Health Equity Index (HEI) in Star Ratings

What CMS decided

The Health Equity Index reward is not being implemented. CMS said it in the Contract Year 2027 Medicare Advantage and Part D final rule, issued April 2, 2026, in one sentence worth quoting in full:

For the 2027 Star Ratings, CMS is not implementing the Excellent Health Outcomes for All reward (previously called the Health Equity Index reward) that was developed to incentivize improved performance for a subset of enrollees and will continue the historical reward factor that encourages consistently high performance for all enrollees across all quality measures.

The rule (CMS-4208-F3 / CMS-4212-F) was published at 91 FR 17384 on April 6, 2026 and took effect June 1, 2026. The regulation confirms it. 42 CFR 422.166(f) now carries the reward factor at (f)(1) and the Categorical Adjustment Index at (f)(2), and there is no (f)(3). Reward factor values are unchanged at 0.4, 0.3, 0.2, 0.1 and 0.0.

If a 2027 Star Ratings forecast still has an equity-based reward in it, take it out. This is the change most often misremembered as an addition, because it was announced as one three years earlier.

What the HEI was designed to be

The 2024 final rule (CMS-4201-F) finalized a health equity index reward beginning with the 2027 Star Ratings. CMS later renamed it Excellent Health Outcomes for All. It was not an add-on. It was a replacement: the HEI would have taken the place of the existing reward factor.

The design worked measure by measure. For each measure, CMS would compare contracts on how well they performed among enrollees with specified social risk factors. Contracts in the top third scored +1, the middle third scored 0, and the bottom third scored -1. Those scores were weighted by measure weight and combined into an index, and the index determined the reward added to a contract's summary and overall ratings.

Size of the affected population decided how much a contract could earn. Contracts whose share of enrollees with those social risk factors was at or above the contract-level median across all contracts would receive the full reward, contracts at or above half the median would receive half, and contracts below half the median would receive nothing, with a positive index score required in every tier. The first HEI scores were to be built on pooled data from measurement years 2024 and 2025, which is why plans spent both of those years building stratified reporting they now own with no reward attached to it.

The population it was built around

CMS specified three social risk factors: dual eligibility for Medicare and Medicaid, receipt of the Part D low-income subsidy, and disability status. These are the same characteristics the Categorical Adjustment Index already uses, which was one of the recurring criticisms of the HEI: a plan could be adjusted for the same population twice, once through the CAI and once through the index.

That criticism is now moot on one side, because only the CAI remains. The CAI sits at 42 CFR 422.166(f)(2), is applied after the reward factor, and adjusts summary and overall ratings for the within-contract disparity associated with dual or low-income subsidy status and disability. Plans serving a high-dual population are still adjusted. They are just not rewarded separately for outcomes in that population.

What the reward factor does instead

The reward factor pays for two things at once: performing well and performing evenly. It is rating-specific and gets added to both the summary and overall ratings.

CMS ranks each contract on the weighted mean of its measure scores and on the weighted variance of those scores, relative to all rated contracts at the same rating level. Under 42 CFR 422.166(f)(1)(ii), relative variance performance is categorized as high (at or above the 70th percentile), medium (between the 30th and 69th percentiles) or low (below the 30th percentile). Relative mean performance is categorized as high (at or above the 85th percentile), relatively high (between the 65th and 84th percentiles) or other (below the 65th percentile). The combination sets the reward.

The strategic read is that consistency is worth as much as peak performance. A contract with one 5-star measure and one 2-star measure is punished by variance in a way an equity index would not have punished it. Two-tenths of a star is often the distance between a 3.5 and a 4.0, and the bonus payment that follows.

What plans should still do

The reward went away. The reasons to look at performance by social risk factor did not.

• The CAI still uses this population. Dual, low-income subsidy and disability status still adjust your final rating. Knowing where your subgroup gaps are is how you predict that adjustment instead of receiving it.
• Variance is now the thing being paid for. A measure where a dual subgroup lags 15 points behind the rest of the panel is a measure sitting low in your distribution. Closing that gap raises the mean and lowers the variance at the same time, which is exactly what the reward factor pays for.
• The measures are scored on everyone. Subgroup underperformance was never a separate scoreboard. It was always inside the rate.
• The measure set got smaller. CMS removed 11 measures in the same rule and is adding a Part C Depression Screening and Follow-Up measure starting with the 2027 measurement year and the 2029 Star Ratings. Fewer measures means each surviving one carries more weight, so a single lagging subgroup does more damage than it used to.
• Keep the stratified reporting you built. The 2024 and 2025 stratification work still tells you which populations move your cut point position. Ripping it out to save a reporting line is a false economy.

Common mistakes

• Modeling 2027 with an equity reward still in it. The most expensive version of this error is a bonus payment forecast built on a reward that will not arrive.
• Assuming the reversal makes 2027 easier. It removes a bonus opportunity for high-dual contracts. It does not lower a single cut point, and Tukey outlier deletion is still pushing them up.
• Confusing the HEI with the CAI. One was reversed before it started. The other has been in the methodology for years and is still applied.
• Dismantling social-risk-factor reporting. The reward is gone; the CAI, the variance math and the underlying rates all still depend on that view of your population.
• Reading “equity is out” as a clinical signal. The rule changed how CMS pays. It did not change which members are hardest to reach, and those members still show up in the denominator of every measure.

How Pelica handles this

The reward factor rewards low variance, which means the practical work is finding the subgroups dragging individual measures and closing those gaps before the measurement year ends.

Pelica's Quality & Stars copilot keeps a single canonical record per member across claims, EHR, pharmacy, lab and ADT, so a gap can be sliced by dual status, low-income subsidy or any other attribute without a BI ticket, and then worked by an agent rather than added to a report. Customers have seen a 41% improvement in quality gap closure, roughly 90% BCS and KED closure in-year, and 96%+ adherence on the triple-weighted Part D measures.

Related terms

Cut points are the thresholds a reward factor gets added on top of, and Tukey outlier deletion is why those thresholds keep rising. CAHPS is the patient-experience survey behind the measures that carry the most weight in the patients' experience category, and it is the slowest one to move. For the full 2027 picture, read the 2027 Star Ratings changes quality leaders should plan for.

Sources

• CMS: Contract Year 2027 Medicare Advantage and Part D Final Rule fact sheet (CMS-4208-F3 / CMS-4212-F)
• Federal Register: Contract Year 2027 policy and technical changes, 91 FR 17384 (April 6, 2026)
• eCFR: 42 CFR 422.166, Star Ratings adjustments including the reward factor and CAI
• CMS: 2024 Medicare Advantage and Part D final rule (CMS-4201-F), which finalized the HEI reward