often a lab result or a vaccine documented as free text that the measure logic cannot see, and fixing the mapping raises the rate before anyone changes a clinical workflow. That is also the moment to reconcile the practice's rate with the payer's care gap list; the two rarely match, and the differences are usually attribution and data lag rather than care.
Clinician-level rows sit below the practice row for each goal measure. This is the part practices hesitate over and the part that produces change. Physicians respond to seeing their own rate next to their partners'; they do not respond to a practice average. Share it in the monthly meeting, unblinded, framed as a shared problem.
A worked example: hypertension control
Take the practice from the opening. Its CMS165 rate in January was 61 percent: 1,140 of 1,870 patients with hypertension had a last recorded blood pressure below 140/90. The national benchmark it chose from its Medicare Advantage plan's quality report was 72 percent. The goal statement: "Increase controlling high blood pressure from 61 percent to 70 percent by December 31, measured monthly with the CMS165 eCQM." Not "improve." A number and a date.
The team then chose changes it could actually make. First, a repeat blood pressure protocol: any initial reading at or above 140/90 is retaken after five minutes by the medical assistant, and the retake is recorded, because the measure uses the last reading of the visit and first readings run high. Second, a monthly outreach list of hypertensive patients with no visit in six months, worked by the care coordinator. Third, a clinician-level report at each monthly meeting. The dashboard row for CMS165 showed 61, 62, 64, 66, 67, 69 across the first six months; the retake protocol alone accounted for most of the early gain. Whether the practice reaches 70 by December depends on the outreach, but every month of that row is evidence of goal setting, action and monitoring, which is what the improvement criteria ask for.
Note what the practice did not do: it did not write a separate QI plan document, a separate PDSA worksheet or a separate evidence narrative. The goal statement, the dated dashboards and a half-page description of the three changes are the evidence, and they exist because the practice ran the dashboard.
Keeping the evidence for annual reporting
PCMH annual reporting asks for the measure list, current performance, and for the improvement criteria, the goal, the actions and the result. Keep a folder with twelve dated dashboard exports, the goal statements with the date they were set, a one-page description of each change and when it started, and the clinician-level report from at least one meeting with a note of who attended. Screenshots of the EHR quality module showing the eCQM name and reporting period prove you used the standardized measure. For resource stewardship, keep the payer report you drew the ED or admission data from. That folder is the annual report; our annual review service is largely a matter of checking it is complete and the numbers reconcile.
For practices pursuing recognition for the first time, the dashboard is also the earliest thing to build, because it takes twelve months to show twelve months of data. Our PCMH transformation work starts there, and the same measures usually feed the care gap outreach the practice is already doing for its payers.
Questions we hear
Our EHR reports differ from the payer's care gap list. Which do we use?
For PCMH, your own eCQM report, because it uses the standardized specification and covers your whole panel, not just one payer's members. Use the payer list as a cross-check and a source of resource stewardship data. When the two disagree on a patient, the chart is the tiebreaker, and the discrepancy usually points to a data mapping problem worth fixing.
How many measures do we need goals on?
The improvement criteria ask for goals and action on several measures across the clinical categories and at least one resource stewardship measure. Three or four active goals is what a small practice can actually work; monitor the rest. Rotating goals annually keeps the dashboard useful.
Do we have to show improvement to keep recognition?
The core criteria require setting goals, acting and monitoring; the criteria that require demonstrated improvement are elective. In practice, a practice that runs the dashboard and works its goals shows improvement on most of them, and where it does not, the documentation of what was tried is still the evidence the core criteria call for.
What to do this week
- Open your EHR quality module and list the eCQMs it can report; mark the ones that fall in each of the four QI 01 categories.
- Pick eight measures using the table above as a pattern, and identify two resource stewardship measures from payer reports.
- Run the reports for last month, validate twenty patients from one measure against the chart, and fix any mapping problem you find.
- Build the one-page dashboard with numerator, denominator, rate, goal, trend and owner columns.
- Write goal statements with a number and a date for three or four measures, each tied to a specific workflow change.
- Put a 30-minute monthly dashboard review on the calendar with clinician-level rates for the goal measures.
