Changes in Persistence to Basal Insulin Following the Medicare Out-of-Pocket Cost Cap

ABSTRACT

Objective: To evaluate changes in persistence to basal insulin following the implementation of the $35 monthly out-of-pocket cost cap for insulin for Medicare beneficiaries.

Study Design: We completed a retrospective cohort analysis of a nationally representative Medicare Advantage sample from 2022 to 2023.

Methods: We compared precap rates of persistence to basal insulin in 2022 with postcap rates in 2023 using Persistence to Basal Insulin, a health plan performance measure used in Medicare Part D quality programs, among the study population and multiple subgroups.

Results: A statistically significant increase (+1.3 percentage points; P < .001) was seen in overall patient persistence to basal insulin in the first year of the implementation of the insulin cost cap. Subgroup analyses showed greater increases in persistence in several subpopulations, including patients aged 45 to 54 and 65 to 74 years, male patients, non-Hispanic Black patients, and patients residing in certain census divisions, especially those with a high prevalence of type 2 diabetes.

Conclusions: In addition to providing financial relief to Medicare patients, the implementation of the insulin cost cap was associated with an improvement in patients’ treatment persistence to basal insulin therapy, a measure associated with positive health and economic outcomes. These results lend useful insight into future policy evaluations and proposals that seek to improve treatment access and address disparities in appropriate medication use.

Am J Manag Care. 2026;32(6):In Press

Takeaway Points

The Inflation Reduction Act of 2022 included a provision that capped monthly out-of-pocket insulin costs at $35 for people with Medicare, effective in 2023. We evaluated changes in persistence to basal insulin among Medicare Advantage beneficiaries in the first year of this cost cap.

• The results revealed a statistically significant increase in overall patient persistence to basal insulin—a measure associated with positive health and economic outcomes—in the first year of the implementation of the cost cap.
• Several subgroups experienced greater improvement in persistence to basal insulin, including patients aged 45 to 54 and 65 to 74 years, male patients, non-Hispanic Black patients, and patients residing in certain census divisions.
• Our early analysis suggests that the Medicare insulin out-of-pocket cost cap may have had a positive impact. The observed differences in insulin use among subpopulations can inform policy makers, health plans, and other stakeholders seeking to better understand relationships between cost exposure and treatment persistence.

As the seventh leading cause of death in the US, diabetes incurs a substantial cost—estimated to be more than $412 billion in 2022—to the health care system.¹ Many adults with diabetes eventually require insulin therapy, including basal insulin, to achieve glycemic targets. Diabetes is particularly impactful in Medicare, where 1 in 3 beneficiaries has diabetes and more than 3.3 million beneficiaries use at least 1 form of insulin.² Rates are even higher for certain demographic groups: More than 40% of Black and Hispanic Medicare beneficiaries have diabetes.²

The rising cost of insulin has attracted significant political and media attention in recent years. In response, the Inflation Reduction Act of 2022 (IRA) included provisions to limit out-of-pocket costs for insulin to $35 per month with no deductible for covered insulin products for Medicare beneficiaries,³ a benefit that took effect in 2023 and was expected to improve patient access.

In this study, we evaluate changes in persistence to basal insulin among Medicare Advantage beneficiaries following implementation of the insulin cost cap using Persistence to Basal Insulin (PST-INS),⁴ a health plan performance measure developed and stewarded by the Pharmacy Quality Alliance (PQA). The PST-INS measure evaluates the percentage of individuals 18 years and older who were persistent to treatment with basal insulin during a measurement year. Here, persistence refers to continuing on basal insulin therapy without exceeding a permissible gap.

The PST-INS measure serves as an important tool within the Medicare Part D quality program, administered by CMS, to evaluate the quality of insulin use as part of public reports on Part D stand-alone and Medicare Advantage prescription drug plan quality. This evaluation of the change in PST-INS measure performance can provide valuable insight into the potential effects of cost-saving measures on improving treatment persistence to medications.

METHODS

This retrospective cohort analysis compared rates of insulin persistence during the years immediately preceding (2022) and following (2023) implementation of the insulin out-of-pocket cost cap in a nationally representative sample of Medicare Advantage beneficiaries. Year-over-year changes in persistence were reported overall and stratified by patient characteristics of interest.

Data Sources

Data from January 1, 2022, to December 31, 2023, were included in the analysis. The data were extracted from Inovalon’s Medical Outcomes Research for Effectiveness and Economics Registry (Inovalon MORE² Registry), which captures patients’ complete enrollment, diagnosis, and treatment histories across time and setting. The MORE² Registry is a real-world database that includes medical and pharmacy closed claims sourced from more than 160 health plans that have been statistically deidentified, representing more than 79 billion medical events. Data are longitudinally matched and cover all major US payer lines of business, including 29% of the Medicare Advantage market. These data were supplemented by member demographic and enrollment data from CMS’ Medicare Beneficiary Summary File.

Study Design

The analysis was designed to align closely with the PST-INS measure specifications used in the Medicare Part D quality program to reflect real-world changes in measure performance. The specifications of the PST-INS measure are based on the methodology developed by Wei et al and later validated by PQA, a national quality organization and measure developer, showing that persistence to basal insulin is associated with key outcomes such as reduced hemoglobin A_1c and lower health care utilization and costs.^5,6

The study population comprised Medicare Advantage beneficiaries meeting the PST-INS measure’s inclusion criteria: individuals 18 years and older with at least 1 prescription claim for basal insulin, a treatment period of at least 91 days (defined as the period between a patient’s earliest prescription start date for basal insulin during the measurement year and the end of the measurement year, death, or disenrollment, whichever occurs first), and no more than a 1-day gap in enrollment during the treatment period. Patients with gestational diabetes, end-stage renal disease, or prescription claims for mixed insulin or regular U-500 insulin, or who received hospice care during the measurement year, are excluded from the measure and were therefore excluded here. In line with the yearly implementation of the PST-INS measure in the Medicare Part D quality program, patients could be included in the eligible population for 2022, 2023, or both. Individuals who refilled all basal insulin prescriptions on or before each prescription’s expected refill date during the measurement year were considered persistent.

Rates of persistence were evaluated in the overall study population and various subpopulations. Subgroup analyses calculated rates of persistence by patient sex, age group, race and ethnicity, census division, original reason for Medicare eligibility, and dual eligibility for Medicare and Medicaid and low-income subsidy (LIS) status. Independent 2-tailed z tests were used to compare persistence rates between years.

RESULTS

The analysis included 169,489 Medicare Advantage beneficiaries in 2022 and 142,898 in 2023 (Table 1). Overall rates of persistence to basal insulin showed a modest, statistically significant improvement from 76.7% in 2022 (preimplementation) to 78.0% in 2023 (post implementation) (+1.3 percentage points; P < .001) (Table 2).

Larger improvements in persistence from 2022 to 2023 were observed among numerous subgroups (Table 2). These included patients aged 45 to 54 years (+1.5 percentage points; P = .018) and 65 to 74 years (+1.5 percentage points; P < .001). We also observed differences among census divisions, with large improvements in persistence seen in the East South Central (+4.4 percentage points; P < .001), South Atlantic (+4.2 percentage points; P < .001), East North Central (+2.3 percentage points; P < .001), and West South Central (+2.2 percentage points; P < .001) divisions.

Although women outnumbered men within the total sample, rates of persistence among women did not increase as strongly as among men (women: +1.1 percentage points; P < .001; men: +1.5 percentage points; P < .001). Our results also revealed differences among racial and ethnic groups. At baseline, rates of persistence were lower among non-Hispanic Black patients (73.2%) compared with non-Hispanic White (77.3%) or Hispanic/Latino (79.0%) patients but showed a greater improvement in the year following the implementation of the insulin cost cap (Black: +2.3 percentage points; P < .001; White: +0.9 percentage points; P < .001; Hispanic/Latino: +0.7 percentage points; P = .025).

Patients who were dually eligible/LIS showed the same year-over-year increase in persistence as those who were not dually eligible/LIS (+1.2 percentage points; P < .001). Improvement in persistence was also comparable between those eligible for Medicare based on age (+1.3 percentage points; P < .001) and those eligible based on disability status (+1.1 percentage points; P < .001).

Notably, no statistically significant decreases in persistence from 2022 to 2023 were observed in any of the subgroups assessed.

DISCUSSION

The results of this analysis revealed a statistically significant increase in overall rates of the PST-INS measure during the first year of the implementation of the insulin cost cap. Although this important policy change was expected to provide financial relief to Medicare patients, our results show that its implementation was also accompanied by an improvement in patients’ treatment persistence to basal insulin therapy, a measure that has been associated with positive health and economic outcomes.⁶ Further, small year-over-year changes in performance on measures used in Medicare quality programs can have potentially significant implications for population health.⁷

Rates of persistence to basal insulin in the overall study population increased by 1.3 percentage points from 2022 to 2023, although more pronounced improvements were seen among several subpopulations. These included patients aged 45 to 54 and 65 to 74 years, male patients, non-Hispanic Black patients, and patients residing in certain census divisions.

Black patients are overrepresented in the Medicare Advantage population,⁸ and the significant improvements in insulin persistence shown here are promising for continued improvement in this subgroup’s care. Contrarily, female patients, who are also overrepresented in Medicare Advantage, saw a smaller absolute improvement in persistence than the overall population. Although women had higher rates of persistence than men at baseline in 2022, the gap in performance between men and women narrowed in 2023.

The census divisions that experienced greater improvements in persistence notably include large swaths of the diabetes belt, a geographic area consisting of 644 counties in which residents are more likely to have type 2 diabetes, according to the CDC.⁹ These improvements demonstrate positive changes to basal insulin treatment in 2023 within areas most severely affected by diabetes, where residents are additionally more likely to be African American and have higher rates of obesity and are less likely to have a college degree.⁹

More than half of the study population was dually eligible for Medicare and Medicaid and/or received the LIS; these markers serve as indicators of greater financial need, as patients in this subpopulation typically receive additional financial assistance for their prescription medications. In this analysis, those who were dually eligible/LIS and those who were not dually eligible/LIS showed the same improvement in persistence, suggesting that a $35 monthly cap may not have yielded significant benefit over 1 year for patients who were already shielded from out-of-pocket costs.

In the broader context of the Medicare program, which included 68 million enrollees as of 2024,¹⁰ even modest improvements in markers of care quality can equate to substantial effects on overall health care utilization and costs. A recent evaluation of the Part D Senior Savings Model (PDSS),¹¹ which similarly limited co-payments to a maximum of $35 per monthly supply of each prescribed insulin for beneficiaries of participating Part D plans, showed that the model was associated with an increased probability of being persistent to basal insulin among insulin users in PDSS-participating Medicare Advantage prescription drug plans in 2021 and 2022. Broadly, the model was shown to increase access to, utilization of, and adherence to insulins by lowering insulin costs to patients, which in turn also reduced Part D costs to Medicare. These findings supplement the results we report here, which lend useful insight to future evaluations of the insulin cost cap mandated by the IRA and other policy proposals that seek to improve treatment access, especially as additional medications are considered for price negotiation and cost caps in the Medicare Part D program. Beyond policy, these findings can also inform benefit design and cost assistance programs as health plans seek to better understand potential relationships between cost exposure and treatment persistence.

Limitations and Future Work

Our analysis has several limitations. We were unable to control for patients’ cost exposure before the insulin cost cap was implemented, limiting our ability to discern the cap’s direct impact. Furthermore, unmeasured factors may have affected the changes in insulin persistence we report here. These factors may include other recent efforts to limit out-of-pocket insulin costs, such as the PDSS,¹¹ or changes to rebates within the Medicare program. Although our results cannot be treated as causal, they provide important insights into policy changes and simultaneous real-world changes in patient behavior as measured by national quality programs. Lastly, patients’ behaviors related to dispensing and taking medications may have been affected by the COVID-19 pandemic, which overlapped with this project’s study period.

Future work is needed to evaluate the full impact of the insulin cost cap, including analyses of a broader population of Medicare beneficiaries and multiple years post implementation of the cap.

CONCLUSIONS

This timely analysis shows improvements in persistence to basal insulin in the Medicare Advantage population over the first year of implementation of the Medicare out-of-pocket cost cap, potentially suggesting that the cap may have had a positive impact. In particular, the results of this analysis show that treatment persistence improved most significantly for men, non-Hispanic Black patients, and those in certain census divisions. An understanding of these differences among subpopulations can inform policies to ultimately address disparities in appropriate insulin use, improve outcomes, and promote more equitable care.

Acknowledgments

The authors wish to thank Irene Nsiah, Elizabeth Dowdy, and Noel Hunt for their contributions to this project.

Author Affiliations: Pharmacy Quality Alliance (RO, BS, MC-B, LP), Alexandria, VA; Inovalon Inc (CT, SB, KS), Bowie, MD.

Source of Funding: This project was supported by Eli Lilly and Company and Novo Nordisk. All work was completed independently by the Pharmacy Quality Alliance and its subcontractor, Inovalon, without influence from the organizations providing financial support.

Author Disclosures: Dr Oueini, Mr Shirley, Dr Castora-Binkley, and Ms Pezzullo are employees of the Pharmacy Quality Alliance. Dr Teigland, Dr Bilder, and Mr Sunkari are employees of Inovalon Inc.

Authorship Information: Concept and design (RO, BS, CT, MC-B); acquisition of data (BS, SB, KS); analysis and interpretation of data (RO, BS, CT, MC-B, SB, KS); drafting of the manuscript (RO, BS, SB); critical revision of the manuscript for important intellectual content (BS, CT, MC-B, LP); statistical analysis (BS, SB, KS); obtaining funding (RO, BS, LP); administrative, technical, or logistic support (BS, CT, SB); and supervision (BS, CT, MC-B, SB, LP).

Address Correspondence to: Razanne Oueini, PharmD, 5911 Kingstowne Village Pkwy, Ste 130, Alexandria, VA 22314. Email: roueini@pqa.org.

REFERENCES

1. Parker ED, Lin J, Mahoney T, et al. Economic costs of diabetes in the U.S. in 2022. Diabetes Care. 2024;47(1):26-43. doi:10.2337/dci23-0085

2. Cubanski J, Neuman T, True S, Damico A. How much does Medicare spend on insulin? KFF. April 1, 2019. Accessed March 25, 2025. https://www.kff.org/medicare/issue-brief/how-much-does-medicare-spend-on-insulin/

3. Inflation Reduction Act of 2022, HR 5376, 117th Cong (2022). Accessed March 25, 2025. https://www.congress.gov/bill/117th-congress/house-bill/5376/text

4. 2024 Measure Manual: Persistence to Basal Insulin (PST-INS). Pharmacy Quality Alliance. Accessed March 25, 2025. https://www.pqa.org

5. Wei W, Pan C, Xie L, Baser O. Real-world insulin treatment persistence among patients with type 2 diabetes. Endocr Pract. 2014;20(1):52-61. doi:10.4158/EP13159OR

6. Nsiah I, Campbell PJ, Parikh MA, Hines LE, Pickering M, Nau DP. Persistence to basal insulin: association with health outcomes in a population with type 2 diabetes. Clin Diabetes. 2024;42(2):257-265. doi:10.2337/cd23-0016

7. Borrelli EP, Park MA, Leslie RS. Impact of star ratings on Medicare health plan enrollment: a systematic literature review. J Am Pharm Assoc (2003). 2023;63(4):989-997.e3. doi:10.1016/j.japh.2023.03.009

8. Medicare Advantage demographics. AHIP. January 2025. Accessed June 1, 2025. https://ahiporg-production.s3.amazonaws.com/documents/202501-AHIP_MA-Demographics-Report-v04.pdf

9. Barker LE, Kirtland KA, Gregg EW, Geiss LS, Thompson TJ. Geographic distribution of diagnosed diabetes in the U.S.: a diabetes belt. Am J Prev Med. 2011;40(4):434-439. doi:10.1016/j.amepre.2010.12.019

10. Medicare monthly enrollment. CMS. Accessed March 25, 2025. https://data.cms.gov/summary-statistics-on-beneficiary-enrollment/medicare-and-medicaid-reports/medicare-monthly-enrollment

11. Taylor EA, Khodyakov D, Buttorff C, et al. Part D Senior Savings Model final evaluation, 2021 to 2023. CMS. September 2025. Accessed September 22, 2025. https://www.cms.gov/priorities/innovation/data-and-reports/2025/pdss-final-eval-rpt