Most GLP-1 users with diabetes who showed no improvement stayed on the drug, HTA data show
Health Transformation Alliance data show 83% of Type 2 diabetes patients with no GLP-1 gains stayed on therapy, against 94% of responders. For employers setting 2027 terms, poor responders mostly do not leave the drug on their own.
The Board Room · Leadership desk
In HTA data on Type 2 diabetes patients using GLP-1s, HbA1c rose 0.1 percentage point in the less effective group and fell 2.6 points in the more effective group. The less effective group gained 2.7% of body weight on average; the more effective group lost 8.4%.
| Figure | Value | Note | Claim |
|---|---|---|---|
| Less effective group, HbA1c change | 0.1 pt rise | 5 | |
| More effective, HbA1c | 2.6 pt fall | 4 | |
| Less effective, weight | 2.7% gained | 5 | |
| More effective, weight | 8.4% lost | 4 |
What happened
- HTA released the analysis of 12.7 million patients on October 13, 2026, as employers weigh GLP-1 coverage for the 2027 plan year.
- The diabetes responders started with HbA1c above 8%, saw it fall 2.6 points and lost 8.4% of body weight on average.
- In the non-improving group, HbA1c rose 0.1 point, weight rose 2.7%, and patients had an emergency room visit or hospital admission before starting.
- Business Group on Health reported that employer coverage of GLP-1s for weight management fell from 72% in 2025 to 60% in 2026, with 14% dropping or planning to drop it by 2027.
Why it matters
- What it costs cost Self-insured employers paying per prescription keep funding fills for non-responders, which hcamag describes as the pattern with the clearest cost implications.
- What it limits constraint The traits tied to weaker response, higher illness burden and recent hospital use, are the ones that raise nondiscrimination questions if written into eligibility or continued-use rules.
- Who has to decide decision Employers settling 2027 terms choose among broad coverage, added continued-use review and a fixed monthly reimbursement carve-out, and the persistence data bears on the review option more than on the carve-out.
- What doesn't add up contradiction HTA says the analysis is not meant to set coverage criteria, yet hcamag presents it as population data for pharmacy benefit manager and clinical discussions.
Persistence is the figure a benefits team can pull from its own claims, and in this data it is a weak filter. At nine months, 17% of the diabetes patients whose numbers did not improve were no longer on therapy, against 6% of responders [16]. The departure rate is about 2.8 times higher [17], but most of the non-responders remain on the drug [1]. A plan that waits for them to leave on its own will still be paying for most of them at month nine.
The diabetes result does not map cleanly onto the weight management decisions employers are making. The share of employers covering GLP-1s for weight management fell 12 points in a year, according to Business Group on Health [20], while the 83% and 94% figures come from patients using the drugs for Type 2 diabetes [1]. The HTA figures nearer to the weight management question are the obesity ones. Persistence at nine months was 80% for the more effective obesity group and 76% for the less effective one [6][7], a gap of 4 points [18]. Of the four groups HTA reported, the diabetes non-responders were the only one whose average weight rose [4][5][6][7]. The report does not say why most of them stayed.
HTA chief executive Rob Andrews framed the purpose as targeting. "If we can better understand who is likely to see a meaningful health benefit, employers can focus on getting the right treatment to the right people rather than treating every prescription the same," he said [10]. Targeting needs a criterion, and the data offer a thin one. The two diabetes groups differ by 0.2 in average Charlson Comorbidity Index score, 2.1 against 1.9 [19], so the prior emergency room or inpatient stay may do much of the separating. The published account does not say how much.
The evidence behind any such criterion is limited. The study was observational, and HTA noted that it did not measure effects on productivity, time off work, disability or employee retention [8]. Follow-up for most of the results ran for roughly a year, so nobody yet knows whether the drugs pay for themselves over a longer stretch [8]. Merative, whose claims data the study drew on, validated the predictive models internally [9]. Cost pressure for 2027 is already set: Marsh projected that employer health costs will rise 8.2 percent next year, the steepest increase since 2003 [14].
What to watch
- Any HTA or Merative follow-up that extends past about one year or measures productivity, absenteeism, disability or retention.
- Whether employers and their pharmacy benefit managers adopt continued-use criteria after legal review, and whether they cite this analysis when they do.
- Whether the 14% of employers that Business Group on Health says dropped or plan to drop weight management coverage by 2027 follow through.
Clarity's read
What the record supports and how the coverage leans. The claims behind it follow.
Reality
- Evidence58
- Adoption30
- Hype gap+10
- Incentives50
- Confidence60
Claim ledger
Ranked by verification strength, evidence, and original report placement.
- [1]
Among patients using GLP-1 drugs for Type 2 diabetes, the group whose blood sugar and weight did not improve stayed on therapy at a rate of 83 percent, against 94 percent for the group that responded well.
- [2]
HTA released an analysis of 12.7 million patients on October 13, 2026, as employers weigh GLP-1 coverage for the 2027 plan year.
- [3]
HTA stated in its published analysis that the work is not intended to predict individual outcomes or establish coverage criteria.
- [4]
The more effective Type 2 diabetes group started with HbA1c above 8 percent, a lower illness burden (average CCI 1.9) and no emergency room or inpatient visits in the prior six months; HbA1c fell 2.6 percentage points, average weight loss was 8.4 percent, and 94 percent stayed on therapy for at least nine months.
- [5]
The less effective Type 2 diabetes group averaged a CCI of 2.1 and had at least one emergency room visit or hospital admission before starting; HbA1c rose 0.1 percentage point, patients gained 2.7 percent of body weight on average, and 83 percent stayed on therapy for nine months or longer.
- [6]
Among patients with obesity but no Type 2 diabetes, the more effective group lost 9.4 percent of body weight on average, and 80 percent stayed on therapy at nine months.
- [7]
In the less effective obesity group, patients still lost 5.3 percent of body weight on average, 64.4 percent reached the 5 percent mark, and 76 percent stayed on therapy for nine months.
- [8]
The study was observational, and HTA noted it did not evaluate workforce outcomes such as productivity, absenteeism, disability or retention. Most results reflect about one year of follow-up, so long-term return on investment remains uncertain.
- [9]
Merative, the analytics firm whose claims data the study drew on, internally validated the predictive models.
- [10]
"If we can better understand who is likely to see a meaningful health benefit, employers can focus on getting the right treatment to the right people rather than treating every prescription the same," said Rob Andrews, chief executive of HTA.
- [11]
The traits linked to weaker response include higher illness burden and recent hospital use, and eligibility rules built on those factors can raise questions under the federal nondiscrimination requirements for group health plans; employers considering changes to prior authorization or continued-use reviews should have legal counsel and their pharmacy benefit manager examine any criteria before adoption.
- [12]
Business Group on Health reported that the share of employers covering GLP-1s for weight management fell from 72 percent in 2025 to 60 percent in 2026, and 14 percent have dropped or plan to drop that coverage by 2027.
- [13]
Some employers have looked at fixed monthly reimbursement carve-outs for GLP-1 drugs to cap spending without ending access.
- [14]
Marsh projected that employer health costs will rise 8.2 percent in 2027, the steepest increase since 2003.
- [15]
For a self-insured employer paying per prescription, the pattern of non-responders staying on therapy has the clearest cost implications.
- [16]
At nine months, 17% of diabetes non-responders were not on therapy, against 6% of responders.
- [17]
The nine-month departure rate among diabetes non-responders is about 2.8 times that of responders.
- [18]
The nine-month persistence gap between the more and less effective obesity groups is 4 percentage points.
- [19]
The average CCI of the less effective diabetes group exceeds that of the more effective group by 0.2.
- [20]
The share of employers covering GLP-1s for weight management fell 12 percentage points between 2025 and 2026.
Get this ledger inside ChatGPT. Ask about any of our stories and the answer brings its figure, each claim’s verdict and the sources. Free, no account.
Sources
1 independent publisher whose own reporting we read for this story.
- hcamag.comGLP-1 diabetes study: patients without gains often stay on drug
1 article · October 15, 2026
Topics and entities
Follow any of these and your For You feed starts watching them — no settings page required.
Topics
- GLP-1 drugsFollow
- Employer Health BenefitsFollow
- Pharmacy benefit managementFollow