A medicine can cost less after Medicare negotiates its price. A generic can also bring down the cost over the same period. Franklin and colleagues used public drug-plan files to compare changes for selected cardiovascular drugs. The story is easiest to read with the comparison group in view.
- The analysis included 6,215 plans, comprising 5,738 Medicare Advantage plans and 477 stand-alone drug plans.
- In the group without a generic option, the estimated extra reduction relative to the comparison group was $56.63 per 30 days. In the group with a newly available generic, it was $7.46.
- The figures use plan rules for the initial coverage phase. Individual pharmacy bills and year-long spending would require different data.
- Paper date
- 19 November 2026
- Plans included
- 6,215 Medicare drug coverage plans
- Cost unit
- US dollars per 30 days of medication
- Policy period
- First year of negotiated prices in 2026
Original publication: 19 Nov 2026 · The date above refers to this brief.
Estimated relative reductions in 30-day cost
US dollars per 30 days, reduction shown as a positive amount
How the cost comparison works
Estimate earlier costs
Calculate an initial-coverage cost for each medication and plan.
Estimate later costs
Observe estimated costs after negotiated prices took effect.
Use the comparator change
Compare the change where prices were negotiated with the change in the comparison group.
What a cost in the plan file represents
The researchers estimated the out-of-pocket cost of a 30-day supply under a drug plan’s coverage rules, for the initial coverage phase. Their dataset contains plan-level estimates. Observed pharmacy bills would require beneficiary records. The findings summarize a set of drugs and plans in the policy comparison.
Source 1 ↗A worked example from the published table
The listed mean for empagliflozin went from $77.29 to $43.21, a drop of $34.08. The mean for the non-negotiated comparator row without a generic went from $76.52 to $99.35, an increase of $22.83. Subtracting that increase from the first change gives −$56.91. This arithmetic uses two rows of the table. The paper’s regression includes additional medication comparisons.
Source 1 ↗Which other price changes were already happening?
The regression’s estimated extra reduction for negotiated drugs without a generic was $56.63 (95% CI $55.48 to $57.78). In the analysis where generic versions appeared, the estimated reduction was $7.46 (95% CI $6.05 to $8.78). In that second analysis, the comparison drug also became available as a generic and its reported mean cost fell. The relative estimate asks how much larger the change was in the negotiated group.
Source 1 ↗What goes into an individual bill
The analyses used different sets of medicines. The two estimates describe how their estimated costs moved relative to their respective comparison drugs. A patient’s cost over time could also depend on the chosen plan, cost-sharing rules, coverage phase and use. The researchers studied a limited set of medicines in the first year of negotiated prices.
Source 1 ↗What a follow-up could measure
Claims linked to enrolment could show how price changes translated into amounts paid by beneficiaries and whether fills changed. A longer time series could help separate changes at generic entry, negotiated price implementation and other coverage changes.
Source 1 ↗Find what was compared before reading a price change
Primary journal research letter. Checked the data source, plan count, estimated 30-day costs, main table, regression comparison and limitations.
01Do the amounts describe pharmacy bills for individual people?
- What was observed
- The reported estimates come from the plan files and apply to initial coverage for 30 days.
- Where the conclusion stops
- Individual and annual costs would require other data.
Source 1 · Methods and Limitations
02Why compare groups with the same generic availability?
- What was observed
- A new generic can change costs in both the negotiated and comparison groups.
- Where the conclusion stops
- Each estimate depends on which drugs and cost changes were included.
Source 1 · Main table and Results
Numbers you can inspect
| Measure | Value & unit | Origin & method |
|---|---|---|
| Plans included | 6215 plans | Reported All included Medicare drug coverage plans. Source 1 · Results |
| Medicare Advantage | 5738 plans | Reported Medicare Advantage drug coverage plans in the analysis. Source 1 · Results |
| Stand-alone drug coverage | 477 plans | Reported Stand-alone plans in the analysis. Source 1 · Results |
| Empagliflozin mean before | 77.29 US dollars per 30 days | Reported Selected unadjusted mean for a worked comparison. Source 1 · Main table, generic not available |
| Empagliflozin mean after | 43.21 US dollars per 30 days | Reported Selected unadjusted mean for a worked comparison. Source 1 · Main table, generic not available |
| Comparator mean before | 76.52 US dollars per 30 days | Reported Mean in the no-negotiation comparator row. Source 1 · Main table, generic not available |
| Comparator mean after | 99.35 US dollars per 30 days | Reported Mean in the no-negotiation comparator row. Source 1 · Main table, generic not available |
| Two-row worked comparison | -56.91 US dollars per 30 days | Calculated (43.21 − 77.29) − (99.35 − 76.52), using reported means. Source 1 · Main table, generic not available |
| Estimated change, generic unavailable | -56.63 US dollars per 30 days | Reported Regression estimate, 95% CI −57.78 to −55.48. Source 1 · Main table and Results |
| Estimated change, generic became available | -7.46 US dollars per 30 days | Reported Regression estimate, 95% CI −8.78 to −6.05. Source 1 · Main table and Results |
Compare the actual experiments
These studies answer different questions. Read the unit and endpoint before comparing results.
| Study | Unit & setting | Readout | Interpretation boundary |
|---|---|---|---|
| Different data for the cost and the care received Source 1 · Methods and Limitations | Drug coverage plans and integrated care boards | Modeled cost and NHS activity claims | A price measure and a service activity count answer different access questions. |
| Using a comparison group to interpret change Source 1 · Methods, difference-in-differences | Plan cost rows and Texas Medicare county outcomes | Estimated medication costs and weekly outcomes | A change during a policy or event means more when the comparison is explicit. |
The export includes claims, available numbers, methods and source locations. It contains our reading notes and published summaries; it is not raw participant data or an independent reanalysis.
Evidence update · 21 Nov 2026
First publication. Any calculations in this brief use reported figures. The reading notes identify possible follow-up questions. AI source check; no human editorial or clinical review.
How much did estimated costs change relative to the comparison?
| Analysis | Estimated change per 30 days | What was compared |
|---|---|---|
| Generic unavailable | −$56.63 (95% CI −$57.78 to −$55.48) | Negotiated versus non-negotiated medicines without a generic. |
| Generic became available | −$7.46 (95% CI −$8.78 to −$6.05) | The comparison also includes new generic availability. |
| Two-row arithmetic example | −$56.91 | One negotiated drug and one comparator row. The full model uses other data as well. |
Model estimates and 95% CIs are from the article’s table. The final row is Press-News arithmetic from selected means.
Your questions, answered
What does the $56.63 estimate represent?
It is the difference between changes in estimated plan costs for negotiated and comparison drugs in the analysis without a generic option.
Where does the $56.91 worked example come from?
Subtract the comparison-row increase of $22.83 from the $34.08 reduction, keeping the signs. The example gives −$56.91.
Does a 30-day estimate describe total annual spending?
The paper estimates 30-day costs for the initial coverage phase. Total spending would also depend on use, coverage and costs throughout the year.
What does the study show about taking medicines as prescribed?
The paper compared plan costs. A study of medication use would need beneficiary-level records or other measures of adherence.
Limits of this interpretation
- The data describe plan rules for an initial coverage phase. They do not track beneficiary-level bills or health outcomes.
- The two models use different groups of drugs and one policy year. Concurrent changes in coverage could affect the estimates.
- The worked example uses two published means. It omits other drugs included in the fitted regression.
Sources & transparency
- Franklin et al. (2026) — Medicare Price Negotiation, Generic Availability, and Out-of-Pocket Costs for Cardiac Medications
Primary journal research letter. Checked the data source, plan count, estimated 30-day costs, main table, regression comparison and limitations. · Accessed 21 Nov 2026
DOI: 10.1001/jamanetworkopen.2026.45071
Prepared and source-checked with AI for Press-news Team. The source notes identify the material available during preparation. The article received no human editorial or clinical review. The source list supports the reported results and the calculations shown.
Source check: AI source check — primary research and stated source limits
Clinical review: Not applicable to this educational guide
Suggest a correction