A strong answer proposes a composite, not a single raw number, and is explicit about why.

The standard single metric — HEDIS CDC HbA1c < 8 (or < 9) rate: percentage of diabetic members with their most recent HbA1c < 8 during the measurement year. Pros: defined, comparable across plans, reportable for STAR. Cons: a one-dimensional snapshot — misses medication adherence, kidney screening, eye exams, blood pressure control.

A better composite (name it something like “Diabetes Care Bundle”):

HbA1c < 8 (clinical control).
Annual nephropathy screening (kidney).
Dilated eye exam in the past 2 years.
Statin adherence (PDC ≥ 80%) for members with diabetes and ASCVD risk.
Blood pressure < 140/90 most recent.

Report as “% of diabetic members meeting all five.” Tough denominator, but the gaps are actionable — you can see which bundle elements are missing.

Justification: diabetes outcomes are driven by sustained multi-component care, not HbA1c alone. A plan optimizing only HbA1c can ship a member whose A1C is controlled via aggressive medication but who’s headed for CKD because no one screened in three years.

What could go wrong (the Goodhart section):

Gaming. If comp or provider bonuses are tied to the bundle, providers may do low-value check-the-box visits to close gaps without engaging the patient. You see rising numerator without rising real outcomes.
Denominator games. Re-defining “diabetic” narrowly excludes hard-to-manage patients and lifts the score without improving care.
Selection bias. Members who don’t engage with care aren’t captured in denominators consistently across plans — cross-plan comparisons get noisy.
Equity blind spots. A flat bundle score obscures outcome gaps by race, income, or language. Report the bundle stratified by demographic subgroup alongside the headline, or you risk publishing an average that masks a crisis.
Measurement lag. HbA1c is a 3-month average; a recent intervention won’t show for a quarter.

What to tell the CEO: “Here’s the bundle score — that’s the headline. Here it is stratified by race and payer. Here are the bundle elements driving the gap. Here’s the trend, and here’s what we’d close first.”

Design a population-health metric for diabetes management.

Draft your answer

Stuck? Peek at a hint

More healthcare questions