Can Preventive Drug Lists Make High-Deductible Health Plans Safer for People With Diabetes?

The high and rising prevalence, morbidity, and expense of diabetes pose substantial financial burdens on individuals, communities, health systems, and society, calling for innovative policy solutions to mitigate its costs. In this issue of Diabetes Care, Lu et al. (1) present results of a natural experiment in health benefit design, examining the effects of preventive drug lists (PDLs) on medication use and total and out-of-pocket health care expenditures among commercially insured individuals.

Currently, 38.1 million (14.7%) adults in the U.S. are living with diabetes (2), costing the U.S. $412.9 billion annually, or one-quarter of all U.S. health care expenditures (3). By 2030, diabetes prevalence is projected to increase, to 54.9 million adults, as are societal costs, to $622 billion (4). Medication costs are a major driver of health care costs, with >$185 billion spent in 2022 on the nearly 1.1 billion medication prescriptions filled by people with diabetes that year (3). Twenty-eight percent of Americans have reported difficulty affording medications, while 21% have not filled prescribed medications due to cost (5). The financial burden of diabetes management on individuals, however, is highly variable and is driven largely by health insurance coverage (5).

High-deductible health plans (HDHPs)—plans with high annual deductibles but, typically, lower monthly premiums (6)—have become increasingly popular over the past two decades. HDHPs were first introduced in 2001 to control health care costs in increasing consumers’ cost-sharing responsibility and thereby disincentivizing the use of (presumably low-value) medical services. In reality, HDHPs often decreased clinically necessary care, particularly preventive care and treatment of asymptomatic conditions (7). The 2024 annual deductible threshold for HDHPs was $1,600 for individual and $3,200 for family plans, yet nearly 40% of Americans do not have $400 in dispensable cashflow on hand to cover unexpected expenses (8).

When employers offer only HDHP coverage, employees are forced to switch to these plans, which—in people with diabetes—have been associated with decreased monitoring of retinopathy (9,10), blood pressure (10,11), and A1C (10,11) and increased odds of severe hypoglycemia, diabetic ketoacidosis, myocardial infarction, stroke, heart failure hospitalization, lower-extremity complications, kidney failure, and proliferative retinopathy events (12,13). With 29% of adults with employer-sponsored insurance enrolled in HDHPs (14), it is essential to make them safer for people with diabetes.

PDLs have been proposed to temper the financial burden of HDHPs (15). PDLs include a curated set of medications that are considered “preventive” for specific chronic health conditions and therefore covered by HDHPs without the requirement that beneficiaries first spend down their deductible. Whether long-term access to PDLs can mitigate the adverse effects of HDHP enrollment on people with diabetes was—up until now—unknown.

In a rigorously conducted quasi-experimental study, Lu et al. (1) address this question by examining the effects of PDL implementation by some, but not all, employers on total and out-of-pocket medication-related expenditures and medication use among commercially insured adolescents and adults with diabetes across the U.S., with further evaluation for differential effects as a function of beneficiary income level and health savings account (HSA) enrollment. With the assumption that employers’ decision to implement PDLs is independent of employee characteristics, this natural experiment seeks to establish a causal relationship under otherwise uncontrolled (real-world) conditions and where a true randomized experiment would be both unethical and infeasible.

The study findings showed that, consistent with the goal of PDLs, overall out-of-pocket medication costs decreased by 29.7% among PDL members compared with control participants, which was driven—as expected—by a 43.1% decrease in out-of-pocket spending for PDL medications. Concurrently, use of all medications and PDL medications increased by 9.4% and 12.5%, respectively, signaling the beneficial effects of lower out-of-pocket costs on accessibility and affordability of care (1).

The authors further found that high- and low-income beneficiaries saw similar savings in out-of-pocket expenditures with implementation of PDLs, while PDL beneficial effects on medication use (i.e., access) were more pronounced among low-income beneficiaries. This underscores the degree to which out-of-pocket costs hinder access to care for individuals with limited financial reserves. Similarly, while beneficiaries with and without HSAs saw similar savings with PDL implementation, only those without HSAs increased their medication use, suggesting that HSAs may shield beneficiaries from some of the overt financial burden of out-of-pocket expenditures. Still, usage of HSAs requires that beneficiaries have the ability to put money into the HSAs up front (as <7% of employers fund the HSA with the full deductible amount [14]), which may not be feasible for all employees, and 30% of HSA-eligible beneficiaries did not use these accounts, due to lack of awareness, employer policies, or financial constraints (16).

This work is timely and informative. It is imperative to rigorously and independently evaluate public policies to ensure that they achieve the intended goals without harm. Indeed, HDHPs were designed to limit health care expenditures without hindering high-value care; yet, subsequent examination revealed many dangerous consequences (7,9–13,17) as beneficiaries are not positioned to identify clinically “low-value” care and limit all care (18). This study offers reassurance that eliminating cost sharing for evidence-based therapies brings the effect of HDHPs closer to the original intent of prioritizing high-value, evidence-based care. Notably, PDLs only exist in the context of commercial plans; Medicare beneficiaries are ineligible but would likely benefit more, as they are generally ineligible for cost sharing reduction interventions like manufacturer copayment reduction cards (19).

Nevertheless, it is unlikely that PDLs can mitigate all of the harms of HDHPs. PDLs are limited to addressing a small number of health conditions and treatments. Services like diabetes self-management education, medical nutrition therapy, and endocrinology care are not covered, and care for other health conditions is not covered either. Moreover, not all employers offer PDLs as part of available coverage, and there is variability in what conditions are eligible for predeductible coverage. In 2021, 76% of employers reported offering PDLs, up from 29% in 2020, though only 8% included all PDL-eligible services (20). Reassuringly for people with diabetes, glucose-lowering medications were the most commonly included PDL drugs (offered by 66% of employers), but in only 30% of plans would people with diabetes face no out-of-pocket costs for glucose-lowering medications.

The study has limitations, and questions remain unaddressed. In addition to the limitations acknowledged by the authors, it is important to note that analyses focus on 2003–2017, so we must consider how well these findings may apply to today’s insurance landscape and medications used. These analyses also end years prior to Internal Revenue Service Notice 2019-45 and Notice 2024-75, which mandated that HDHPs allow predeductible coverage of glucose-lowering medications, statins, several blood pressure medications, glucose meters, A1C testing, retinopathy screening, LDL cholesterol testing, and blood pressure monitors (15,21). It is likely that contemporary cost savings of PDLs to beneficiaries are greater than observed in this study. Most importantly, we do not know how PDLs impacted medication adherence, choice of pharmacotherapy, and health outcomes, which should be examined in future research.

Health benefit design is an essential part of affordability and accessibility of health care. While there are limitations to this work, research such as this is essential to understanding how health insurance drives health care use and needs to inform policies going forward. Research such as this also underscores the importance of routinely screening people with diabetes for financial barriers to care (22), leveraging the health care team to address these barriers (19), supporting people with diabetes to choose health insurance coverage that meets their needs, and advocating for policy solutions to mitigate the financial burden of diabetes management.