OBJECTIVE: To decompose the overall effect of multitiered formularies on drug utilization and spending into the following 2 observed effects on consumer behavior: first, higher copayments on drug equivalents create an incentive to reduce the number of prescriptions, and, second, wider differential copayments between drug equivalents create an incentive to use a greater proportion of generics.
STUDY DESIGN: We merged drug claims for 352 760 retired Medicare enrollees having employersponsored health insurance with benefit information. Our unit of analysis was the enrollee. We used cross-sectional variation in incentive-based formularies to compare the effects of increased copayment amounts for drug equivalents with those of increased copayment differentials between drug equivalents. The study sample may not be representative of the Medicare population.
METHODS: Multivariate regression analysis using the 2002 MarketScan Medicare Supplemental and Coordination of Benefits database and Benefit Plan Design database.
RESULTS: A 10% increase in copayments for drug equivalents was associated with a 1.3% reduction in total drug spending, a 16.0% increase in out-of-pocket expenditures, a 2.0% reduction in the number of prescriptions filled, and a 0.7% reduction in proportion of prescriptions filled with generics. A 10% increase in copayment differentials between drug equivalents was associated with a 1.0% reduction in total drug spending, a 4.1% increase in out-of-pocket expenditures, a 1.0% reduction in the number of prescriptions filled, and a 0.7% increase in proportion of prescriptions filled with generics.
CONCLUSION: Increasing copayment differentials between drug equivalents is as effective a strategy for reducing total drug spending as increasing copayment amounts for drug equivalents but better maintains access to prescription medications.
(Am J Manag Care. 2007;13(part 2):353-359)
The results of this study show that focusing on cost-sharing differentials between tiers (as opposed to raising all copayment amounts) and linking them to lower-priced drug equivalents is a similarly effective strategy for reducing drug spending but does a better job of maintaining access to prescription medications.
Most existing studies indicate that multitiered formularies reduce total drug expenditures and the number of prescriptions purchased.
However, most studies fail to decompose the overall effect of multitiered formularies into its absolute price effect (ie, an increase in copayments for all drug equivalents) and its relative price effect (ie, an increase in the copayment differential between drug equivalents).
Understanding the dual incentives of multitiered copayment structures helps identify new opportunities for designing drug formularies based on value.
Incentive-based formularies have become the most common tool for controlling health plan spending for prescription medications.1,2 Almost three quarters of workers covered by employer-sponsored drug plans face incentive-based formularies. Incentive-based formularies are designed to control spending on prescription medications by assigning differential copayments or coinsurance rates to multiple classes, or tiers, of drugs based on the type of drug prescribed, contracts between plans and drug manufacturers and dispensing agents, and relative discount prices between brand-name drugs and generic equivalents.3 Multitiered formularies are expected to reduce plan exposure and to lower spending by introducing incentives for consumers to use less expensive preferred brand-name drugs or generic equivalents when available, to limit their purchase and refill of prescriptions with minor or uncertain therapeutic value, and to shift some of the costs of purchased drugs from plan to consumer.4
Evidence shows that the implementation of incentive-based formularies has been successful in slowing the rate of growth in total drug spending per person per year.5-8 Studies also demonstrate that more aggressive cost sharing lowers the number of prescriptions purchased.9-15 The reduction in the number of drugs purchased raises concern about the effect of multitiered formularies on health outcomes, particularly among enrollees who rely on consistent and timely access to prescription medications for the treatment of chronic conditions. While these studies imply that access to medications may be harmed by enrollee cost sharing, they fail to consider the dual incentives created by multitiered formularies and to assess their conflicting effects on drug utilization and costs.
Incentive-based formularies create 2 incentives. First, by increasing the copayment amount for drug equivalents, multitiered plans are designed to promote more efficient use of prescription medications. By requiring members to pay a larger proportion of the costs of all drugs, plans assume that enrollees will curtail the consumption of unnecessary medications (and physicians will cease prescribing drugs with uncertain medical benefit16,17). Second, by widening the cost differential between drug equivalents, multitiered plans are intended to encourage the consumer to substitute cheaper generic drugs for more expensive brand-name drugs (and less expensive multisource brand-name drugs for their higher-priced equivalents). By requiring enrollees to pay more for brand-name drugs, plans hope that enrollees will be more willing to buy generic equivalents if they exist or will choose preferred brand-name drugs over nonpreferred brand-name drugs if generics are unavailable.
The effect of incentive-based formularies on access to medications and, ultimately, on health outcomes depends on the relative effect of each of these 2 incentives on consumer behavior. The first incentive, which we term an absolute price effect, is expected to achieve a reduction in overall drug spending by reducing the total number of prescriptions purchased. The second incentive, which we term a relative price effect, is expected to achieve a reduction in drug spending by substituting lower-cost generic or preferred brand-name drugs for more expensive drugs without necessarily affecting the total number of prescriptions filled.
The hypothesized effects of the 2 incentives are summarized in Table 1. By increasing the price paid by consumers for drug equivalents, consumers' real income is reduced, and the absolute price effect is assumed to lead to a reduction in total drug spending and a reduction in the number of prescriptions filled. Given that demand for prescription medications is fairly inelastic,18 the absolute price effect should also lead to an increase in enrollee out-of-pocket expenditures and, assuming the relative price of drug equivalents remains constant, should have little effect on the proportion of prescriptions filled with generics. By widening the difference in prices between drug equivalents, the relative price effect should also reduce total drug spending. However, by encouraging reliance on generics, it should have a smaller effect on the total number of prescriptions purchased. Assuming that effective drug equivalents are available at no additional cost, an increase in the copayment for a brand-name drug relative to its generic equivalent should have little effect on out-of-pocket expenditures and, in fact, may lower enrollee spending as consumers substitute generics for brand-name equivalents.
The objective of this study was to decompose the overall effect of incentive-based formularies on drug utilization and spending into its absolute and relative price effects. The absolute price effect is measured by a change in drug use and spending for a given change in the copayment amount for drug equivalents. The relative price effect is measured by a change in drug use and spending for a given change in the copayment differential between drug equivalents. The decomposition analysis is intended to identify strategies for containing drug costs without undermining access to medications.
Sources of Data
The sample consists of 5 plan categories, each with a different copayment structure for prescription drugs. Two plan categories have a 1-tier system (with $5 and $10 copayments), and 3 plan categories have a 3-tier system (with $5/$15/$25, $10/$15/$30, and $10/$25/$35 copayments) (Table 2). For our sample of elderly Medicare beneficiaries, there were no 2-tier plans in the MarketScan database. Each firm in our sample offered only 1 copayment structure for retail purchases, eliminating the risk of bias caused by enrollees' self-selecting into plans based on health status and drug utilization patterns. However, evidence suggests that firms may design their drug benefit based on the health status and service needs of their retirees.19 For instance, firms with older retirees or retirees with known and significant medical conditions requiring maintenance medications may adopt more aggressive cost sharing as a way of limiting plan payments. Our data show that retirees enrolled in 3-tier plans seem to be sicker than those enrolled in 1-tier plans. To test for firm-level selection, we ran our models with and without the health status control variable, and the results were similar, suggesting that remaining bias after adjusting for observable health status is likely to be small.
Sampled BeneficiariesThe sample includes 352 760 elderly Medicare enrollees with employer-sponsored retiree drug coverage. The study is based on retired Medicare beneficiaries and their dependent spouses 65 years and older who receive prescription drug coverage under an employer-sponsored health plan. Actively working enrollees and those younger than 65 years are likely to use prescription medications differently from the older retired population and are therefore excluded from the analysis. Most members remained enrolled for all 12 months, and more than 90% of enrollees submitted at least 1 pharmacy claim.
The models were estimated using generalized least squares to adjust for firm-level correlation in the error terms. The dependent variables (except for proportion of prescriptions filled with generic drugs) were annualized by dividing by the proportion of the 12-month period an individual was enrolled in the plan. Because of the large sample size and the high proportion of drug claimants, expenditure models were estimated on actual (nonlogged) US dollars over all enrollees. Nonclaimants were assigned a value of $0. The unit of analysis was the individual.
Table 3 summarizes the sample distribution by type of prescription drug plan. Eighty-two percent of the sample was enrolled in a 1-tier plan and 18% in a 3-tier plan. Members of 3-tier plans were more likely to be older and female. Moreover, 3-tier plan enrollees had higher mean risk scores than those in 1-tier plans, although both groups had lower risk scores than the average Medicare beneficiary nationally. Ninety-one percent of all enrollees in 1-tier plans submitted at least 1 claim, compared with 89% of those in 3-tier plans. However, individuals in plans with more aggressive cost sharing filled fewer prescriptions on average than those in plans with lower cost sharing, even after controlling for days' supply. Beneficiaries in 1-tier plans filled on average 45.5 prescriptions within the year, compared with 37.9 prescriptions among 3-tier enrollees. The mean total drug costs in 1-tier plans were $2188, compared with $1823 in 3-tier plans. In contrast, the mean enrollee drug payments in 3-tier plans were almost double the amount paid by beneficiaries in 1-tier plans: enrollees in 1-tier plans paid on average $245 per year, compared with $469 among those in 3-tier plans. The planlevel differences were significant at the 1% level.
The effects of higher copayment amounts (absolute price effect) and of wider copayment differentials (relative price effect) are given in the last 2 rows of Table 4. An increase in the absolute price for drug equivalents and an increase in the relative price between drug equivalents result in a reduction in total payments. However, the source of the savings differs. The savings from an increase in the copayment amount are achieved mainly through a reduction in the number of prescriptions filled, while the savings from an increase in the copayment differential are achieved through greater reliance on generic drugs, as well as a reduction in the number of prescriptions filled. In fact, widening the copayment differential increases the proportion of prescriptions filled with generics.
To compare the relative magnitudes of the absolute and relative price effects, the regression results were used to simulate the following 2 equal changes: a 10% increase in the copayment amount and a 10% increase in the copayment differential. The results, summarized in Table 5, show that increasing the copayment amount for drug equivalents and increasing the copayment differential between drug equivalents are broadly similar in their effectiveness in achieving total drug savings. A 10% increase in the copayment amount resulted in a 1.3% reduction in total drug spending, compared with a 1.0% savings generated by a 10% increase in the copayment differential. Given the inelastic demand for prescription medications, raising the price of drug equivalents resulted mainly in a transfer of costs from the plan to the consumer. A 10% increase in the copayment amount for drug equivalents increased out-of-pocket expenditures by 16.0%. In contrast, as consumers are incentivized to shift to generic substitutes, a 10% increase in the copayment differential between drug equivalents increased out-of-pocket spending by only 4.1%.
The simulation further highlights that increasing the copayment differential has less of an adverse effect on access to prescription medications. A 10% increase in the copayment differential between drug equivalents reduces the number of prescriptions filled by 1.0%, compared with a 2.0% reduction in purchased prescriptions associated with a 10% increase in the copayment amount for drug equivalents (Table 5). These results suggest that consumers are better able to maintain their prescription medication utilization pattern by substituting less expensive generics for more costly brandname drugs. A 10% increase in the copayment differential between drug equivalents increases the proportion of total drugs filled with generic substitutes by 0.7%, whereas a 10% increase in the copayment amount for drug equivalents actually reduces the proportion of drugs filled with generics by 0.7%, although this was not statistically significant.
By shifting more of the cost of prescription medications onto the consumer, incentive-based formularies are designed to lower drug spending and to limit plan exposure. However, underlying multitiered formularies are 2 incentives with different and conflicting effects on consumer behavior, namely, an incentive to reduce the number of prescriptions purchased and an incentive to use a greater proportion of generic substitutes. The results of this analysis suggest that the second of these incentives (ie, a change in the copayment differential between generics and brand-name equivalents or between preferred and nonpreferred brand-name equivalents) provides a similarly effective strategy for reducing total drug spending but better maintains access to essential medications.22 The findings indicate that increasing the copayment amount for drug equivalents reduces utilization of prescription medications and passes much of the cost of prescription medications onto consumers. In contrast, by encouraging generic substitution, increasing the copayment differential between drug equivalents reduces total drug spending, with less of an adverse effect on the number of drugs purchased or on out-of-pocket spending.
The study illustrates that how incentive-based formularies are structured has a major effect on access to prescription medications. By focusing on differences in copayment amounts between tiers and by structuring those tiers around drugs with generic substitutes, plans may be better able to control spending without creating barriers to medication access. Furthermore, by showing that consumers are willing to alter their consumption behavior in response to relative price incentives, the study lends support to recent recommendations for replacing price-based formularies (in which copayments are tied to the price of the drug) with value-based formularies (in which copayments are tied to the costeffectiveness or therapeutic value of the drug).22-25 By building value-based measures of drug equivalents into copayment structures, plans may be better able to lower costs, maintain access, and improve therapeutic outcomes.
Several limitations of this study should be noted. First, some plans offered alternative cost-sharing options for mail order and for out-of-network purchases. Plans also varied along other cost-containment dimensions such as deductibles, out-ofpocket maximums, formularies, utilization review, and prior authorization, as well as in their controls on medical service use and costs. Although we control for differences in some of these mechanisms, the full range of utilization management strategies is likely to influence drug consumption in ways not accounted for in our model. Second, MarketScan's retiree sample is drawn primarily from large unionized firms in the manufacturing and durable goods industries located in a few states. Furthermore, sampled beneficiaries have employer-sponsored supplemental health insurance with a prescription drug benefit. Although the findings help illustrate the relative effect of absolute versus relative price effects under incentive-based formularies, they may not be generalizable to the Medicare population. Third, firms are likely to design health plan benefits around the medical needs of their retirees, and unobservable firm-level selection might cause an underestimation of the effect of relative price incentives. Fourth, the sample was drawn before the implementation of Medicare Part D. The availability of prescription drug coverage under Medicare affects the consumption behavior of beneficiaries and the offer rate and design of retiree health benefits by employers in ways that need to be addressed.
AcknowledgmentsWe thank Robert Baker, BA, for his excellent computer programming skills and Susan Haber, ScD, whose initial ideas provided the basis for this study.
Author Affiliations: From Mathematica Policy Research, Inc, Cambridge, Mass (BHG), and RTI International, Waltham, Mass (JK).
Funding Source: This study was funded by an internal grant from RTI International. The authors have no financial interests in the results of this study. Preliminary results of the study were presented at the AcademyHealth 2006 Annual Meeting; June 25, 2006; Seattle, Wash.
Correspondence Author: Boyd H. Gilman, PhD, Mathematica Policy Research, Inc, 955 Massachusetts Ave, Ste 801, Cambridge, MA 02139. Email: email@example.com.
Author Disclosure: The authors (BHG, JK) report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter discussed in this manuscript.
Authorship Information: Concept and design (BHG, JK); acquisition of data (BHG, JK); analysis and interpretation of data (BHG, JK); drafting of the manuscript (BHG, JK); critical revision of the manuscript for important intellectual content (BHG, JK); statistical analysis (BHG, JK).1. Kaiser Family Foundation and Health Research and Educational Trust. Employer Health Benefits 2005 Employer Survey. September 2005. Available at: http://www.kff.org/insurance/7315/exhibits/index.cfm. Accessed April 4, 2007.
3. Thomas CP. Incentive-based formularies. N Engl J Med. 2003;349:2186-2188.
5. Joyce GF, Escarce JJ, Solomon MD, Goldman DP. Employer drug benefit plans and spending on prescription drugs. JAMA. 2002;288:1733-1739.
7. Fairman KA, Motheral BR, Henderson RR. Retrospective, long-term follow-up study of the effect of a three-tier prescription drug copayment system on pharmaceutical and other medical utilization and costs. Clin Ther. 2003;25:3147-3161.
9. Huskamp HA, Deverka PA, Epstein AM, et al.The effect of incentivebased formularies on prescription drug utilization and spending. N Engl J Med. 2003;349:2224-2232.
11. Goldman DP, Joyce GF, Escarce JJ, et al. Pharmacy benefits and the use of drugs by the chronically ill. JAMA. 2004;291:2344-2350.
13. Landsman PB,Yu W, Liu X,Teutsch SM, Berger ML. Impact of 3-tiered pharmacy benefit design and increased consumer cost sharing on drug utilization. Am J Manag Care. 2005;11:621-628.
15. Nair KV, Wolfe P, Zaluck RJ, et al. Effects of a 3-tiered pharmacy benefit design on the prescription purchasing behavior of individuals with chronic disease. J Manag Care Pharm. 2003;9:123-133.
17. Zhan C, Sangl J, Bierman AS, et al. Potentially inappropriate medication use in the community-dwelling elderly. JAMA. 2001;286: 2823-2829.
19. Gilman BH, Kautter J. Impact of multi-tiered copayments on the use and cost of prescription drugs among Medicare beneficiaries. Health Serv Res. In press.
21. Thomas CP, Wallack SS, Lee S, Ritter GA. Impact of health plan design and management on retireesâ€™ prescription drug use and spending, 2001. Health Aff (Millwood). 2002;suppl Web exclusives:W408-W419.
23. Neumann PJ. Evidence-based and value-based formulary guidelines. Health Aff (Millwood). 2004;23:124-134.
25. Fendrick AM, Chernew ME. Value-based insurance design: aligning incentives to bridge the di vide between quality improvement and cost containment. Am J Manag Care. 2006;12:SP5-SP10.