5

How Does the VA National Formulary Compare with Private Insurance Formularies for Drugs and Devices and with Other Government Formularies?

INTRODUCTION

The majority of Americans are covered by private-sector pharmacy benefits with formularies and formulary systems. These benefits must be flexible and responsive to public needs and preferences if they are to compete in the private marketplace. The VA has explicitly attempted to model its recent reorganization and pharmacy benefits after many aspects of those in the private sector (Kizer, 1996, 1999; Ogden et al., 1997). The committee found that comparisons of the VA National Formulary and formulary systems with private-sector plans are the most informative. There are clear and substantive differences between the VHA and MCOs (as described elsewhere in this report), which suggest the need for caution in making comparisons. Nevertheless, the VA and private-sector plans have similar formulary control objectives, as noted in Chapter 2 of this report. The committee also concluded that specific details of MCO formularies and state statutory controls on MCO performance would be helpful in understanding MCO formularies and formulary systems in comparison to the VA National Formulary.

That information is reviewed in this chapter. The information to characterize these formularies was gathered from the open, peer-reviewed literature, company websites, and the annual surveys supported by Hoechst Marion Roussel, Novartis, and Wyeth-Ayerst. The IOM committee did not find any surveys of MCO and PBM formularies that were similar to the survey of pharmacists about hospital P&T committee activities done by Mannebach et al. (1999). To address this deficiency, a special survey responding to questions concerning important controls or elements of restrictiveness that the committee identified was carried out by the Academy of Managed Care Pharmacy, as noted earlier



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 149
DESCRIPTION AND ANALYSIS OF THE VA National Formulary 5 How Does the VA National Formulary Compare with Private Insurance Formularies for Drugs and Devices and with Other Government Formularies? INTRODUCTION The majority of Americans are covered by private-sector pharmacy benefits with formularies and formulary systems. These benefits must be flexible and responsive to public needs and preferences if they are to compete in the private marketplace. The VA has explicitly attempted to model its recent reorganization and pharmacy benefits after many aspects of those in the private sector (Kizer, 1996, 1999; Ogden et al., 1997). The committee found that comparisons of the VA National Formulary and formulary systems with private-sector plans are the most informative. There are clear and substantive differences between the VHA and MCOs (as described elsewhere in this report), which suggest the need for caution in making comparisons. Nevertheless, the VA and private-sector plans have similar formulary control objectives, as noted in Chapter 2 of this report. The committee also concluded that specific details of MCO formularies and state statutory controls on MCO performance would be helpful in understanding MCO formularies and formulary systems in comparison to the VA National Formulary. That information is reviewed in this chapter. The information to characterize these formularies was gathered from the open, peer-reviewed literature, company websites, and the annual surveys supported by Hoechst Marion Roussel, Novartis, and Wyeth-Ayerst. The IOM committee did not find any surveys of MCO and PBM formularies that were similar to the survey of pharmacists about hospital P&T committee activities done by Mannebach et al. (1999). To address this deficiency, a special survey responding to questions concerning important controls or elements of restrictiveness that the committee identified was carried out by the Academy of Managed Care Pharmacy, as noted earlier

OCR for page 149
DESCRIPTION AND ANALYSIS OF THE VA National Formulary (see Appendix B and Table 2.1). This survey covered two small MCOs and six PBMs serving 176 million covered lives. The committee also relied on information from the personal and institutional experiences of committee members in the managed care and pharmacy benefits management industry (E. Dichter; J. Jones; O. Wolke; and A. Zimmerman, personal communications, 2000). MCOs or PBMs may know what effect prior authorization, therapeutic interchange or generic substitution, or copayments and other cost controls that are commonly used have on choice of drugs for their patients. The committee found no national quantitative quality or cost data in the peer-reviewed literature, however, and therefore could not compare these restrictions to the VA National Formulary effects on utilization (examples of which are displayed * in the figures in Chapter 3 and Chapter 4 of this report and discussed in both chapters). Clearly, these controls are intended to restrict choice and direct prescribing. It is well known that copayment size is inversely related to utilization of health care services (Brook et al., 1983), including prescription drugs (Leibowitz et al., 1985; Smith, 1993), and may possibly affect health outcomes (Johnson et al., 1997). Most MCOs/HMOs offer prescription drug benefits (98.1%, although only about 90 to 92% of enrollees buy these benefits). They usually have formularies (92.9%, but 97.8% of all HMO enrollees are covered by plans that have formularies), and 92% of HMOs contract with PBMs to handle part or all of their pharmacy programs. This most often involves claims processing (Hoechst Marion Roussel, 1998), but PBMs carry out formulary management for 46% of MCOs and 63.2% of employer plans according to Novartis (1998). Most formularies are closed (26.9%) or partially closed (45.4%) (Novartis, 1999). Similar results—35% closed, 24% partially closed—were reported by Luce et al. (1996). Kreling and Mucha (1992) reported 60% restricted or restricted with exceptions. These figures document the increase in restrictive formularies in managed care (72.2%); most hospital formularies have been restrictive for some time. Only 17.5% of MCOs provide brand name drugs without a penalty when generics are available, which occasionally (17%) amounts to the entire cost of the brand drug, but usually (66.4% [or 44.4%; Novartis, 1998]) to the difference in cost (Hoechst Marion Roussel, 1998). Employer plans are more generous, but 20% require dispensing generics when available; otherwise the enrollee must pay either the entire prescription cost or, more often, the difference in cost between the brand and the generic drug (Wyeth-Ayerst, 1998). MCO and PBM coverage of members of major drug classes is usually more extensive than the VA National Formulary, but financial penalties for nonformulary drugs, non-preferred drugs, or brand name drugs when generics are available are more and more common. This cost control was applied by 86.4% of HMO pharmacy benefit programs estimated in 1999, but less frequently by PBMs (Lipton et al., 1999). MCO copayments on formulary genetic prescriptions average $6.17, and on formulary brands $9.65. Nonformulary genetic copayments average $7.32, and nonformulary brands $13.77. About 80 to 90% of MCOs require prior authorization for some drugs (Novartis, 1998, 1999).

OCR for page 149
DESCRIPTION AND ANALYSIS OF THE VA National Formulary IMPLEMENTATION OF DRUG MANAGEMENT STRATEGIES IN MANAGED CARE MCO drug management strategies include formularies and formulary systems, generic substitution, therapeutic interchange, tiered copayments, DUR, and prior approval, among others. These, as in the case of the VA, are intended to direct prescribers and patients to lower-cost, but similarly safe and effective drugs and to help in price negotiations to obtain such lower-cost drugs. As described later, some of these strategies are applicable to Medicaid managed care recipients within certain limits. Others, such as tiered copayments, are not. MCO formularies also restrict access by prior approval or lists of excluded drugs, ranging from a few to more than 200 different agents. Earlier in this chapter, MCO formulary systems were discussed. Here, some specific formularies are examined. The committee reviewed MCO or PBM formularies in whole or in part from six Mountain State MCOs (R. Valuck, personal communication, 1999), five major Massachusetts MCOs (Massachusetts Outpatient Formulary Guide, 1999), Geisinger Health Plan (PennState Geisinger Health Plan, Formulary, 1999), PCS Health Systems (1999), Humana (http://www.humana.com, under member services), United (www.uhc.com, under pharmacy programs), and Aetna/U.S.Health Care (www.aetnaushc.com, under members and consumers) and from the experiences of committee members. About 70 million covered lives and hundreds of formularies are represented by MCO or PBM officials serving on the IOM committee. Because managed care formularies undergo constant revision, these formularies will undoubtedly be different by the time this report is published. MCOs usually list more drugs than the VA National Formulary, but they also exclude more drugs, have more drugs on required prior authorization, and occasionally have quantity or volume limits. They may also list only one agent in a class, which may be different or the same as that listed in the VA National Formulary (although, according to the AMCP survey [see Table 2.1 ], the surveyed PBMs and MCOs usually do not limit closed classes to only one agent). For example, in Massachusetts, Fallon Community Health Plan listed lansoprazole, Neighborhood Health Plan, omeprazole; and BlueCrossBlueShield, Harvard Pilgrim Health Care, and Tufts Health Plan, both lansoprazole and omeprazole in the very popular and costly PPI class (Massachusetts Formulary Guide, 1999). At the same time, some of these plans listed five or six ACEIs or HMG CoA RIs compared to two or three in these classes in the VA National Formulary. Tufts Health Plan publishes its list of noncovered drugs as part of its Prescription Alternative Program (www.tufts-healthplan.com, under member information). The list of noncovered drugs includes both the restricted product and the health plan's suggested alternative(s). For example, the HMG CoA RIs cerivastatin (Baycol), lovastatin (Mevacor), and simvastatin (Zocor) are on the noncovered list. The suggested alternatives are atorvastatin (Lipitor), pravas

OCR for page 149
DESCRIPTION AND ANALYSIS OF THE VA National Formulary tatin (Pravachol), and fluvastatin (Lescol). A modest number of other drugs have quantity limits or require prior authorization. Humana's on-line drug formulary notes 12 products and some injectables requiring prior approval. These products include the ache drug isotretinoin (Accutane), alendronate sodium (Fosamax) for osteoporosis, antifungals terbinafine tabs (Lamisil) and itraconazole (Sporanox), finasteride (Proscar) for prostatic hypertrophy, troglitazone (Rezulin) for diabetes, and NSAIDs celecoxib (Celebrex) and rofecoxib (Vioxx). United Health Care's on-line list of preferred drugs includes 19 products that require prior approval and 69 that have some type of quantity limit. The prior approval list includes antihypertensives losartan (Cozaar) and losartan/hctz (Hyzaar), etanercept (Enbrel) for arthritis, antifungals fluconazole (Diflucan), terbinafine (lamasil), and itraconazole (Sporanox); and the antidepressant bupropion (Wellbutrin SR). Drugs subject to some sort of prescribing limit include: interferon beta 1B (Betaseron); alprostadil (Caverject) for erectile dysfunction; SSRIs paroxetine (Paxil), citalopram (Celexa), fluoxetine (Prozac), and sertraline (Zoloft); fluticasone propionate (Flonase) for allergy; alendronate sodium (Fosamax); drugs for migraine sumatriptan succinate (Imitrex), rizatriptan benzoate (Maxalt), and zolmitriptan (Zomig); and HMG CoA RIs simvastatin (Zocor) and atorvastatin (Lipitor). Aetna US Health Care lists 46 drug products that require precertification and more than 120 that are excluded from the formulary. Of the newer antidepressants, Aetna excludes payment for fluvoxamine (Luvox) and citalopram (Celexa) and lists paroxetine (Paxil), fluoxetine (Prozac) and sertraline (Zoloft) as alternatives. For HMG CoA RIs, Aetna excludes cerivastatin (Baycol), atorvastatin (Lipitor), pravastatin (Pravachol), lovastatin (Mevacor) and includes simvastatin (Zocor) and fluvastatin (Lescol) as alternatives. A number of drugs such as rofecoxib (Vioxx) and celecoxib (Celebrex) require step therapy under the Aetna formulary plan. A comprehensive analysis of MCO formularies is not available. The restrictiveness data assembled for this report from the AMCP survey and from publicly available sources are the most extensive and up-to-date currently available. MCOs may have different formularies, prior approval, and copayment provisions for different clients such as employers, individuals, Medicare, and Medicaid programs. Some insurers have numerous health plans, each with a unique drug benefit structure. For example, the Aetna US Healthcare on-line formulary notes exceptions for California (injectable drugs require precertification) and Indiana (no precertification program) residents. The Humana on-line formulary lists unique formularies for Tampa, Florida, South Florida, and Illinois. Other large MCOs are similarly variable. VA formularies are also variable, as noted elsewhere in this report. This variability among formularies in different health care sectors and the variability in controls or elements of restrictiveness prevented the committee from reaching a definitive conclusion about comparative restrictiveness. In some respects the VA National Formulary is more, and in some respects less, restrictive than comparison formularies.

OCR for page 149
DESCRIPTION AND ANALYSIS OF THE VA National Formulary IMPACT OF STATE LEGISLATION ON MCO PRACTICE Reacting to the studies cited in this report, to professional and trade groups, and to consumers, legislators in many states have become actively involved in the issues of managed care and the use of formularies. To date, 33 states (see Figure 5.1) have passed legislation authorizing the use of formularies. The majority of these states have also required public disclosure of the formulary and nonformulary process. California Law SB625 requires the filing of the nonformulary procedure and that the process be expeditious (http://www.leginfo.ca.gov/pub/97–98/bill/sen/sb_0601–0650/sb_625_bill_19980622_chaptered.html). Maine Law 63-2550.4 requires nonformulary request approvals within 24 hours or provision of a 72-hour supply of the prescribed drug. Fifteen states require access to nonformulary drugs if, (1) they are medically necessary, (2) they are prescribed by a physician, and (3) the preferred drug is ineffective or reasonably expected to cause an adverse or harmful reaction (see www.ganet.state.ga.us/cgi-bi...code/g/33/20A/9 for Georgia's law). Tennessee Bill SA0684 amends SB0637 and prohibits managed care organizations from either switching or discontinuing an enrollee's prescription drug unless the patient's provider determines that this change would not harm or prolong the patient's treatment (www.legislature.state.tn.us/Bills/100gahtm/l00_amnd/sa0684.htm). California Bill AB974 requires that health plans continue coverage of a drug that is appropriately prescribed and medically necessary (www.leginfo.ca.gov/pub/97–98/bill/asm/ab_0951–1000/ab_974/bill/19980622_chaptered.html). States have also addressed off-label usage. When scientific results reported in the FIGURE 5.1 States specifically authorizing formularies.

OCR for page 149
DESCRIPTION AND ANALYSIS OF THE VA National Formulary FIGURE 5.2 States enacting off-label usage legislation. medical literature support the off-label use of a drug for a medical condition, MCOs in 15 states must provide coverage. An additional 15 states specifically require off-label coverage for the treatment of cancer (Figure 5.2). In 1999, 12 states (Connecticut, Florida, Iowa, Maryland, New Hampshire, North Carolina, North Dakota, Oklahoma, Rhode Island, South Dakota, Texas, and Virginia) passed legislation, in some cases further restricting MCOs. Connecticut legislation prohibits a managed care plan from increasing cost sharing or copayment conditions or eliminating or decreasing covered prescriptions during the contract year (www.cga.state.ct.us/ps99/fc/pdf/1999sb-00125-r000430-fc.pdf). Similar legislation was passed in Texas and was effective as of September 1999 (http://tlo2.tlc.state.tx.us/tlo/76r/billtext/SB01030F.htm). Other provisions enacted include establishing quality-of-care and performance measurement systems (Maryland) or requiring formulary approval by a special committee (Virginia) (Centeon, 1999). Most of the provisions of these state laws are not relevant to the VA National Formulary, which is characterized by complete disclosure (www.dppm.med.va.gov), off-label coverage, low or no copayments, and relatively stable drug coverage. Nevertheless, the basic thrust of many of these provisions is to ensure that access to drugs is not restricted unless there is a reasonable decision on medical necessity, a concern for all formularies including the VA National Formulary.

OCR for page 149
DESCRIPTION AND ANALYSIS OF THE VA National Formulary PUBLIC-SECTOR PROGRAMS The VHA is a government health care program. For this reason, two relevant public-sector programs with formularies and formulary systems were reviewed in addition to MCO programs. The VA and Congress had specified such a comparison and specifically suggested the Medicaid program. The Medicaid program also is the largest government pharmacy benefit. It has been functioning with formularies for several decades. Its effects have been studied extensively, and it is described annually in some detail in surveys supported by the National Pharmaceutical Council (NPC). The committee appreciates that Medicaid serves a population with entirely different demographics (principally women and children), has different eligibility standards (primarily economic rather than national service), and is subject to quite different federal and state law and regulatory program requirements. Information on Medicaid was gathered from an extensive consultant's report (Brown, 1999), committee experience, the NPC surveys, and the open, peer-reviewed medical literature. The committee also decided to review the Department of Defense (DOD) formulary and formulary system. The DOD health system and pharmacy benefit budgets are similar in size to the VHA and VA Pharmacy budgets, and there is considerable interest in coordinating DOD and VA formularies and pharmacy benefits or in making changes that would bring the DOD pharmacy benefit design closer to that of the VA (see P.L. 105-85, P.L. 106-65, and P.L. 106-117). Although the DOD health system serves a larger total population with somewhat younger active duty personnel and more women, there are similarities because of the large numbers of retired military personnel using DOD health care. DOD and VA programs are already cooperating in purchasing for example, but as described below, the formularies are quite different. Data for the committee 's comparison of the DOD system came primarily from GAO audits and personal communications (GAO, 1998) and from the publicly available DOD website. Medicaid The Medicaid outpatient pharmacy benefit is administered through state Medical Assistance programs under Title XIX of the Social Security Act (SSA) and applicable state laws. The program provides drugs to recipients through the U.S. retail pharmacy market, which dispensed about 2.97 billion prescriptions in 1999. It spent slightly less than $12 billion in this marketplace in FY 1997 (Baugh et al., 1999; www.nacds.org/news/releases/nr_082999_projections.html; www.hcfa.gov/medicaid/mstats.htm). In comparison, the total VHA pharmacy budget for FY 1997 was $1.3 billion (GAO, 1999). Medicaid Fee for Service In this section, the committee discusses Medicaid fee for service. Medicaid managed care is taken up later in this chapter. The Medicaid program was enacted in 1965 to provide health care services to the poor and is jointly funded by the

OCR for page 149
DESCRIPTION AND ANALYSIS OF THE VA National Formulary federal government and the individual states. Originally a fee-for-service program, it is now increasingly managed care (discussed later in this chapter). Federal law and regulation require states to provide Medicaid benefits to specific categorically needy groups in order to receive federal matching funds. These groups include those eligible for Aid to Families with Dependent Children (AFDC) prior to July 16, 1996; the aged, blind, and disabled receiving Supplemental Security Income (SSI); and children, pregnant women, and elderly who meet certain income criteria. The Personal Responsibilities and Work Opportunity Reconciliation Act of 1996 (P.L. 104–193) repealed the AFDC program and replaced it with Temporary Assistance for Needy Families (TANF). Most persons covered by TANF are eligible to receive Medicaid benefits, but it is not an automatic entitlement. States may also extend Medicaid benefits to other groups that do not meet the basic federal eligibility guidelines, for example, the medically needy. Although the Medicaid program covers a wide range of eligible groups, the Medicaid population is not representative of the U.S. population. The majority of Medicaid recipients are poor women and children. Of the 41.5 million Medicaid eligibles in FY 1997, 58% were women; 45% were less than 14, 55% less than 21, and 77% less than 45 years of age (www.hcfa.gov/medicaid/mstats.htm). This compares with a VA population that averages 53 years of age and is 95% male (see VA Annual Reports). In general, 26% of Medicaid eligibles are SSI aged, blind, or disabled and more than 68% are adults or children meeting income standards, for example, TANF or AFDC eligibles. The aged Medicaid eligibles, however, obligate a disproportionate share of Medicaid expenditures, primarily due to long-term (nursing home) care, which alone consumes about 25% of the total Medicaid budget. These Medicaid recipients resemble the VA population more closely. Only 55% of Medicaid recipients remain continuously eligible for the full year. The rest enroll and disenroll throughout the year. Over 34.8 million (83%) Medicaid eligibles received one or more health care services during 1997 (www.hcfa.gov/medicaid/mstats.htm). Although VHA eligibles are to some extent stratified by income, as veterans they remain eligible for life. Only 3.3 million of a total of about 26 million veterans are enrolled in the VA health care system, and they tend to be older, disadvantaged, and minorities (VA Annual Report, 1998; Fonseca et al., 1996). Title XIX of the SSA specifies the types of services that must be provided to Medicaid recipients if states are to qualify for federal matching funds. These include inpatient and outpatient hospital care; physician services; vaccines for children; prenatal care; x-ray and laboratory services; and nursing home care for recipients over 12 years of age. Title XIX also lists a variety of optional services eligible for federal matching funds. These include diagnostic and screening services; optometrist services; eyeglasses; intermediate-care facilities for the mentally retarded; rehabilitation and physical therapy services; transportation to and from medical services; and outpatient prescription drugs. Most of these services are provided by the VHA, including some highly specialized care for spinal cord injury and blindness, among others.

OCR for page 149
DESCRIPTION AND ANALYSIS OF THE VA National Formulary The basic Medicaid program is described as a fee-for-service system, but states have wide latitude in determining payment methods and rates. Federal law requires that payment be sufficient to enlist enough providers to reasonably deliver the care required by recipients. In addition, providers must accept Medicaid payment as full payment for their services. States may require small deductibles, coinsurance, or copayments for certain recipients, but cost sharing cannot be imposed on pregnant women, children, or nursing home patients, or required for emergency or family planning services. States may apply for waivers allowing them to enroll recipients in prepaid managed health care plans. These waivers, called Section 1915 and Section 1115 waivers, allow states to experiment with different health care delivery systems. Medicaid programs operate within broad federal guidelines. As a result, state programs have developed in various ways over time. Some states have enrolled all Medicaid recipients in managed care; and others have not enrolled any. Some states have imposed strict drug utilization and payment guidelines, while some have only nominal drug cost and use control procedures. Some states have expanded eligibility to include a broader range of needy individuals. In addition, 17 states have implemented expanded drug coverage programs for elderly and/or disabled individuals who would not normally qualify for Medicaid coverage. Nine of these programs were implemented more than 10 years ago. Total Medicaid vendor payments amounted to more than $123.5 billion in FY 1997. This figure does not include payment by Medicaid of aged recipients' Medicare Part B premiums (see Glossary) or Medicaid managed care premiums, or payments to “disproportionate share” hospitals (those with large numbers of poor patients). Almost 25% of vendor payments in 1997 were associated with nursing home care, and 20% went to hospitals. Pharmaceuticals accounted for 9.7% of vendor payments. Almost two-thirds of Medicaid recipients, about 21 million individuals, used the pharmacy benefit in 1997 (www.hcfa.gov/medicaid/mstats/htm). Since Medicaid is an entitlement, its budget is open ended. If prescription drug costs increase, they will be paid without affecting other care budgets. In the VHA, a fixed appropriation requires savings in one or more parts of the budget to offset overruns in other parts. This may make VA providers and managers more sensitive to budget overruns and to the need for controls that demonstrably achieve their intended purposes. Medicaid Prescription Drug Benefit Every Medicaid program provides coverage for prescription drugs. Federal law and regulations set the basic requirements for this coverage. Each state has leeway in designing drug benefits within these federal requirements. Federal law regarding prescription drug coverage has changed over the past decade. In addition to Title XIX of the SSA, the current Medicaid prescription drug benefit is governed by provisions enacted in the Omnibus Budget Reconciliation Acts of 1990 and 1993 (OBRA 1990, 1993). Prior to the passage of OBRA 1990, states faced fewer limits on the implementation of drug management strategies, such as for

OCR for page 149
DESCRIPTION AND ANALYSIS OF THE VA National Formulary mularies, prescription limits, generic substitution, prior approval systems, refill limits, and copayments (NPC, 1989). To address rising drug expenditures in the late 1980s, Congress proposed that a national P&T committee create a national formulary and designate therapeutic interchanges. Pharmaceutical companies responded to this proposal by offering to pay rebates on drug purchases to state Medicaid programs in exchange for a statutory prohibition on restrictive state formularies. Under a compromise enacted in OBRA 1990, Congress prohibited restrictive state Medicaid formularies, allowed prior approval under certain conditions, required Medicaid programs to reimburse all new drugs for at least 6 months after FDA approval, and required manufacturer rebates based on the lesser of a discount of about 15% below AWP or the best price offered to other purchasers. After OBRA 1990 and the elimination of restrictive formularies, many states reported increases in drug expenditures, and Congress began considering a repeal of the prohibition on restrictive formularies. In addition, other drug purchasers complained that the best price language of the Medicaid rebate agreement led to an increase in the prices they paid as manufacturers tried to reduce their rebate liability (CBO, 1996). Provisions in OBRA 1993 responded to these concerns, repealing the prohibition on restrictive formularies in OBRA 1990 as well as the requirement to cover all drugs for 6 months after FDA approval. OBRA 1993 and 1990 mandated that manufacturers sign rebate agreements to qualify their products for reimbursement. Standardization of state formularies and prior approval systems is required. Medicaid now covers products of manufacturers that have signed rebate agreements (as essentially all do) and may only exclude products in accordance with specific federal regulation described below. An excluded product must be made available through prior approval, but certain statutorily designated classes of drugs may be excluded from reimbursement and the prior approval requirement at the discretion of the states. The best-price language of the rebate agreements continues to require that the best price a manufacturer gives to any other purchaser, including any cash or volume discount or rebate, is automatically given to every Medicaid program. A minimum rebate amount is now set at 15.1% below AWP (42 USC Section 1396r–8(c)(1)(C)). Prices charged under the Federal Supply Schedule, depot prices, and single-award contract prices are excluded from the best-price calculations. Also specifically excluded are covered (that is, brand name) drug prices charged to the Indian Health Service, the Department of Veterans Affairs, the Department of Defense, and the Public Health Service, according to the Veterans Health Care Act of 1992 (42 USC Section 1396r-8(c)(1)(C)). The current Medicaid prescription drug benefit also allows prior approval of any drug as long as the prior approval system provides a response within 24 hours and a 72-hour emergency supply of the drug under review. Requirements for formularies are more complex. States may create a formulary if it is developed by a committee consisting of physicians, pharmacists, and other appropriate individuals appointed by the governor of the state, and it includes the drugs of any manufacturer that has entered into a rebate agreement unless the product is excluded in accordance with other regulatory requirements.

OCR for page 149
DESCRIPTION AND ANALYSIS OF THE VA National Formulary These latter requirements specify that a covered outpatient drug may be excluded only if it does not have a significant, clinically meaningful therapeutic advantage in terms of safety, effectiveness, or clinical outcome over other drugs included in the formulary and there is a written explanation available to the public of the basis for the exclusion (42 USC Section 1396r–8(d)(4)(C)). In any event, excluded drugs must be made available under a prior approval program unless they are members of a specifically listed class. These are (a) agents when used for anorexia, weight loss, or weight gain; (b) agents when used to promote fertility; (c) agents when used for cosmetic purposes or hair growth; (d) agents when used for the symptomatic relief of cough and colds; (e) agents when used to promote smoking cessation; (f) prescription vitamins and mineral products, except prenatal vitamins and fluoride preparations; (g) nonprescription drugs; (h) covered outpatient drugs for which the manufacturer seeks to require as a condition of sale that associated tests or monitoring services be purchased exclusively from the manufacturer or its designee; (i) barbiturates; (j) benzodiazepines. These federal statutory and regulatory requirements have left prior approval as one of the last flexible drug management strategies available to state Medicaid programs. After meeting the requirement for a 24-hour response and a 72-hour emergency supply, states have significant leeway in designing and implementing prior approval. States may adopt their own clinical or nonclinical criteria for approving a request. States may restrict a particular drug to patients of a certain age, to those with a specific and verified diagnosis, or even to those who have been treated with other drugs prior to the request. Prior approval and formulary programs vary depending on factors such as the specific drugs or classes restricted, the criteria for approval, and the system for granting exclusions to the regulations. No two Medicaid drug benefit programs are exactly alike, but no state has the flexibility in designing formularies and formulary systems, including exceptions processes, enjoyed by the VA. Medicaid Controls Prior Approval and Formulary Systems Formularies, prior approval systems, copayments, exclusions, and prescription limits or quantity controls are the most common restrictions found in Medicaid programs. Table 5.1 outlines the basic control elements of each state's Medicaid fee-for-service drug benefit. Seven states (California, Colorado, Illinois, Michigan, Montana, Ohio, and South Dakota) report a closed formulary, although all states restrict some medications or therapeutic classes. Every state restricts some or all of the OBRA 1990 excludable drugs (see Table 5.1), most commonly amphetamines, barbiturates, antihistamines, drugs used for cosmetic purposes, and benzodiazepines. Although not required, some states allow coverage of these products with prior approval. States also require prior approval for drugs other than those listed in OBRA 1990, although these requirements change periodically. These restricted drugs and drug classes include the antipsychotic drug clozapine (Clozaril), growth hormones (for example, Protropin),

OCR for page 149
DESCRIPTION AND ANALYSIS OF THE VA National Formulary The investigators surveyed assessments by panels of physicians of the therapeutic importance of expanded availability of a selected subset of 18 of the top 200 drugs. The surveyed physician panels agreed that expanded access to four of the 18 drugs resulted in a net therapeutic benefit. They agreed that expanded access to four other drugs provided no additional therapeutic benefit. The remaining products provided questionable therapeutic benefit or produced no agreement among the panels. Medicaid programs around the country reported increases in drug expenditures after the elimination of restrictive formularies. These authors warned that formularies are complex policy instruments that may lead to intended and unintended consequences (Walser et al., 1996). Dranove (1989) used Illinois Medicaid data to study the effect of lifting formulary restrictions on a group of new anti-infective drugs on office and out-patient hospital physician visits and costs for eight disease categories (all infections). He found significant increases in drug costs for three of the eight disease categories (genital tract infection, severe acne, and chronic lung disease). There was also an increase in the number of physician visits for patients with respiratory infections. Interestingly, physicians tended to prescribe the newer and more expensive drugs only after trying the older and cheaper ones. The study did not address the possibility that new drugs may provide patients with improved health more quickly, offsetting their increased cost. Dranove concluded that neither an open nor a closed formulary is best. The study did not control for previous trends in drug utilization and cost and had no comparison state. Effects of Medicaid Prior Approval Systems Two recent studies have focused specifically on the effect of prior approval for NSAIDs (Kotzan et al., 1993; Smalley et al., 1995). Both studies reported savings from increased use of generic NSAIDs as a replacement for proprietary, single-source drugs. Kotzan et al. analyzed 19 months of utilization and cost data (12 months before and 7 months after beginning prior approval) for 80,000 continuously enrolled patients in the Georgia Medicaid system. They found an immediate decrease in the number of single-source NSAID prescriptions and an increase in the use of multiple-source drugs. The increase in multiple-source prescriptions replaced slightly more than half of the single-source prescriptions. This absolute reduction in prescribing accounted for a total cost savings for NSAID therapy of $3 million. There was no increase in the use or cost of physician or hospital services during the 7 months after the program began. The results of the study are limited by the lack of a comparison group and the short postintervention period. The findings of Smalley et al. (1995) were similar. They studied the effect of the Tennessee prior approval program for NSAIDs, initiated in October 1989, on the use and cost of pharmacotherapy, outpatient, and inpatient services. There was no comparison state to control for general trends in the use of study drugs. They found that expenditures fell by 53%, resulting in a savings of approximately $12.8 million, due to increased use of generic NSAIDs and an overall

OCR for page 149
DESCRIPTION AND ANALYSIS OF THE VA National Formulary reduction of 26% in the total number of days of NSAID use. Regular users of single-source NSAIDs experienced a 28% decrease in the number of days of NSAIDs use and an associated 64% decrease in NSAID cost. There was no change in the use of other health care services associated with the prior approval program. Both measures began to rise within 3 months after the prior approval program began and continued to increase throughout the 2-year follow-up period. The authors concluded that prior approval for NSAIDs may reduce NSAID costs. The savings may be a function of the large number of generic substitutes, similar efficacy and safety profiles of most NSAIDs, and the large variations in cost among the various drugs. The claim that NSAIDs have similar safety profiles was questioned in a subsequent letter to the editor (Lehman, 1995). The writer noted that two of the generic preparations in the study by Smalley et al. (1995)—fenoprofen and phenylbutazone—“have higher rates of severe organ toxicity than other NSAIDs. ” In a reply to the letter, the authors acknowledged that use of one of the two drugs noted (fenoprofen) doubled after the implementation of the prior approval program. Concern that prior approval restricts access to new and effective therapies and may result in the substitution of less safe alternatives has been expressed by academic investigators (Soumerai and Lipton, 1995) and the public at large (Pear, 1993; The Pink Sheet, 1993; PMA Newsletter, 1993b). Bloom and Jacobs (1985) studied the effect of restricting the availability of cimetidine (Tagamet) on Medicaid expenditures. They found that prior approval led to a 79% reduction in per-patient per-month Medicaid pharmaceutical expenditures for peptic ulcer disease and an 85% reduction in cimetidine use. However, the average per-patient per-month expenditures for physician and in-patient hospital services increased 3 and 24%, respectively, during the same period, and total expenditures rose in every health service category except out-patient pharmaceuticals. There was also a significant increase in the number of surgical procedures for patients with newly diagnosed peptic ulcer disease. Although the study had a number of methodological flaws (for example, it lacked a comparison group and had a short follow-up period), the results support the conclusion that denying access to medically needed, effective drugs is shortsighted policy. At the time of the study, cimetidine was the only member on the market of a drug class (H2R blockers) that is highly effective for peptic ulcer disease. Effects of Medicaid Prescription Limits Several studies have focused on the effect of prescription limits, or quantity controls, on utilization and expenditures in Medicaid (Martin and McMillan, 1996; Soumerai et al., 1987, 1991). In general, these studies have found that prescription limits lead to an immediate reduction in pharmaceutical utilization, but not necessarily to a reduction in overall Medicaid spending, and they may adversely affect clinical outcomes. A study by Soumerai et al. (1987) was the first in a series of studies on the consequences of a three-prescription-per-month limit for New Hampshire Medicaid recipients in 1981. This policy was replaced

OCR for page 149
DESCRIPTION AND ANALYSIS OF THE VA National Formulary 11 months later by a $1.00 copayment for all prescriptions. The investigators used a time-series, comparison-series analysis of patient-level changes in the number of prescriptions dispensed (controlling for prescription size) and expenditures. New Jersey was used as the comparison state. The time series covered 48 months: 20 months before the prescription limit, 11 months during the limit, and 17 months after the limit was replaced by the $1.00 copayment. The study cohorts consisted of Medicaid recipients who were continuously enrolled for 10 or more months during each of the 4 study years. More than 10,000 patients in New Hampshire and 74,000 patients in New Jersey fit the study criteria. These investigators found that the prescription limit reduced prescriptions from 1.1 to 0.7 per patient per month. Analysis of a cohort of primarily disabled or elderly, female, multiple-drug recipients (n = 860) found that their utilization fell from 5.2 to 2.8 prescriptions per patient per month. Although the report noted that the greatest reduction in use was for ineffective drugs, there were large reductions also for essential medications such as insulin (28%) and furosemide (30%). When New Hampshire replaced the prescription limit with a copayment, prescriptions increased to just below the prelimit level (Soumerai et al. 1987). A second study by Soumerai et al. (1991) extended their 1987 findings. This study focused on the effect of the prescription limit on admissions to nursing homes and hospitals for a group of high-risk patients. The study (n = 411) and comparison (n = 1,375) cohorts consisted of elderly, multiple-drug users (more than three medications per month) taking at least one medication for certain chronic diseases during the baseline year. The authors reported that the prescription limit reduced drug use by 35%, but it was associated with a significant risk of being admitted to a nursing home (relative risk 1.8; 95% confidence interval, 1.2–2.6) compared to similar Medicaid recipients in New Jersey. Although the use of the study medications returned to near baseline after the prescription limit was replaced with a copayment, the patients admitted to nursing homes remained there. The authors concluded that limiting reimbursement for effective drugs puts patients at increased risk of institutionalization and may increase Medicaid costs (Soumerai et al., 1991). A third report by Soumerai et al. (1994) studied the effect of the New Hampshire prescription limit on the use of psychotropic agents and acute mental health services for schizophrenics. They found that the use of antipsychotic drugs, anxiolytic and hypnotic agents, and antidepressants and lithium fell 15, 37, and 49%, respectively. Per-patient per-month visits to community mental health centers increased, as did the use of emergency mental health services. The reduction in drug expenditures was more than offset by an increase in cost for other mental health services, estimated at $1,530 per patient or about 17 times the saving in drug costs. Health care utilization returned to baseline levels after the policy was discontinued. The findings of a study by Martin and McMillan (1996) on a prescription limit regulation in Georgia are consistent with previous findings of Soumerai et al. (1987, 1991, 1994). In November 1991, the Georgia Department of Medical

OCR for page 149
DESCRIPTION AND ANALYSIS OF THE VA National Formulary Assistance reduced the number of monthly reimbursable prescriptions from six to five. These two investigators conducted interrupted time-series analyses on 12 months (6 months before and 6 months after the reduction) of patient-level data for 743 high prescription drug users. There was no comparison cohort. Overall, Martin and McMillan found a 6.6% reduction in total prescriptions, a 9.9% reduction in prescriptions reimbursed by Medicaid, and a 9.7% increase in prescriptions paid out of pocket. Each of these changes was abrupt and continuous except for the out of pocket, which was temporary. Miscellaneous, palliative, and pulmonary drugs experienced a significant, abrupt, and sustained decrease in the number of prescriptions filled. It is likely that the policy reduced the use of cardiovascular drugs, but the effect was not immediate. Chemotherapy, central nervous system, gastrointestinal, and hormone prescriptions did not change. These authors concluded that prescription limits alter prescription regimens, potentially predisposing elderly Medicaid recipients to clinical consequences. The lack of a comparison cohort and the short follow-up period limit interpretation of these results. The impression left by many of these studies, despite their methodological flaws, is that denying patients access to medically needed drugs is neither good medicine nor good economics. These denials occur when formularies exclude or deny reimbursement for drugs or classes of drugs or limit prescriptions without providing exceptions or therapeutic alternates or when drugs that represent therapeutic advances are not available. Although costs for these drugs or drug classes may decrease, costs for substitute drugs or substitute care modalities or settings may escalate and patient outcomes deteriorate. The evolution of Medicaid formulary management over the past two decades reflects this realization. As a result, legislation has been enacted in many states to address some of these issues in managed care, as reviewed earlier. Other studies are more difficult to assess, show mixed results, or may indicate some effects of controls. With few exceptions, these reports should be interpreted with caution. The committee also observed that most of the formulary systems and controls studied in these reports differ from the VA National Formulary and their restrictions are generally not those used by the VA, for example, prescription limits, or prior approval of the only member of a class. These studies, therefore, are not very helpful in evaluating the restrictiveness of the current VA formulary system. Medicaid Managed Care: Background Medicaid managed care is not new, but it has expanded dramatically in the past 10 years. In 1991, less than 10% of all Medicaid recipients were enrolled in Medicaid managed care plans. Currently, more that half are enrolled (21,167,485 recipients in 585 plans as of June 30, 1998; www.hcfa.gov/medicaid/plansum8.htm). All of the states except Alaska and Wyoming are enrolling Medicaid recipients in managed care plans (National Academy for State Health Policy, 1999). Two states, Tennessee and Arizona, administer their entire Medicaid programs through managed care plans. Provisions in the Balanced Budget Act of

OCR for page 149
DESCRIPTION AND ANALYSIS OF THE VA National Formulary 1997 made it easier for states to enroll Medicaid recipients in managed care, so it is likely that the trend will continue (NPC, 1998). Most Medicaid managed care plans are authorized by HCFA through approval of Section 1915b Freedom of Choice waivers or Section 1115 Research and Demonstration waivers. Section 1915 waivers allow states to forgo some requirements of Medicaid law, such as freedom of choice, comparability, and statewide access. They permit states to increase access to managed care plans, but they do not allow an expansion of benefits or coverage of populations that would not meet the eligibility requirements of Title XIX of the SSA. In addition, they are generally limited to a small geographic area, such as a county, and are approved for 2-year periods. Section 1115 waivers allow states broader exclusions from Medicaid law and are approved for 5-year periods. These waivers permit states to implement statewide programs to test new health care delivery and financing systems, expand coverage to different populations, and expand benefit packages. States are required to demonstrate that the waiver will not increase Medicaid spending; that is, the program must be budget neutral. HCFA has approved 19 Section 1115 waivers for Medicaid managed care. The Balanced Budget Act of 1997 further eased the way for states to enroll Medicaid recipients in managed care by adding a new section (Section 1932) to Title XIX of the SSA, which allows states to enroll recipients in managed care without a waiver if they comply with the new section. With some exceptions, states must offer a choice between at least two managed care organizations or primary care case managers, or at least one plan and one primary care case manager. Lesser restrictions apply in rural areas. States may also “lock in” recipients for 12 months instead of 1 month unless the recipient can show just cause for disenrollment. The act also removed the requirement that contracted health plans have no more than 75% of enrollees from Medicaid. Plans have enrolled mostly women and children eligible through AFDC or TANF, not the more vulnerable and costly SSI aged and disabled population. The Medicaid managed care programs that have begun to enroll the SSI populations have started with small pilot programs limited to a specific geographic area (National Academy for State Health Policy, 1999). As with state Medicaid formularies, it is not easy to characterize the complex and variable benefits provided by the average managed care organization that will affect Medicaid recipients. Drug Benefit in Medicaid Managed Care There is no standard prescription drug benefit package in Medicaid managed care. It is possible that each of the hundreds of different Medicaid managed care programs has a unique drug benefit plan. The benefit provided to Medicaid managed care enrollees is based on the specific contract(s) negotiated between the state and the plans and, when relevant, on Section 1915 and Section 1115 waivers approved by HCFA. Some states have carved out the drug benefit from their managed care benefit package and continue to provide drug benefits through the fee-for-service system.

OCR for page 149
DESCRIPTION AND ANALYSIS OF THE VA National Formulary In general, the drug benefit found in any particular Medicaid managed care plan is similar to the fee-for-service benefit found in that state. The Center for Health Policy Research at George Washington University asked states whether their managed care contracts or requests for proposals (RFPs) included language regarding drug formularies (Center for Health Policy Research, 1998). The level of specificity in the contracts was found to vary widely. For example, the Michigan RFP noted only that health plans may use a drug formulary, whereas the Florida Medicaid managed care contract specified that the plan should not have a pharmacy benefit more restrictive than Medicaid fee for service. The plan could use a preferred formulary as long as adherence was voluntary. New Jersey permits health plans to have a formulary but requires them to provide all medically necessary legend and nonlegend drugs covered by the Medicaid program and to ensure the availability of quality pharmaceutical services for all enrollees. The contract mandates that health plans include in their formularies new drugs that will have a significant impact on patient care and permits exclusion without prior approval only of drugs and drug categories listed in OBRA 1990. The Pennsylvania RFP is another example of the limits placed on Medicaid managed care plan drug benefit design (Commonwealth of Pennsylvania, HC-SW PH RFP #10-97, 1997). This RFP requires that formularies be developed by P&T committees; include drugs in the therapeutic categories currently covered in the fee-for-service program; provide access to all new drugs within 10 days of FDA approval (either through inclusion on the formulary or by prior approval); exclude coverage of all DESI drugs; and exclude any drug marketed by a company that does not participate in the fee-for-service rebate agreement. The document also requires prior approval systems to abide by the OBRA 1990 requirements of a 24-hour response time and access to a 72-hour emergency supply of any reviewed product. The RFP prohibits therapeutic substitution without explicit consent from the attending physician. Similarly, the South Carolina contract requires that service limits such as a drug formulary may not be implemented unless there is a mechanism to cover drugs outside the formulary that are determined to be medically necessary in the treatment of a particular Medicaid managed care enrollee. The Center for Health Policy Research survey responses and the sample RFPs and contracts indicate that although there is variation in the contract language from state to state, the drug benefit in Medicaid managed care plans is generally limited by the conditions outlined in OBRA 1990 and 1993. Perhaps most importantly, the contracts described in this report provide that drugs not listed on the managed care formulary must be made available through a prior approval program. Furthermore, these prior approval programs are generally required to adhere to the statutory requirements for Medicaid fee-for-service prior approval programs listed in OBRA 1993, that is, 24-hour response time and access to a 72-hour emergency supply (Center for Health Policy Research, 1998). OTC coverage was provided by 25 states through their managed care contracts. A more detailed study of Medicaid managed care plans conducted by the National Academy for State Health Policy (1998) found similar variations in the

OCR for page 149
DESCRIPTION AND ANALYSIS OF THE VA National Formulary drug benefit across states and types of enrollees. Of the 45 Medicaid systems reporting information, including Washington, D.C., 34 provided prescription drug coverage. The remaining 11 states carved out their Medicaid pharmacy benefit from the Medicaid managed care system, and subsequently 2 more states (Massachusetts and New York) have done so. States that carve out the drug benefit are able to retain participation in the Medicaid rebate program and guaranteed best-price discounts that otherwise would be lost for the managed care portion of their Medicaid pharmacy benefit. Presumably, state legislation on disclosure of formularies and nonformulary processes, continuation of needed drugs, and off-label coverage, reviewed earlier in this report, would apply to Medicaid managed care but would have relatively minor effect since Medicaid fee-for-service and therefore managed care programs do not usually present barriers that these laws are designed to address. DEPARTMENT OF DEFENSE Health Care System and Pharmacy Benefit The Department of Defense's primary medical mission is to maintain the health of 1.6 million active duty personnel and to provide health care during military operations. The secondary mission is to offer health care to 6.6 million non-active duty beneficiaries, including dependents of active duty personnel and military retirees, their dependents, and survivors. Most care is provided in about 587 military hospitals and clinics (military treatment facilities, MTFs) operated worldwide by the uniformed services. This system is supplemented by care that is paid for mostly by DOD but provided by civilian physicians under the TRICARE program. The TRICARE health care system provides services to both active duty and retired military personnel and their families. TRICARE is DOD 's managed care replacement for CHAMPUS (Civilian Health and Medical Program of the Uniformed Services), which was phased out between 1995 and 1998. TRICARE offers military and retired military beneficiaries three health care options: TRICARE Extra, Standard, or Prime which are preferred provider, standard fee-for-service, and health maintenance organization benefits, respectively (C. Kirby, GAO, personal communication, 2000). DOD health care services were provided at a cost of $14.7 billion in FY 1998. The DOD pharmacy benefit cost $1.3 billion (9%) in FY 1998 and is available through the 587 military treatment facility outpatient pharmacies, TRICARE contractor network or nonnetwork retail pharmacies, and a national contractor mail order program. MTF pharmacies filled about 55 million prescriptions and consumed about three-fourths of the pharmacy budget in FY 1997. Many retired personnel obtain their health care in the private sector but use the DOD pharmacy benefit to fill their private physician prescriptions at no or reduced cost, a practice not condoned by the VHA for veterans (C. Kirby, GAO, personal communication, 2000). Like the VA and the private sector, DOD pharmacy costs are increasing more rapidly than overall health care costs.

OCR for page 149
DESCRIPTION AND ANALYSIS OF THE VA National Formulary The structure and management of the DOD pharmacy benefit will undergo considerable change under the terms of the National Defense Authorization Act (NDAA) for Fiscal Year 2000 (P.L. 106-65, Section 701, enacted October 5, 1999). The following description applies to the current status of the DOD pharmacy benefit. Relevant potential changes and their implications are noted as appropriate. MTF drugs are free to eligible personnel, but varying copayments, which can reach 50% of prescription cost (and a $300 deductible) for nonnetwork retail drugs, are imposed in other settings, retail pharmacies, and the mail order program. MTF drugs are distributed through a prime vendor at prices negotiated by the Defense Supply Center Philadelphia (sic). These prices vary from the Federal Supply Schedule (FSS) discount for covered drugs of 24% below average manufacturer price (AMP) to as much as 70% lower than AMP. TRICARE contractor network pharmacies provide discounted drugs (although at lesser discounts than the FSS), but nonnetwork pharmacies are reimbursed at retail prices. According to GAO (1998), the data collected to monitor and operate the DOD pharmacy benefit are uncoordinated and inadequate. Therefore, it is difficult to be sure of program costs, drug utilization statistics, appropriateness, or safety. The pharmacy benefit also varies significantly across the United States (GAO, 1998). DOD Formulary and Formulary System MTF pharmacies devise and monitor their own closed formularies. According to the Policy for Basic Core Formulary and Committed Use Requirements Contracts (HA Policy 98-034, 27 April 1998), all MTF full-service pharmacies are required to stock and list on their formularies all the items on a Basic Core Formulary (BCF). When there are joint VA and DOD committed- use contract drugs, they represent the mandatory source and use for MTFs just as for VA facilities, and they may be the only members in formulary closed drug classes. The BCF, as of the last quarter of 1999, is a limited formulary with 159 listings (some differ only with respect to dosage or routes of administration forms) in 41 categories (www.pec.ha.osd.mil/BCF/BCFqckr.htm). This formulary requires a single mandatory source for amoxicillin, albuterol inhalers, extended-release diltiazem and verapamil, oral captopril, lisinopril, nortriptyline, cimetidine, and ranitidine. HMG CoA RIs and PPIs are closed classes and limited to cerivastatin and simvastatin, and omeprazole, respectively. At least one SSRI is required. Two ACEIs; two H2R blockers; two alpha blockers; three CCBs; a limited, mostly generic, group of antibiotics; and only first-generation cephalosporins are listed. Under the terms of the NDAA, DOD is required to establish by October 1, 2000, a Uniform Formulary ensuring the availability of drugs in the complete range of therapeutic classes. The details of the new formulary are currently under development, but it will apply to the MTF, retail, and mail order pharmacies. BCF policy requires a quarterly update by the DOD P&T committee, and changes may be requested by MTF P&T committees. The DOD P&T committee meets at the DOD Pharmacoeconomic Center in Texas and consists of armed

OCR for page 149
DESCRIPTION AND ANALYSIS OF THE VA National Formulary forces physicians and pharmacists, representatives from the VA, the Defense Supply Center Philadelphia, TRICARE contractors and the mail order program contractor, and the DOD pharmacoeconomic center. According to the NDAA, this P&T committee must be supplemented by a Uniform Formulary Beneficiary Advisory Panel that will include DOD beneficiary representatives and provide comments to the Secretary of Defense on the new Uniform Formulary. Currently, nonformulary exceptions to MTF formularies require submission of a request and approval of the MTF commander. Mail order exceptions are handled according to the contractor's exceptions and prior approval procedures. The NDAA requires the Secretary of Defense to establish nonformulary procedures, including an appeals process. Since there are no formulary system restrictions, except for copayments and deductibles, at TRICARE network or non-network retail pharmacies, eligible personnel (that is, non-Medicare, non-active duty) can short-circuit formularies and the nonformulary exceptions or prior approval processes by filling prescriptions at these pharmacies at increased cost to themselves and DOD. Presumably, TRICARE network pharmacies will provide drugs according to the new Uniform Formulary in FY 2001, however. The DOD mail order program provides drugs to patients with chronic health conditions at Defense Supply Center Philadelphia negotiated prices using Merck-Medco as the contractor. All categories of military personnel (except Medicare retirees, although the NDAA has ordered a study of a more comprehensive pharmacy benefit for these retirees) are eligible for the program and, except for active duty personnel, are subject to copayments of $4.00 to $8.00 for each 90-day supply. Until the Uniform Formulary is implemented, the mail order program has a limited formulary of preferred and nonpreferred drugs that was devised by the DOD P&T committee. The formulary lists covered drugs, excluded drugs, and drugs under review. As at MTFs and network pharmacies, generic substitution is mandatory. The mail order formulary, as of the last quarter of 1999, had about 80 preferred, contract, injectable, and OTC listings and about 30 nonpreferred or noncontract listings. Less than 10 individual drugs were excluded, as were 10 drug classes, such as immunizations, smoking deterrents, and anabolic steroids, among others. Drugs for cosmetic, weight reduction, or investigational use were also excluded. The committee concluded that the present DOD BCF, mail order formulary, and multiple MTF formularies are not comparable counterparts to the VA National Formulary and formulary system. The DOD formularies cover a limited number of products, and nonformulary exceptions do not appear to be based on the kinds of clinical criteria and review process required by the VHA. The committee has little information on MTF formularies except that they are free to add drugs in many categories or drug classes to the BCF and undoubtedly do. They are also free to restrict access to new or expensive drugs and are reported to do so (GAO, 1998). Presumably, this creates a highly variable pharmacy benefit. Little information was available about the nonformulary exceptions process, except that a request approved by the MTF commander is required. Presumably this, too, might create highly variable access to drugs. Since active duty

OCR for page 149
DESCRIPTION AND ANALYSIS OF THE VA National Formulary FIGURE 5.3 Effect of contract on relative market shares of CCBs. SOURCE: Department of Defense P&T committee minutes (Aug 1999). and Medicare-eligible retired personnel are not eligible for TRICARE contractor retail pharmacy drugs, they must deal with this fragmented and restrictive system as best they can. The committee has no information on the level of satisfaction or health implications of this. The BCF formulary itself is highly restrictive in comparison to the VA National Formulary, which has about eight times as many listings. The effects of this formulary system on quality of care are not known as far as the committee could determine. Some data on cost implications have been collected, for example, Figure 5.3 from the minutes of the DOD P&T committee. Existing DOD databases do not allow detailed cost analyses according to GAO (1998), although the NDAA mandates implementation of the Pharmacy Data Transaction Service to address this problem. Furthermore, as noted earlier in this chapter, restrictions on access to medically indicated drugs by the BCF or some MTFs may divert utilization and costs to other settings, that is, retail pharmacies, other MTFs, or in the case of Medicare-eligible retirees, to the VA (although they would have to see VA physicians to obtain VA pharmacy benefits), where these treatments or alternatives are available. The committee concluded that the DOD BCF and other formulary and formulary systems are in earlier stages of devel-

OCR for page 149
DESCRIPTION AND ANALYSIS OF THE VA National Formulary opment than the VA National Formulary and do not provide useful comparisons. How they will compare after full implementation of improvements under the NDAA is unknown and presumably will depend on the details of changes. GENERAL COMMENTS ON COMPARISONS In examining private-and public-sector formularies and formulary systems in comparison to the VA National Formulary and formulary systems, the committee concluded that some formularies are more open. For example, Medicaid programs are required to offer all drugs on the Federal Supply Schedule. Some formularies are more restricted. For example, they require prior approval for many drugs and entirely exclude some drugs or drug classes. Drugs or drug classes are not excluded in the VA systems. All are variable, and some perhaps much more so (DOD) than the VA. Access barriers are sometimes more frequent, and sometimes less, and they vary in how burdensome they are in both the private and the public sectors. Some controls that may present real impediments to obtaining needed drugs, especially for low-income patients, such as relatively costly copayments and deductibles, are not features of the VA formulary system. Other controls, such as generic substitution and therapeutic interchange are in common use in many systems. The committee could not reach blanket conclusions on the relative restrictiveness of these highly variable comparison formularies and formulary systems. If a formulary appropriately controls drug costs, it may be more important to the VA since cost overruns and cost shifting have entirely different implications in a fixed budget system like the VA (or to some extent managed care) than they do in an open-ended entitlement like Medicaid. As has been concluded so often in this report, the key issue does not appear to be the details of a formulary, providing it is of reasonable size, inclusiveness, and quality. The important element for quality and restrictiveness is timely availability of a safe and effective, medically necessary drug, if not listed, through an exceptions process. A good formulary supports this element and, through its capacity to make quality choices, enhances price negotiations and prudent purchasing.