The Unintended Cost of High Cost Sharing in Medicare—Assessing Consequences for Patients and Options for Policy

Cost sharing (e.g., copayments, coinsurance, and deductibles) in health insurance involves a trade-off between financial protection to patients and the costs of potentially inefficient health care use.¹ Low cost sharing protects patients from high and unexpected medical expenditures – a key function of insurance – but it also raises a concern that patients may overuse low-value care when they pay less than its full cost, an inefficiency economists term “moral hazard.” Conversely, higher cost sharing may mitigate moral hazard, yet it exposes patients to greater financial risk and could cause some patients to avoid beneficial care due to concerns about cost.² The challenge of insurance benefit design is to calibrate cost sharing for different services and patient populations to promote the use of services with high clinical benefit, deter low-value care, and provide adequate financial protection to patients.

However, research consistently finds that cost sharing, in practice, is a blunt tool – placing too high a financial burden on vulnerable patients who may benefit from care – rather than a finely-tuned instrument designed to promote efficient health care use.³ A study by Madden and colleagues in this issue of JAMA Health Forum demonstrates that Medicare is no exception.⁴ Medicare includes substantial patient cost sharing, and traditional Medicare does not cap out-of-pocket spending, leaving beneficiaries exposed to potentially thousands of dollars of out-of-pocket costs. Using the 2017 Medicare Current Beneficiary Survey, Madden and colleagues found that 16% of Medicare beneficiaries reported difficulty affording health care, defined as delaying care due to concerns about cost or having trouble paying medical bills. One of the paper’s most striking findings is how affordability burdens were concentrated among medically and socially vulnerable beneficiaries. For example, the non-elderly disabled (i.e., aged <65 years), those with multiple chronic conditions or functional limitations, and the near-poor (i.e., beneficiaries with $15,000-$25,000 in annual income), had 2-3 times the odds of reporting difficulty affording care than healthier or wealthier individuals.⁴ Paradoxically, Medicare beneficiaries who may be least able to forgo care are also the least able to afford care.

Madden and colleagues did not study health outcomes, but prior research demonstrates that even moderate cost sharing can have adverse effects on health when vulnerable patients forgo essential care that costs more. For example, a study of Medicare supplemental plans in California found that patients reduced physician visits and medication use following the introduction of copayments for physician visits ($10 per service, in 2001 dollars) and increases in copayments for prescription drugs (by $5 per fill). Hospital stays increased, particularly among sicker patients, suggesting worsening health.⁵ A separate study of Medicare Advantage plans from 2001 to 2006 found that a doubling of copayments for outpatient visits (from a mean of approximately $7 to $14 for primary care and from $13 to $22 for specialty care) led to a 3% reduction in outpatient use and a nearly 9% increase in hospital admissions, likely increasing overall medical spending. The increase in hospitalizations was greatest among patients residing in low-income ZIP codes and with chronic conditions such as diabetes.⁶

Conversely, the RAND Health Insurance Experiment found that low-income patients with poor baseline health benefited from enrolling in health plans with no cost sharing, for whom free care facilitated access to physicians and treatments for health conditions such as hypertension.¹ More recent experimental research shows that eliminating cost sharing for prescription drugs can improve health outcomes among high-risk patients.⁷ Viewed in the context of this evidence, the findings by Madden and colleagues underscore concerns that Medicare’s cost-sharing structure may be inefficient because high-risk patients may be unable to afford care that is important for managing their health.

Cost sharing in Medicare is a blunt tool, but there is also substantial heterogeneity in who pays these costs, which may account for patterns in the findings Madden and colleagues reported. Medicare beneficiaries with incomes ≤100% of the federal poverty level (FPL) and limited assets qualify for Medicaid supplemental insurance, which eliminates cost sharing for inpatient and outpatient care. Beneficiaries with incomes ≤150% of FPL may be eligible for prescription drug subsidies.⁸ Medicare beneficiaries whose income exceeds these thresholds are ineligible for this assistance, which may explain why Madden and colleagues found that the near-poor were especially likely to report difficulty affording care.⁴ Beneficiaries may also limit their out-of-pocket costs by obtaining supplemental insurance through a former employer or a Medigap plan (the latter is limited to beneficiaries with traditional Medicare). However, these sources of supplemental coverage largely exclude non-elderly disabled beneficiaries. Disabled Medicare beneficiaries are less likely to have accumulated sufficient work history to receive retiree health benefits, and Medigap plans are generally not required to issue coverage at community-rated premiums (i.e., premiums that are not adjusted for health risks) until a person turns age 65.⁹ Alternatively, Medicare beneficiaries can enroll in a Medicare Advantage plan in lieu of traditional Medicare. Medicare Advantage plans often have lower cost sharing than traditional Medicare, and out-of-pocket costs are capped. However, the average out-of-pocket spending limit among Medicare Advantage plans exceeds $5,000.

Thus, Medicare beneficiaries who face higher out-of-pocket costs tend to be socially and medically vulnerable individuals for whom Medicare’s cost sharing may be inefficient. Because the haphazard structure of who pays this cost sharing is partly an artifact of policy (e.g., underwriting regulations in the Medigap market and eligibility rules for Medicaid), it can be remedied with policy. Recent legislative proposals include important reforms. For example, the Elijah E. Cummings Lower Drug Costs Now Act, recently re-introduced before the 117^th Congress, proposes to cap Medicare beneficiaries’ annual out-of-pocket costs on prescription drugs.¹⁰ Other legislation proposes to increase eligibility for prescription drug subsidies to near-poor Medicare beneficiaries (incomes up to 200% FPL). In addition to these reforms, two other policy changes may be useful. First, expanding eligibility for Medicaid’s cost-sharing assistance to near-poor Medicare beneficiaries could help lower costs for those who may not have affordable alternatives for supplemental coverage.⁸ Second, introducing a standard out-of-pocket maximum in Medicare that is tied to income (e.g., a higher dollar limit for people with higher incomes) would limit patients’ exposure to catastrophic costs, an essential function of insurance. This change would also establish greater consistency between Medicare Advantage, where out-of-pocket costs are capped, and traditional Medicare, which does not feature such financial protection.

Cost sharing can serve an important purpose in controlling health care spending, but how it is implemented can be detrimental to vulnerable patients. Ultimately, poorly designed insurance can be costly to patients (both in terms of finances and health) and to insurers that absorb the higher costs of care for individuals with poorly managed health conditions. In Medicare, a blunt cost-sharing structure collides with a series of fragmented supplemental insurance programs and policies that place a higher cost burden on sicker patients with low-to-modest incomes. Madden and colleagues have identified populations where this strain is greatest and where targeted reforms could yield tangible benefits.