Effects of Co-payment on Prescription Drug Demand

Switching from a proportional coinsurance payment formula to a formula based on flat co-payments in the soon-to-be-enacted Medicare prescription drug benefit may lead to increased compliance, while maintaining budget-neutrality.

In the United States, overall drug spending in the private sector rose by 15-20 percent per year during the 1990s. In 2002 alone, national spending on prescription drugs exceeded $160 billion. Because of these increased costs, some employers and insurers are redesigning health insurance plans to allow for greater sharing of costs with consumers. In particular, they are moving from insurance systems with flat co-payments (where policy holders pay a specific dollar amount per prescription) to systems of coinsurance (where policy holders pay a fixed percentage of the total prescription price).

Studies have shown that greater cost sharing reduces consumer spending on prescription drugs and therefore lowers employer costs. Others have argued that this trend does not undermine overall health, since patients facing rising costs will reduce consumption of nonessential medications more than consumption of vital chronic care drugs.

However, the case of medications with preventive benefits, that is that prevent future illnesses and complications, may be different. Here, patients do not always understand the long-term benefits of continued use. "In this case, underutilization may be the problem, and 'too much' cost-sharing may lead to a loss of welfare," Avi Dor and William Encinosa argue in Does Cost Sharing Affect Compliance? The Case of Prescription Drugs (NBER Working Paper No.10738.)

The authors explore the impact of cost sharing on "compliance" -- that is, on the adherence to refilling prescriptions for preventive care drugs. They examine the effect of rising cost sharing as well as the differing impact of co-payment and coinsurance systems. They focus on diabetes, a chronic condition affecting 16.2 million Americans and a leading cause of adult blindness, kidney failure, amputations, and heart disease. "As the incidence of diabetes reaches epidemic proportions, leading to spiraling costs, the need to undertake preventive measures is becoming even more pronounced."

Dor and Encinosa collect a sample of 27,057 adult individuals with chronic type II diabetes who need ongoing oral medications. Of these, 20,494 people are covered by insurance plans requiring consumers to pay a flat co-payment per prescription, while the 6,563 remaining individuals are covered by coinsurance plans. The authors divide the individuals into three groups: non-compliers, who do not refill their prescription within 90 days of the first prescription running out; partial compliers, who buy at least one more prescription within 90 days but not enough to cover the full 90 days; and fully compliant individuals, who buy one or more prescriptions within 90 days and cover all 90 days.

Under the co-payment regime, about 54 percent of patients were fully compliant within the first week of the 90 days following the end of the prior prescription. By the fourth week, more than 60 percent were fully compliant. This figure "tapers off" to about 58 percent by the end of the 90 days. Under the coinsurance model, individuals display a similar pattern, except that compliance is systematically about 10 percent lower. In the first week following the end of the prescription, about 44 percent fully complied. By the fourth week, 50 percent were fully compliant, tapering off to 48 percent by the end of the 90 days. And over the course of the 90 days, about 14 percent of the initial non-compliers eventually became partially compliant, similar to the proportion (15 percent) under the co-payment regime.

What happens when consumers shoulder an increased portion of the costs? In the co-payment sample, a simulated increase from $6 to $10 in the up-front co-payment results in a 6.2 percent increase in noncompliance and a 9 percent reduction in the share of fully compliant persons. In the coinsurance sample, a proportional increase in the coinsurance rate (from 20 percent to 75 percent) results in a 9.9 percent increase in non-compliance and a 24.6 percent reduction in fully compliant individuals. In another test, holding the out-pocket share constant at $15 in both samples leads to a much higher non-compliance in the coinsurance sample (45.3 percent compared with 35 percent in the co-payment sample).

Dor and Encinosa then perform a rough cost-benefit analysis of an increased co-payment. Under a $6 co-payment, the national costs of compliance are equivalent to $632.7 million per 90 days, while the costs fall to $601.5 million under a $10 co-payment. The savings from increasing the co-payment thus reach $31.2 million per 90 days, or $124.8 million per year, equal to 4.9 percent of annual national spending on anti-diabetic medications. However, the increase in non-compliance rates as a result of the higher co-payment would also increase the rate of diabetic complications (such as blindness, heart disease, and the like) by an estimated $360 million per year, far exceeding the initial cost savings.

"The results suggest that increasing cost sharing leads to greater non-compliance, and to lower compliance in both regimes," explain the authors. However, the negative effects of cost sharing on non-compliance are much larger in the coinsurance regime than under the co-payment system. The authors hypothesize that this effect is attributable to greater uncertainty regarding the out-of-pocket costs for consumers under the coinsurance regime.

Dor and Encinosa conclude that efforts to impose higher cost sharing for prescription drugs may ignore the role of prescription drugs in preventing future medical complications from chronic conditions. Also, since non-compliance is higher under coinsurance than under fixed co-payment systems, both private and public insurance providers may be able to reduce overall medical costs by switching from coinsurance to co-payment systems. "Apparently, many employers are moving in the wrong direction," Dor and Encinosa note. In 2002, some 19 percent of employers offering prescription drug benefits to employees switched from co-payments to coinsurance when the contracts expired. The authors also suggest that switching from a proportional coinsurance payment formula to a formula based on flat co-payments in the soon-to-be-enacted Medicare prescription drug benefit may lead to increased compliance, while maintaining budget-neutrality

-- Carlos Lozada