Date of Award

January 2024

Document Type


Degree Name

Medical Doctor (MD)



First Advisor

Kasia J. Lipska


Nearly thirty million people in the United States have been diagnosed with diabetes. Around 2 million of these individuals have type 1 diabetes and require daily insulin injections to survive. Among the remaining 28 million people with type 2 diabetes, about 20% use insulin while more than 80% are eligible for branded medications, including GLP-1 receptor agonists and SGLT-2 inhibitors, to control their disease and reduce the risk of complications. Medications for diabetes account for more than 200 million prescriptions annually. The number of people with diabetes is expected to grow substantially, and so is their projected medication use. However, prices of insulin and branded medications for diabetes pose significant threats to affordability and access. The list prices of insulin have increased dramatically in the last two decades (e.g., Humalog (insulin lispro)’s price has gone up from $26 to $250 per vial), while list prices for newer branded agents, such as SGLT-2 inhibitors remain high (e.g., Jardiance (empagliflozin) price starts at $578 per 30-day supply). These trends suggest that without major policy reforms, the financial burden associated with diabetes pharmacotherapy will continue to grow for patients and health payors. At the individual level, cost-related nonadherence is already common and results in poor clinical outcomes. Despite this growing problem, existing studies describing spending on medications for diabetes are limited. First, studies are often conducted in specific settings, without presenting national estimates across insurance coverage groups. Second, existing studies fail to consider spending on medications in the context of a patient’s available resources. Finally, few studies have evaluated or estimated the impact of recent policy changes directed at curbing out-of-pocket spending for patients with diabetes. We conducted three analyses to (1) describe spending on medications for diabetes using nationally representative data, (2) examine spending on medications in the context of a patient’s family income, and (3) estimate projected savings under proposed policy solutions. First, we examined out-of-pocket spending on insulin in the U.S. relative to patients’ post-subsistence (or disposable) family income. In this analysis, we identified risk factors that contribute to “catastrophic” spending on insulin, defined as spending more than 40% of post-subsistence income on insulin alone. Second, we estimated out-of-pocket savings for individuals who require insulin, under the provisions of the Inflation Reduction Act (IRA) of 2022, which includes a $35 monthly spending cap on insulin. Third, we examined out-of-pocket spending on the four diabetes medications that will be included in Medicare price negotiations under the IRA, in the context of a patient’s family income and as a proportion of total healthcare expenditures. In our first analysis, we show that, among Americans who use insulin, 14.1% reached catastrophic levels of spending on insulin over one year, representing almost 1.2 million people. Risk of catastrophic spending was higher among those with lower income levels. Nearly two-thirds of patients who experienced catastrophic spending were Medicare beneficiaries. Catastrophic spending was 61 percent less likely among Medicaid beneficiaries than among Medicare beneficiaries after adjustment for various factors (including income), suggesting that insurance design may play an important role in protecting patients from catastrophic spending. The second analysis estimates substantial cost savings under a monthly $35 spending cap on insulin. We estimated that Medicare-insured persons are projected to save nearly one-fifth of their expenditures on insulin under the IRA’s spending cap. Additionally, we found that, if the caps were applied to individuals with private insurance or those without prescription drug insurance, they would save 18.6% and 41.0% respectively on insulin. Our third analysis examines out-of-pocket spending on the four diabetes medications included in Medicare price negotiations (Jardiance, Januvia (sitagliptin), Farxiga (dapagliflozin), and Fiasp/Novolog (insulin aspart)) in the context of a patient’s resources. Many individuals taking these medications reached catastrophic levels of spending, defined as spending greater than 40% of their annual post-subsistence income on one of these medications (8.8% of individuals on Jardiance, 15.8% on Januvia, 11.4% on Farxiga, 15.1% on Fiasp/Novolog). Additionally, out-of-pocket spending on these four medications constituted a large percentage of an individual’s total out-of-pocket spending on all prescription medications. For example, among patients on Jardiance, out-of-pocket spending on Jardiance constituted 24% of all of their prescription medication spending (corresponding percentages were 36% for those on Januvia, 17% on Farxiga, 26% on Fiasp/Novolog). In conclusion, our studies support recent policy reforms, but reveal that additional work is needed to address affordability issues for patients. Although the $35 copayment cap has the potential to significantly curb out-of-pocket spending on insulin for individuals with Medicare, other individuals, particularly the uninsured and those with low annual incomes, need additional protective measures. Newer diabetes medications, which lack generic options, still incur high costs to individuals. More comprehensive safeguards against catastrophic spending are needed for individuals with low incomes and those without prescription drug insurance coverage, who appear particularly vulnerable under the current system.


This thesis is restricted to Yale network users only. This thesis is permanently embargoed from public release.