Medicare Health Outcomes Survey (HOS)

What the Medicare Health Outcomes Survey is

HOS is a survey CMS sponsors and Health Services Advisory Group administers, fielded to Medicare Advantage enrollees. It was the first patient-reported outcomes measure used in Medicare managed care, and that description is still the useful one: HOS asks members about their own health and function rather than about the plan.

The standard baseline sample is 1,200 enrollees per Medicare Advantage organization. Since 2019 an organization can request oversampling when contract enrollment allows more than the standard 1,200. Respondents are drawn at random, so a plan does not choose who is surveyed and cannot work a list.

How the two-year cohort works

A new baseline cohort is surveyed every year. Two years later the same respondents are surveyed again, and the pair of answers becomes the measurement. Cohort 26 is the current example: the baseline was fielded in 2023 and the follow up in 2025, and those two rounds together produce the HOS outcome measures used in the 2027 Star Ratings.

A HOS outcome measure is closed before you see it. By the time the 2027 rating publishes, the care that moved it happened between 2023 and 2025. There is no fourth-quarter push on HOS.

Which Star measures come from HOS

Six measures come off the HOS questionnaire for the 2027 Star Ratings. Five are Star Ratings measures and one is display only. CMS split them into two families in its July 31, 2026 HOS memo, and the split matters because the two families read different cohorts.

Three of the six are HEDIS measures collected through the HOS instrument, which is why they carry the HEDIS HOS label. For the weight that applies to those three in a given rating year, check the measure table in the CY2027 Advance Notice for your measurement year; do not assume it carried over.

The lineage column is the part teams get wrong. The two Improving or Maintaining measures and the three HEDIS HOS measures are not reading the same members, so a contract can watch one family move while the other sits still and conclude something is broken when nothing is.

Why HOS matters more for the 2027 Star Ratings

Improving or Maintaining Physical Health and Improving or Maintaining Mental Health were weight 1 in the 2026 Star Ratings because they came back as new measures after a specification change. CMS said in the 2026 Star Ratings fact sheet that they would have a weight of one for the 2026 Star Ratings and a weight of three beginning with the 2027 Star Ratings, and Table IV-1 of the CY2027 Advance Notice lists both at weight 3.

Two measures tripling in the same year is a real change in the arithmetic of a rating. It puts the HOS outcome measures in the same weight class as Plan All-Cause Readmissions, Controlling Blood Pressure, Diabetes Care Blood Sugar Controlled and the three Part D adherence measures. See triple-weighted measures for how weight compounds into a summary rating, and the 2027 Star Ratings changes for the rest of the measure set.

CMS case-mix adjusts both CAHPS and HOS measures, and stated in the CY2027 final rule that Part D adherence measures join them beginning with the 2028 Star Ratings. So a HOS score already accounts for how sick and how old a contract population is relative to others.

HOS compared with CAHPS

Both are member surveys and both feed the Star Ratings, and they are routinely treated as the same thing in planning meetings. They are not.

CAHPS tells you what members think of you. HOS tells you what happened to them. A plan can be well liked and still lose ground on HOS.

Where the HOS data lives

Contract-level reports sit in HPMS under Quality and Performance, then HOS. As of July 31, 2026 that includes the Cohort 26 performance measurement report, the Cohort 26 Star Ratings validation report, the Cohort 26 aggregate score analysis, the 2025 Round 28 HEDIS HOS Effectiveness of Care report, and the beneficiary-level Cohort 26 performance measurement data. Older material is there too: performance measurement reports for Cohorts 23 through 26, HEDIS HOS Rounds 25 through 28, baseline reports for Cohorts 24 through 27, and HOS-M for 2021 through 2024.

Beneficiary-level HOS data is the exception. It does not come from HPMS. It comes from Health Services Advisory Group at HOS@hsag.com, free, delivered as an encrypted CSV over secure file transfer, and the request needs a designated recipient, the plan address, a recipient phone number and your CMS contract numbers. HSAG support is (888) 880-0077 and CMS policy questions go to hos@cms.hhs.gov. Most teams do not know to ask, so they work from aggregate reports when the member-level file was available the whole time.

One more scheduling note: the 2025 Cohort 28 HOS baseline arrives in HPMS in mid-November 2026. If your contract did not participate in Cohort 26 performance measurement or Round 28 HEDIS HOS, there are no contract-specific reports for you this cycle.

What operations teams can actually do

You cannot survey your way to a better HOS score. The sample is random, the questions ask the member about themselves, and the window closed two years before the rating publishes. What is left is the care that happens in between, which means HOS lives in the annual wellness visit and the care management workflow, a long way from a Stars campaign.

1. Work the three HEDIS HOS topics inside routine care. Falls risk, bladder control and physical activity are ordinary visit conversations that often go undocumented and undiscussed. They are also the HOS measures closest to something a team controls.
2. Use the aggregate score analysis as your diagnostic. It benchmarks the contract against state, region and national on average physical and mental component summary scores, share of members reporting worse health than a year ago, problems with two or more activities of daily living, two or more chronic conditions, depressive symptoms, BMI of 30 or higher, sleep, and share deceased within two years of baseline. That last row tends to explain more score movement than anything a quality team did.
3. Request the beneficiary-level file. Aggregate benchmarks tell you where you sit. The member-level file tells you which populations moved.
4. Read the current cycle as a two-year plan. Cohort 28 baseline reports do not arrive until mid-November 2026, so Cohort 26 is what you plan against until then.

For what landed in HPMS on July 31, 2026 and what to check first, see the HOS validation report post.

Related terms

See triple-weighted measures for what a weight 3 measure does to a summary rating, cut points for how a rate becomes a star, and 2027 Medicare Star Ratings changes for the full measure and weight picture that surrounds HOS.

Sources

• Medicare Health Outcomes Survey official site (CMS sponsored, administered by Health Services Advisory Group)
• CMS HPMS memos, week of July 27 to 31, 2026 (2026 Medicare HOS: HPMS update and availability of HOS data and reports, July 31, 2026)
• CMS: CY2027 Advance Notice (Table IV-1, measure list and weights for the 2027 Star Ratings)
• CMS: 2026 Medicare Advantage and Part D Star Ratings fact sheet (weight of one for 2026, three beginning with the 2027 Star Ratings)
• CMS: Contract Year 2027 Medicare Advantage and Part D Final Rule fact sheet (case-mix adjustment of CAHPS, HOS and adherence measures)