Health Affairs, 23, no. 1 (2004): 160-167 doi: 10.1377/hlthaff.23.1.160 © 2004 by Project HOPE	New Online   How Would Obama, McCain Cover The Uninsured?   Debating Cost Of Uninsured   Try Medicare-For-All   HA Blog Top 10

This Article Abstract Figures Only Reprint (PDF) Erratum Erratum (v23,p288) Submit a response to this article Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services E-mail this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to Citation Manager Reprints & Permissions Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (9) Citing Articles via Google Scholar Google Scholar Articles by Smith, C. Search for Related Content PubMed PubMed Citation Articles by Smith, C. Related Collections Pharmaceuticals Health Spending

Health Tracking

TRENDS

Retail Prescription Drug Spending In The National Health Accounts

Abstract

 
Recent rapid spending growth for retail drugs has largely arisen from increased use of new drugs, rather than from increasing prices of existing drugs. A sizable shift in the payment from consumers to third parties has also contributed to faster growth. Strategies such as negotiating for rebates and using tiered copayments have sought to slow spending growth but simultaneously have complicated the estimation of spending in the National Health Accounts (NHA). NHA estimates show that retail pharmaceuticals’ share of health spending is not much different than it was in 1960, although its share of gross domestic product (GDP) has tripled.

Estimates of U.S. retail drug spending are measured annually in the Centers for Medicare and Medicaid Services (CMS) National Health Accounts (NHA), which report total health spending by type of service or product and by payer. Estimates are primarily based on census data and sample surveys of private retail pharmacy sales. To accurately measure net spending by all payers, these estimates adhere to government national income accounting standards requiring that estimates be tabulated in a mutually exclusive and exhaustive fashion, which has solidified their reputation for high quality.³ Sales are consistently adjusted over time to exclude rebates received by insurers and to include noninstitutional populations not covered by household surveys.⁴ Further, these estimates constitute the official government estimates of overall prescription drug expenditures and are the starting point for cost estimates calculated by the CMS and the Congressional Budget Office (CBO) for policies such as Medicare prescription drug expansions.

Estimates of drug spending in the NHA document the growing importance of drugs in the U.S. health care system. These estimates are sometimes grouped with the more inclusive "nondurable medical products" category, but in this paper the discussion pertains only to products requiring a prescription.⁵ The interpretation of these statistics and their underlying trends can be improved by understanding the measurement concepts, data sources, and methods used to produce them.

What’s In The Estimates

Top Editor's Notes What’s In The Estimates U.S. Historical Perspective Who Pays For Drugs... Role Of Managed Care... The Baseline Rebates Out-Of-Pocket And Insurance... Notes

 
Spending estimates in the NHA’s prescription drug category report sales in retail outlets such as independent and chain pharmacies, mass-merchandiser and department stores, food stores, and mail-order pharmacies. These retail sales measure spending for human use and dosage-form prescription drugs that are sold to consumers directly and do not include drugs billed through clinics, hospitals, and long-term care institutions. However, drug spending through health care institutions is included in each provider sector’s estimate.

U.S. Historical Perspective

Top Editor's Notes What’s In The Estimates U.S. Historical Perspective Who Pays For Drugs... Role Of Managed Care... The Baseline Rebates Out-Of-Pocket And Insurance... Notes

 
Recent rapid increases in retail drug purchases, which rose at an average annual rate of 15.6 percent between 2000 and 2002, have been a top concern of policymakers (Exhibit 1). Much of the rise in spending over the past decade has been attributed to steady growth in use, which incorporates a shift in consumption to newer drugs at higher prices. A smaller portion has been accounted for by price increases among existing drugs (Exhibit 2).⁶

View larger version (23K): [in this window] [in a new window]   EXHIBIT 2 Factors Accounting For Growth In U.S. Prescription Drug Spending, 1980–2002

View larger version (12K): [in this window] [in a new window]   EXHIBIT 3 Prescription Drug Spending As A Share Of U.S. Health Spending, 1960–2002

Top Editor's Notes What’s In The Estimates U.S. Historical Perspective Who Pays For Drugs... Role Of Managed Care... The Baseline Rebates Out-Of-Pocket And Insurance... Notes

Role Of Managed Care In Insulating Consumers

Top Editor's Notes What’s In The Estimates U.S. Historical Perspective Who Pays For Drugs... Role Of Managed Care... The Baseline Rebates Out-Of-Pocket And Insurance... Notes

 
Prior to the introduction of managed care, fee-for-service (FFS) plans constrained growth in drug usage through the use of deductibles, coinsurance provisions, and ceilings on annual prescription benefits. FFS plans typically paid a share of drug expenses after an overall deductible was met.⁹ As managed care companies sought increased market share, they began to offer drug coverage for relatively low copayments, setting the stage for increased purchase and use of outpatient prescription drugs. Managed care’s focus on well care also increased physician visits and complementary purchases of prescription drugs, contributing to increased use and rising private health insurance payments. Since many beneficiaries faced relatively low out-of-pocket costs of $5 or $10 for drugs, with most participants subject only to the major medical limits of the health plan, insurance often paid for the majority of the costs.¹⁰

PBMs. More recently, managed care companies hoped to rein in total health care costs through improved management of benefits, including pharmaceuticals. The increasing trend of managing drug benefits spawned specialized companies called pharmacy benefit managers (PBMs), whose tasks evolved from solely processing claims to also monitoring patients’ refill requests, notifying physicians of alternative therapies, and providing information services to clients. Perhaps the most important function of PBMs was the volume discounts they negotiated, which complicated the estimation of private health insurance spending for prescription drugs.

Multi-tier plans. Managed care plans also introduced multi-tier plans, which encouraged the use of certain drugs by charging different copayments for brand-name and generic drugs on and off plans’ formularies. These new plans also often brought increased consumer cost sharing and other cost control methods such as generic incentive programs, prior authorization, and more intense drug utilization review. Savings in these plans arise from increased consumer cost sharing, which in some instances simply shifts more costs to consumers and in other instances could reduce use. This has narrowed the gap between private health insurance and out-of-pocket drug spending growth in recent years.

Use of generics. Generic drugs are gaining ground in the prescription drug market. Access to the market might be further hastened by a rule passed in 2002 that limits the number of automatic thirty-month periods of marketing exclusivity granted to brand-name drugs, thereby reducing delays of generic drugs.¹¹ Prescriptions dispensed in generic form increased from 19 percent of all prescriptions in 1984 to 47 percent in 2001.¹² However, the impact of generics on the share of total sales (rather than the number of prescriptions) was rather modest—only 18 percent in 2001—because of the large disparity between the average prices of brand-name and generic products. Meanwhile, incentives to consume less costly generic drugs are in place through tiered payment systems, and budgetary pressures have accelerated the likelihood that health plans will pursue these and other money-saving strategies.

Conversions to OTC status. Savings also might originate from the conversion of prescription drugs to over-the-counter (OTC) status. When one prescription drug in a therapeutic class converts to OTC status, sponsors must then decide whether to move all prescription products in that class to the highest copayment tier, exclude the entire class of drugs from coverage, or implement yet another strategy. Still, these conversions account for only a small part of the $31.7 billion in OTC products, also estimated within the NHA.

The Baseline

Top Editor's Notes What’s In The Estimates U.S. Historical Perspective Who Pays For Drugs... Role Of Managed Care... The Baseline Rebates Out-Of-Pocket And Insurance... Notes

 
Methods for estimating retail prescription drug spending have evolved over time as the quality and availability of data sources have changed.¹³ When methodologies changed, earlier estimates were recalculated to produce a consistent time series. Over all years, estimates are benchmarked to the most comprehensive source of information on retail drug sales: the Census of Retail Trade. This survey reports sales data from retailers disaggregated by merchandise type, one of which is prescription drugs. We make a few small adjustments to these prescription drug sales figures to match NHA accounting definitions.¹⁴

For the years following the last census, estimated sales for prescription drugs are then made using growth in National Prescription Audit retail sales and in wholesale mail-order sales tracked by IMS Health, adjusted for intra–health system transfers and rebates. (Some retail sales flow to institutions such as nursing homes, whose purchases are separately accounted for in the NHA, and some mail-order sales filter through the Department of Veterans Affairs [VA].)¹⁵ Since retail sales at the pharmacy do not include the rebate amount returned to health insurers and Medicaid programs, retail payments are higher than net payments by health insurers. Consequently, sales are adjusted to deduct these manufacturer rebates.

Rebates

Top Editor's Notes What’s In The Estimates U.S. Historical Perspective Who Pays For Drugs... Role Of Managed Care... The Baseline Rebates Out-Of-Pocket And Insurance... Notes

 
Rebates are payments negotiated between PBMs and manufacturers when the manufacturers’ products are given preference on a formulary, and they are typically shared between the PBM and the insurer. Although some generic drugs offer rebates, brand-name drugs offer the majority of them.¹⁶ The two most common types for brand-name drugs are access rebates, which are paid merely for placement on a formulary, and market-share rebates. The lure of rebates might have steered customers to more or less costly drugs, as PBMs increasingly rely on revenues earned from rebates rather than from claims processing fees.¹⁷

The CMS internal calculation of rebates in the brand-name drug market is partly informed by annual reports of rebates paid in the Medicaid program. Manufacturers paid a minimum Medicaid rebate equal to 15.7 percent of the average price they charged retail pharmacies for brand-name drugs in 1993, 15.4 percent in 1994, 15.2 percent in 1995, and 15.1 percent in 1996 and beyond.¹⁸ If a manufacturer gives a more substantial discount to a private purchaser, Medicaid must receive the equivalent discount (a requirement known as the "best price" provision). Manufacturers report to the CMS both the average manufacturer price (AMP), which measures the price received for a drug sold to retail pharmacies, and the lowest, best price charged for a brand-name prescription drug to any private purchaser. The percentage difference between the AMP and the best price is the best-price discount.

Using these reported data on brand-name drugs sold in the Medicaid program, the average best-price discount for drugs was weighted each year by the mix of drugs paid for by Medicaid.¹⁹ Although top-selling brand-name drugs in Medicaid overlap top-selling drugs in the private third-party market, some differences in consumption are expected. Similarities are implied by one report that seven of the top ten Medicaid drugs ranked by per member per year spending were also among the top ten drugs in spending within the commercial/group managed care market.²⁰ In the absence of actual private rebate data, we relied on the weighted average Medicaid best-price discount for determining annual rebates in the private market. We then estimated that, on average, private purchasers earned about half the percentage rebate that Medicaid earned, because not all private insurers can negotiate the largest reported discount or qualify for rebates on every drug.²¹ In consultation with industry experts, we estimated that approximately 80 percent of all rebates are passed back to insurers, lowering their benefit payments. The resulting estimate for private third-party rebates as a percentage of third-party retail drug spending was 8.5 percent in 1993, falling to 7.5 percent by 1995, then rising to 8.1 percent by 2002. When aggregating rebates earned in the private third-party and Medicaid markets, we estimated total rebates as a percentage of third-party expenditures (including all other public support in the denominator) at 9.3 percent in 2002, nearly unchanged from 9.0 percent in 1995.

Out-Of-Pocket And Insurance Payments

Top Editor's Notes What’s In The Estimates U.S. Historical Perspective Who Pays For Drugs... Role Of Managed Care... The Baseline Rebates Out-Of-Pocket And Insurance... Notes

 
A remaining issue concerns the manner in which private payments are distributed among consumers, private health insurers, and public programs. In the NHA, public expenditures for prescription drugs are estimated from government records. To assist in the allocation of private funding, the CMS looks to the IMS Health National Prescription Audit Retail Method of Payment Report, which shows the value of prescriptions paid in retail pharmacies by consumers, insurers, or governments. This report captures information based on the primary payer for the drug. The consumer category includes direct consumer purchases, which might later be reimbursed by traditional indemnity providers; purchases of non-covered prescriptions by insured people; and purchases by uninsured customers. Insurance, or third-party coverage, includes all third-party transactions other than public payments.

If a person presents insurance coverage when purchasing drugs, the entire payment is attributed by IMS to insurance even though a copayment is often required at the pharmacy. Therefore, other deductions from insurance are also necessary in addition to the deduction of insurance for estimated rebates: (1) a copayment share that is transferred to "out of pocket"; and (2) Medicare and Medicaid health maintenance organization (HMO) expenditures, since these HMO premiums are categorized as government programs in the NHA.

In determining the copayment share, the CMS used the out-of-pocket share for prescription drug purchases with private insurance payments reported in the person-based surveys of the National Medical Expenditure Survey (NMES) of 1987 and the more recent Medical Expenditure Panel Survey (MEPS) through 1997. Since the data are released by MEPS with about a two-year lag, Scott-Levin (now Verispan) data are used in concert with industry-reported statistics to determine the most recent growth in copayments for 1998 through 2002. Based on the increasing importance of three-tier copayment structures, a weighted-average copayment share was developed based on information from the Scott-Levin Managed Care Formulary Audit.²² To form this share, the average copayment shares incurred by two- and three-tier plans were weighted by the portion of covered lives in these plans.²³ Enrollment shifting into three-tier plans contributes to a steady rise in copayment shares, as these shares are four to five percentage points higher on average than copayments in two-tier plans.

Following these changes, the adjusted IMS out-of-pocket and private health insurance data are used to allocate NHA private prescription drug spending. Finally, growth in out-of-pocket prescription drug spending is compared with similar information from the Bureau of Labor Statistics (BLS) Consumer Expenditure Survey.

Following rapid growth in prescription drug spending since the mid-1990s—caused in part by rapid growth in the number of new drugs—signs of slowing have recently appeared. Patent expirations of blockbuster drugs, for example, and the use of tiered copayment benefits have grown, two factors that have already begun to slow increases in consumption.²⁴ Yet growth in prescription drug spending is projected to continue to outpace overall health spending by nearly four percentage points over the next ten years, causing its share of health spending to increase from 10.5 percent in 2002 to 14.5 percent in 2012—even before anticipated growth from Medicare drug legislation is taken into account. Perhaps more importantly, the rate of spending growth is expected to continue to outstrip gains in GDP. This situation will intensify interest in research to evaluate whether purchases of specific drugs raise or lower overall health spending.

Editor's Notes

Top Editor's Notes What’s In The Estimates U.S. Historical Perspective Who Pays For Drugs... Role Of Managed Care... The Baseline Rebates Out-Of-Pocket And Insurance... Notes

 
Cynthia Smith is an economist with the National Health Statistics Group, Office of the Actuary, at the Centers for Medicare and Medicaid Services in Baltimore, Maryland.

The author thanks Katie Levit, Mark Freeland, and several anonymous peer reviewers for their helpful comments. The opinions expressed here are the author’s and do not necessarily represent those of the CMS.

Notes

Top Editor's Notes What’s In The Estimates U.S. Historical Perspective Who Pays For Drugs... Role Of Managed Care... The Baseline Rebates Out-Of-Pocket And Insurance... Notes

 

1. S. Heffleret al., "Health Spending Projections for 2002–2012, 7 February 2003, www.healthaffairs.org/WebExclusives/Heffler_Web_Excl_020703.htm (6 June 2003). Projections are based on an econometric model that uses disposable personal income, relative prescription drug prices, new drug introductions, and direct-to-consumer advertising as independent variables. The model results are then adjusted for factors not easily picked up in this type of model, such as the effect of a blockbuster drug losing patent protection.
2. Worldwide sales data are from IMS World Review, and U.S. sales are from IMS Retail and Provider Perspective. Dividing the U.S. sales by the worldwide sales reveals that more than three-quarters of the top ten drug sales are consumed by the United States. IMS Health, "Top-Line Industry Data," www.imshealth.com/ims/portal/front/indexC/0,2773,6599_5264_0,00.html (26 June 2003). Three of the top ten drugs consumed in 2001 were among the top ten in direct-to-consumer advertising in 2001.National Association of Chain Drug Stores, Chain Pharmacy Industry Profile (Washington: NACDS, 2003), 50. Retail prices of the top drugs are available from National Institute for Health Care Management Research and Educational Foundation, "Prescription Drug Expenditures in 2001: Another Year of Escalating Costs" (Washington: NIHCM, 2003), 13.
3. P.B. Ginsburgand J.D. Pickreign, "Tracking Health Care Costs," Health Affairs (Fall 1996): 140–149.
4. Household surveys typically cover the noninstitutionalized population only; are subject to recall error by respondents, which can lead to undercounting; and do not reflect rebate transactions between manufacturers and insurers. Instead, the strength of household surveys comes from the microdata available on individuals, allowing researchers to study interactions among demographic and health factors, income, and health care purchases.
5. Estimates of prescription drug and the more inclusive nondurable medical product category of spending are available annually. Other nondurable medical products include nonprescription drugs and medical sundries. Nonprescription drugs sold over the counter include those marketed to the general public and those promoted to the medical professions and comprise products such as vitamins, analgesics, and cough and allergy medications. Finally, medical sundries primarily include such items as sanitary napkins and tampons, surgical and medical instruments, surgical dressings, and diagnostic products such as needles and thermometers.
6. Calculations by the Centers for Medicare and Medicaid Services. Price growth is measured by the Consumer Price Index for Drugs, which measures the composite price growth in the same product from year to year. Price growth in the introduction of new drugs is not captured until its second year of existence.
7. Pharmaceutical Research and Manufacturers of America, Pharmaceutical Industry Profile 2000 (Washington: PhRMA, March 2000), 32–36.
8. J.D. Kleinke, "The Price of Progress: Prescription Drugs in the Health Care Market," Health Affairs (Sep/Oct 2001): 43–60.See also F.R. Lichtenberg, "Are the Benefits of Newer Drugs Worth Their Cost? Evidence from the 1996 MEPS," Health Affairs (Sep/Oct 2001): 241–251.
9. As late as 1989, 74 percent of medical plan participants among full-time employees of midsize and large private establishments were subject to at least one of three options: a ceiling on benefits offered, coinsurance, or a deductible before reimbursement was offered.C. Bakerand N. Kramer, "Employer-Sponsored Prescription Drug Benefits," in Employee Benefits Survey: A BLS Reader, Bulletin 2459 (Washington: U.S. Government Printing Office, February 1995), 127–131.
10. Bureau of Labor Statistics, Employee Benefits in Medium and Large Establishments, 1995, Bulletin 2496 (Washington: GPO, April 1998), 70. Tables 73 and 74 indicate that about half of the participants in non-HMO plans paid small copayments (less than $10) or paid coinsurance, and that a majority of those covered by HMOs paid copayments, nearly half less than $5.
11. U.S. Food and Drug Administration, "FDA’s New Regulation to Speed Access to Lower Cost Generic Drugs about to Take Effect," Press Release, www.fda.gov/bbs/topics/NEWS/2003/NEW00932.html (23 October 2003).
12. CMS, Health Care Industry Market Update: Pharmaceuticals, 10 January 2003, www.cms.hhs.gov/reports/hcimu/hcimu_01102003.pdf (23 October 2003).
13. J.S. Genuardi, J.M. Stiller, and G.R. Trapnell, "Changing Prescription Drug Sector: New Expenditure Methodologies," Health Care Financing Review (Spring 1996): 191–204.
14. For example, sales taxes imposed on prescription drugs in Illinois and New Mexico are added, as are prescription sales for the small number of retail outlets that have no employees and are not covered by the census. Additionally, retail drug purchases by nursing homes are deducted from total retail sales to avoid double-counting expenditures under nursing home services.
15. In 2002, $3 billion in purchases by nursing home pharmacy sales were transferred to the nursing home estimate.
16. Because of generics’ much smaller share of the market, rebates for generic drugs have thus far not been incorporated because of their small amounts.
17. B. Martinez, "Pharmacy-Benefit Managers at Times Toil for Drug Firms," Wall Street Journal, 14 August 2002. Competition has driven down what PBMs can charge health plans and employers for processing claims, from a per claim charge of roughly $1 a decade ago to just 20–30 cents today.
18. Office of the Assistant Secretary for Planning and Evaluation, Department of Health and Human Services, Memorandum from Anna Cook, "Analysis of Medicaid Rebates and Best Price Discounts, 1994 to 2000," 31 January 2001.
19. Ibid., Table 1.
20. Novartis Pharmacy Benefit Report, Facts and Figures, 2000 Edition (Totowa, N.J.: Emron, 2001), 18.
21. We applied approximately half of the best-price discount percentage in Medicaid to private third-party pharmacy sales after subtracting retail gross profits.
22. This is a semiannual study of the formularies and related prescribing controls at the country’s largest managed care organizations. Data for this audit are collected on second- and third-tier copayment amounts, formulary inclusion, preferred status, prior authorization, and other variables.
23. As three-tier plans grow more prevalent, an increasing number of surveys yield information on copayments and use. Among those surveys reporting the data are the Henry J. Kaiser Family Foundation/Health Research and Educational Trust (HRET) employer surveys (available since 2000), the Atlantic Information Services survey of PBMs (since 2000), Scott-Levin’s Managed Care Formulary Audit (since 1998), and the Novartis Pharmacy Benefit Report (since the mid-1990s). Average copayments reported by the surveys have tended to be very close to one another, especially in the first and second tiers.
24. E. Long, "Slower, but Strong, Rx Growth Predicted over Next Five Years," Drug Topics (3 March 2003): 69.

                      What's this?

This article has been cited by other articles:

				S. Morgan, K. Bassett, and B. Mintzes Outcomes-Based Drug Coverage in British Columbia Health Aff., May 1, 2004; 23(3): 269 - 276. [Abstract] [Full Text] [PDF]

				J. K. Iglehart The New Medicare Prescription-Drug Benefit -- A Pure Power Play N. Engl. J. Med., February 19, 2004; 350(8): 826 - 833. [Full Text] [PDF]