Among Medicare beneficiaries over 65, one in six lives with depression, according to CMS’s own coverage guidance. Most of them go untreated. Starting with the 2027 measurement year, this persistent gap in care will have a meaningful impact on Star Ratings.

CMS finalized a new measure called DSF-E, Depression Screening and Follow-Up, in the Contract Year 2027 Medicare Advantage and Part D rule. It draws on 2027 performance, lands in your 2029 Star Ratings, and flows into 2030 quality bonus payments. That sounds far off. It isn’t. The infrastructure to screen a full Medicare Advantage population and manage what comes after takes longer to stand up than most plans expect, and the 2027 measurement year opens in a matter of months.

What DSF-E Actually Measures

DSF-E scores two measures and averages them into a single rate.

Screening. Did your plan screen enrolled members for depression using a validated tool and capture the result as a structured, coded score? There’s no requirement that screening happen at a visit. A member who completes a screener by text, at home, with no provider involved, counts the same as one screened in an office.

Follow-up. Did members who screened positive get documented, qualifying follow-up within 30 days? Six pathways count, including a behavioral health encounter, a documented case management contact, a confirmed antidepressant fill, or a same-day step-up from a brief screener to a full one that lands negative.

Here’s the detail that changes how a plan should approach this measure. Claims can’t produce the screening rate. A billed visit tells CMS a member showed up. It says nothing about whether anyone handed them a validated instrument or what they scored. Only a structured clinical result, delivered through a real screening encounter, counts. Plans that lean on their existing claims-based quality infrastructure will not meet the requirement.

Poor DSF-E Performance Means Lost Stars 

DSF-E carries a single measure weight out of roughly 85 points that make up a full MA-PD rating, so it’s worth about 1.2% of the total. But because a Star Rating is an average, what matters isn’t the measure’s weight. It’s how significantly your DSF-E score differs from your overall Medicare Star Ratings.

Take a plan sitting at a 4.0 average across all other categories. A 1-star result on DSF-E pulls the overall rating down by roughly 0.035 stars. A 5-star result only adds about 0.012 stars. The downside is close to three times the upside. Doing nothing, given where national screening rates sit today, is the likeliest path to that 1-star outcome.

That asymmetry matters most near a rounding line. CMS rounds the overall rating to the nearest half star, and the money sits almost entirely on the far side of 3.75. At 4 stars, CMS raises the benchmark a plan bids against by up to 10% in some counties, and the rebate share a plan keeps steps from 50% at 3 stars or below to 65% at 3.5 and 4.0. Quality bonuses across Medicare Advantage totaled more than $12.7 billion in 2025, close to $490 per member. For a plan with 100,000 members, crossing into bonus territory is worth roughly $50 million a year. A plan sitting at 4.20 has room to absorb a bad DSF-E score. A plan at 3.73 does not.

What Closing the Gap Actually Takes

Screening a Medicare population at scale and managing a 30-day follow-up clock are two different operational problems, and most plans are set up for neither.

Screening is the biggest area of growth for DSF-E Star Ratings. According to NCQA’s analysis of 2024 HEDIS data, health plans only screened 17% of eligible members for depression, but they successfully followed up with 67% of positive screens. Increasing screening volumes is a relatively easy and meaningful way to increase DSF-E Star Ratings.

NeuroFlow delivers PHQ-2 and PHQ-9 screenings directly to members over SMS, email, a member app, and AI-driven voice calls, with a same-session step-up from PHQ-2 to PHQ-9 built into the flow. Members verify with a date of birth. No account, no sign-in, no clinical staff time required to administer it. This is the same infrastructure behind NeuroFlow’s Behavioral Health Lab program with the Department of Veterans Affairs, which reached a person-to-person connection rate above 60% in 2025 and delivered 3.9 million completed assessments, on pace for 4.3 million in 2026.

One assumption worth retiring here is that older members won’t engage digitally. In a six-month program with a Florida Medicare Advantage population of roughly 50,000 members, the most engaged were members aged 70 to 79, ahead of every other age group measured.

Claims data can’t meet the 30-day follow-up window. Medical claims typically arrive 60 to 90 days after a visit, well past DSF-E’s 30-day follow-up window. A plan monitoring this measure through its normal claims cycle will discover a gap after it’s already closed or already missed. NeuroFlow runs on a clinical feed instead, tracking screening results and case management encounters in near real time so a plan can see an open gap while there’s still time to act on it. 

Across NeuroFlow’s screening population of more than 6.5 million members, 61.4% have completed a screening, with 82.7% of referred members successfully contacted and 51.2% reaching a completed follow-up. The average time from a positive PHQ-9 to completed follow-up runs 15 days, and 11 days for members who complete the process through self-service channels alone, well inside the 30-day window.

Follow-up also has to reach members with no provider in the picture. The hardest case in this measure isn’t the member already in therapy, who will likely see their provider inside 30 days without any intervention at all. It’s the member who screens positive through a text message with no PCP relationship, no referral, and no prescription on file. For that member, a documented case management encounter is the best way to meet the follow-up requirement. 

NeuroFlow provides this level of care navigation. A separate matched study with Independence Blue Cross found that members connected through NeuroFlow’s care navigation model were 68% more likely to receive outpatient behavioral health care. These members were also 43% less likely to be admitted for a behavioral-health-related inpatient stay, and 35% less likely to have a behavioral-health-related emergency visit, with both reductions growing over 18 months of follow-up.

The Window to Act Is Now

CMS typically runs a new measure through a display period before it counts toward Star Ratings, and that window is where high performers separate themselves from everyone else. Plans that wait for the measurement year to start reporting, figure out their eligibility logic, their exclusion criteria, and their data feed under live conditions. This leaves them with no room to fix what doesn’t work. Plans that start now find those problems while they’re still cheap to fix.

DSF-E won’t move your Star Rating on its own. But a bad score on it will, and the plans best positioned in 2029 are the ones building the screening and follow-up infrastructure today, not after the measurement year has already started counting.

Talk to our team about preparing for DSF-E.

Related posts:

The 3 Causes of Behavioral Health Referral Leakage And What They’re Costing Your Health System How We Use Data to Guide Product Design at NeuroFlow 19% of Prudential Disability Claimants Reported Previously Unidentified Mental Wellness Needs NeuroFlow Named to TIME’s World’s Top HealthTech Companies 2025