Printer Friendly Version

Hearings » Is the U.S. Department of Veterans Affairs (VA) Meeting the Pharmaceutical Needs of Veterans? An Examination of the VA National Formulary, Issues of Patient Safety, and Management of the Pharmacy Benefits Program

Statement of Jack Hoadley, Ph.D.

Research Professor
Health Policy Institute
Georgetown University, Washington, DC

Good morning Mr. Chairman and Members of the Subcommittee.  My name is Jack Hoadley, and I am a Research Professor at Georgetown University’s Health Policy Institute.  As a long-time analyst of issues surrounding prescription drug coverage, I have conducted a variety of research projects with regard to formularies and other approaches to managing the use of prescription drugs in Medicare, Medicaid, the VA, and private-sector health plans.  I appreciate the opportunity to speak to the Subcommittee on these important issues.

During congressional debates over the Medicare Part D prescription drug benefit, the role of the VA National Formulary has been commonly invoked.  Some have pointed to the role of the VA Formulary in helping to achieve low prices for the VA, whereas others have made the claim that access to drugs is more restricted in the VA system compared to private plans, especially the private plans offering drug coverage through Medicare Part D.

A report prepared in December 2006 for the Pharmaceutical Research and Manufacturers of America (PhRMA) concluded that “the application of a VA-style formulary process to the Medicare prescription drug program would significantly reduce physician and patient choice of drugs,” and that “a reduction in choice of prescription drugs could be of special concern for the Medicare population.”[1]  In an April 2007 memo, Greg D’Angelo of the Heritage Foundation wrote that “if Congress fixes prices in Medicare and establishes a restrictive national formulary, the program would be less responsive to the diverse and ever-changing needs of beneficiaries.”[2]  As a result of those statements, my colleagues and I decided to shed further light on these issues by conducting an objective comparison of the VA National Formulary to formularies used by Part D plans. 

We examined a sample of 160 commonly prescribed drugs and compared their status on the VA National Formulary to comparable unrestricted coverage for a variety of Medicare Part D plans.  In general, we found that the VA listed fewer drugs on formulary, but our analysis of off-formulary prescribing at the VA suggests that this does not translate into less access to the drugs than exists under Medicare Part D.

Background

Formularies were used by veterans’ medical centers as early as 1955 to help manage pharmacy inventories.  In 1995, the VA took steps to consolidate its bargaining power with drug makers and thus reduce spending.  It combined the formularies of local VA medical centers to create a single formulary in each of 21 Veterans Integrated Service Networks (VISNs).  Two years later, it implemented a national formulary, and by 2007 the VA had completed the process of phasing out the VISN drug lists, making the national formulary the definitive and only VA drug list.[3]

The VA National Formulary functions somewhat differently than most formularies maintained by private organizations.  Since the VA is an integrated system – meaning that veterans go to VA facilities to see VA doctors and fill prescriptions at a VA formulary – the VA has a great deal of leverage to promote the use of a single formulary within its facilities and patient pool.  This distinguishes the VA formulary from those used by most private insurers.  In most private plans, a drug’s formulary status might not be known by a physician when a prescription is written, but instead only determined at the point of sale by the pharmacist, when the patient brings a prescription to be filled at the local pharmacy.  In those cases, the prescribing physician has no particular relationship with the insurer or health plan.  The VA system is more like group and staff model HMOs such as Kaiser Permanente, where the formulary is viewed as a clinical tool to be used by physicians, rather than an enforcement tool of the plan applied at the pharmacy.

The VA National Formulary is managed by practicing VA physicians and regional formulary managers and takes into account safety, efficacy, and cost in deciding what drugs to list on the formulary.  As part of the process, VA clinicians have an opportunity to provide input on the decisions, which helps to create a sense of buy-in for them.  In addition, VA physicians and pharmacists prepare comprehensive written reviews that summarize recent published research on the safety and efficacy of drugs in specific drug classes.  These reviews may be used to make recommendations on a drug’s status within the formulary.  For example, a 2003 review of oral bisphosphonates (typically used to treat osteoporosis) concluded that since Fosamax and Actonel “produce similar results… the VHA should consider these two drugs equivalent clinically, and choose one for use based on best value.”[4] 

The VA considers some classes “closed,” that is, drugs in that class are only covered if they are listed on the formulary.  Generally, only a few drugs in a “closed” class are listed on the national formulary.  The VA justifies the exclusion of others on the grounds that they are therapeutically interchangeable – “equivalent in terms of efficacy, safety and outcomes” – to the drugs on the formulary.  The VA can then obtain lower prices for on-formulary drugs through competitive contracts by committing to use them whenever clinically appropriate.[5]  Most often, these are classes where multiple brand-name drugs are available, but few or no generic alternatives.  Drugs in other classes may have restrictions, meaning that physicians are encouraged to prescribe certain “preferred” drugs in that class over others.[6]  Such restrictions are used both to create leverage in negotiating prices and to restrict inappropriate use of certain drugs. 

According to the VHA, drugs not listed on the national formulary may be prescribed through a non-formulary request process designed to ensure timely, evidence-based decisions.  Non-formulary drugs may be approved for use by a patient if:

  1. Formulary agents are contraindicated;
  2. Formulary agents have caused adverse reactions;
  3. All formulary alternatives have demonstrated therapeutic failure;
  4. No formulary alternative exists; or
  5. “The patient has previously responded to a non-formulary agent and risk is associated with a change to a formulary agent.”

Through this process, drugs not listed on the VA formulary can be prescribed.  An informal survey by the VA pharmacy benefit manager in 1998 found that “88 percent of waiver requests [for non-formulary medications] were approved.”[7]  Nevertheless, VA formulary compliance is very high.  The VA estimates that overall use of drugs not on the formulary is about 5 percent.  This high compliance, especially for the closed classes, is aided by the sense of buy-in by VA clinicians and the reliance on clinical evidence as a key component of the decision-making process.  Adherence for a closed class sometimes reaches 90 percent within three months of a formulary change and over 98 percent within six months.[8]

Several studies over the past decade have asked whether the VA National Formulary too strictly limits the prescription drugs available to veterans.  After a 1999 report by the House Committee on Appropriations raised such concerns, a mandated report by the Institute of Medicine, released in 2000, concluded that “the VA National Formulary is not overly restrictive, and the limited available evidence suggests that it has probably meaningfully reduced drug expenditures without demonstrable adverse effects on quality.”  The reviewers also found, however, that the National Formulary lacked “essential systems to assure that new drugs are expeditiously reviewed” and that more needed to be done to ensure “that a responsive process for assuring access to medically necessary exceptions to the formulary is consistently in place.”[9]

In 1999 and 2000, the VA commissioned two independent surveys of VA prescribers’ perceptions of the VA National Formulary.  The first survey identified a subset of physicians who believed this formulary was more restrictive than most in the private sector or that it impinged on providing quality care to their patients.  But nearly two-thirds thought they were able to prescribe needed drugs and that their patients could obtain nonformulary drugs when needed.[10]  Respondents to the second survey also indicated general satisfaction with the formulary and agreed that it was important for containing costs and ensuring good value.  Although they reported that roughly 90 percent of waiver requests for off-formulary drugs were approved, about one-third of the physicians indicated that approvals took three days or longer.[11]  In 2001 the Government Accountability Office told the Congress that “prescribers reported that the national formulary generally contains the drugs their patients need or, when necessary, prescribers can usually get non-formulary drugs.”[12] 

In a 2005 report, economist Frank Lichtenberg concluded that the VA was tardy in its addition of newly FDA-approved drugs to the VHA formulary, and noted “that only 19% of the drugs approved since 2000 were on the VHA formulary.”[13]  Lichtenberg attributed the use of older drugs in the VA formulary to a reduction in veterans’ mean age of death, by 2.04 months.  But Lichtenberg’s report was rebutted by the VA’s Michael Valentino in a 2007 presentation.[14]  He offered evidence that veterans experienced greater life expectancies at birth than other populations.  Arguing that “newer is not always better,” he pointed out that “many ‘new’ drugs are actually ‘me too’ drugs” that are essentially the same as – and not necessarily more effective than – treatments already on the market.  Valentino also noted that 23 new drugs (including some that Lichtenberg lists as not covered by the VA) were taken off the market for safety reasons between 1980 and 2005, and concluded, “what is the rationale for exposing patients to drugs with unknown risks, when there is little or no clinical advantage?”  Valentino insisted that the “VA reviews all new molecular entities for consideration for national formulary listing in a timely fashion,” and that it based its decisions on cost and efficacy data.

Comparing VA and Medicare Part D Plan Formularies

To compare the VA National Formulary with those offered by the Medicare Part D prescription drug plans, my colleagues and I focused on a pre-selected sample of 160 drugs representing more than half the prescription volume for Medicare beneficiaries.  The sample includes all drugs in 14 drug classes, as well as other commonly prescribed drugs, and has nearly equal numbers of generic (n=76) and brand-name (n= 84) medications.[15]  For Medicare, our analysis covers 47 standalone prescription drug plans offered on a national or near-national basis in 2007, as well as two formularies offered in different regions by a group model HMO participating in the Medicare Advantage program.  We considered the number of plans that listed a drug, whether the drug was placed on a generic, preferred, or specialty tier, and whether the drug was restricted through any utilization management tools (prior authorization, step therapy, or quantity limits).  For the VA, the analysis considered whether a drug was listed on formulary and the yearly volume of prescriptions written for that drug.[16] 

A simple count of the 160 sample drugs on the VA formulary suggests that it lists fewer drugs (82 drugs) than any of the national or near-national Part D plans (median number of drugs listed = 136; minimum = 99).  However, there are key reasons that make this simple accounting misleading.  First, veterans have unrestricted access to all drugs listed on the VA formulary and access to additional drugs if they seek authorization.  By contrast, Part D plans include drugs on their formularies that may only be available to beneficiaries subject to utilization restrictions, such as prior authorization, step therapy, or quantity limits.  In the VA system, patients only need to seek prior authorization or undergo step therapy for drugs not listed on the VA National Formulary.

The TNF Inhibitor class, which includes three expensive specialty drugs primarily used for treating rheumatoid arthritis (Enbrel, Humira, and Remicade), illustrates this dynamic.  Neither the VA nor the median Part D plan covers any of these drugs outright.  In the VA, the drugs are considered off formulary, but they are available when patients or their doctors request authorization.  Most Medicare drug plans list these three drugs on formulary, but require prior authorization before the drug is actually covered.  In these two situations, patients face a similar level of restricted access, but the formulary status is different.

Second, many Part D plans list drugs on a “non-preferred” tier with a higher level of cost sharing.  This system creates a financial incentive for beneficiaries to adhere to their plan’s formulary, even when their doctor may not know which drugs are preferred or not preferred.  By contrast, the VA does not use tiers; cost sharing for any drug is limited to $8 for a month’s supply for those patients subject to the copayment.  Because VA doctors use only the VA formulary, they can become familiar with its coverage; financial incentives are not needed to steer use. 

Because of these system differences, it is most relevant to compare the VA’s formulary to the list of drugs that are on a preferred tier in a Part D plan’s formulary, without designations for either prior authorization or step therapy.[17]  For the national and near-national plans, the median number of unrestricted on-formulary drugs is 104, compared to 136 when restricted drugs are included.  By this criterion the typical Medicare plan formulary comes closer to the 82 drugs listed on the VA National Formulary (Table 1).

Table 1.  Number of Drugs with Unrestricted Coverage, VA Formulary and Selected Medicare Plans, 2007

 

All Drugs (N=160)

Generic Drugs (N=76)

Brand Drugs (N=84)

On VA Formulary

82

56

26

Median, 47 national Part D plans

104*

72

37

     Minimum, 47 national Part D plans

80*

48

20

     Maximum, 47 national Part D plans

149

76

73

Top 10 Part D plans, by 2006 enrollment

 

 

 

     AARP Medicare Rx Basic

112

75

37

     Community Care Rx Basic

98

70

28

     Humana/ Complete

121

76

45

     Humana/ Enhanced

121

76

45

     Humana/ Standard

149

76

73

     Medicare RX Rewards Value

117

76

41

     Prescription Pathway Bronze

118

74

44

     Silverscript (Caremark)

102

63

39

     United Healthcare Rx Basic

97

73

24

     Wellcare/Signature

94

74

20

Group Model Medicare Advantage Plans

 

 

 

     Kaiser Permanente, Northern California

77

45

32

     Kaiser Permanente, Southern California

79

50

29

*Median and minimum for “all drugs” are measured directly and not the sum of the brands and generics values.

Note:  Unrestricted coverage is defined as coverage on a preferred tier with no prior authorization or step therapy requirements.

One view of the difference between the VA’s formulary and those used by Part D plans is how they treat the generic drugs in our sample (Table 1).  The typical Part D plan lists over 90 percent of generic drugs, and several of the ten most popular plans list all of the generics we studied.  By contrast, the VA lists only 56 of 76 sample generic drugs.  This reflects the different perspective of those designing these formularies.  The VA chooses preferred drugs among competing generics based on a combination of clinical evidence and price.  By contrast, Part D plans have incentives to list on formulary most or all of the generics in a particular class.  Lacking any close relationships with prescribing physicians, they must rely on enforcement at the pharmacy to encourage use of one particular generic over another.  Doing so risks alienating their enrollees for minimal financial gain and may discourage enrollment if drugs are listed as off-formulary on the Medicare’s online Drug Plan Finder.

As noted above, the approach to formulary design in an integrated health plan is more like the VA system than other Part D plans.  It is not surprising, therefore, that Kaiser Permanente formularies were similar in scope to the VA formulary (Table 1).  In both of its California regions, Kaiser Permanente listed slightly fewer sample drugs on formulary (77 and 79) than the VA National Formulary.  Like the VA, the Kaiser plans are more likely than other Part D plans to omit generic drugs from their formularies.

Comparisons at the Drug Class Level

Comparisons between the VA National Formulary and the Part D plan formularies vary considerably by drug class (Table 2).  One reason for the variation is that some drug classes (e.g. beta blockers) consist mostly of generic drugs.  Most Part D plans list nearly all generics on formulary, whereas the VA is more likely to omit generic drugs from its formulary for such classes.  Specific program rules also affect the comparisons.  For example, Part D plans include more anti-depressants at least partly because Medicare guidelines require that nearly all unique anti-depressants be listed on formulary, although the guidelines do not require that coverage be unrestricted.

Table 2.  Formulary Listings by Class, VA and Part D Formularies, 2007


Drug Class

Drugs Studied

VA National Formulary

Median, 47 National Plans*

KP Northern CA

KP Southern CA

Anti-Dementia Agents

6

5

4

3

3

Anti-Depressants

30

18

23

24

24

Beta Blockers

15

7

14

6

9

Calcium Channel Blockers

9

5

6

4

3

Cholesterol Agents

18

7

11

4

5

Diabetes Agents

16

5

12

6

8

Proton Pump Inhibitors

6

1

2

2

2

Hormonal Agents

12

6

4

7

6

Renin-Angiotensin

18

8

10

3

3

TNF Inhibitors

3

0

0

3

3

Other Common Drugs

27

20

22

14

12

TOTAL, SELECTED DRUGS

160

82

104

77

79

* Drugs listed on a generic or preferred tier and without prior authorization or step therapy restrictions.  Note that the median value for the total is not the total of the class medians.

The pattern is similar for the Kaiser Permanente formularies, but the details are different.  Kaiser’s formularies, for example, list fewer drugs in categories such as anti-cholesterol agents or the renin-angiotensin drugs used to treat hypertension.  But Kaiser lists more anti-depressants than the typical Part D plan.  Kaiser also lists more drugs without restriction in the TNF inhibitor class than either the VA or the typical Part D plan, because they do not require prior authorization for these drugs.

The Class of Cholesterol Drugs

The contrasting ways that formularies work in the VA compared to the standalone Part D plans can be illustrated with the class of cholesterol drugs.  The VA’s coverage of cholesterol agents has been criticized because the formulary does not list some popular anti-cholesterol drugs such as Crestor and Lipitor, while the majority of national Part D plans list them without restrictions.[18]  As shown in Table 2, the typical Part D plan lists 11 of the anti-cholesterol agents in our sample, whereas the VA lists 7 drugs and Kaiser Permanente lists just 4 in one region and 5 in another.

Whenever possible, the VA suggests that a “high potency” formulary statin should be the first-line treatment prescribed for the patient.  If he or she fails to meet clinical goals on that drug, physicians are advised to consider a second-line therapy (such as niacin or non-formulary Zetia) or a switch to a non-formulary statin, such as Lipitor.[19]  These guidelines mean that non-preferred or even non-formulary drugs are recommended and accessible to veterans, and they may be prescribed in greater numbers than their on-formulary or unrestricted counterparts.  In fact, as Table 3 shows, there is more utilization of Zetia, considered a second-line therapy not listed on formulary, than the two on-formulary drugs in the Bile Acid Sequestrants group.  Similarly, prescribing of Lipitor, another second-line therapy that is not on formulary, is higher than for Lescol, one of the on-formulary statins.[20]

Table 3.  Formulary Listing of Cholesterol Agents, 2007

Drug*

Generic Name

On VA Formulary?

# of 47 Part D plans with unrestricted coverage

VA Utilization

FY 2006

Bile Acid Sequestrants  

CHOLESTYRAMINE

CHOLESTYRAMINE

Y

46

52,249

Welchol

COLESEVELAM

N

20

4,714

Colestid

COLESTIPOL

Y

18

168,976

Cholesterol Absorption Inhibitors 

Zetia

EZETIMIBE

N

36

369,783

Fibrates

Tricor

FENOFIBRATE

N

37

130,181

GEMFIBROZIL

GEMFIBROZIL

Y

47

1,771,658

Nicotinic Acid

Niaspan ER

NIACIN

Y

41

1,258,306

Omega-3 Fatty Acids 

Omacor

OMEGA-3 ACID

N

13

827

Statins

Lipitor

ATORVASTATIN

N

34

711,138

Lescol

FLUVASTATIN

Y

8

500,954

Altoprev ER

LOVASTATIN

N

9

12

LOVASTATIN

LOVASTATIN

Y

47

1,424,081

Mevacor

LOVASTATIN

N

4

376,688

Pravachol

PRAVASTATIN

N

1

56,255

PRAVASTATIN

PRAVASTATIN

N

33

**

Crestor

ROSUVASTATIN

N

33

144,341

SIMVASTATIN

SIMVASTATIN

Y

45

**

Zocor

SIMVASTATIN

N

5

16,487,514

* Drug names in all capital letters are generic drugs.

** Data are not available for the newly approved generic versions of pravastatin and simvastatin.  Previously, the brand version of Zocor was on formulary.

Comparative effectiveness reviews of clinical evidence have led to a similar conclusion to that of the VA.  The Consumer Reports “Best Buy Drugs” report, based on research by the Drug Effectiveness Review Project, recommends use of one of three generic statins, with Lipitor as an alternative for patients who have had a heart attack or acute coronary syndrome together with highly elevated LDL.[21]  Among the 47 national and near-national Part D plans, 34 list Lipitor and 33 list Crestor, while nearly all list generic drugs lovastatin and simvastatin.

Another factor that affects the formulary comparisons is the Medicare program’s guidance requiring that Part D plans list on formulary at least one drug in each subgroup of cholesterol drugs, although plans may use coverage restrictions.  The VA has no such mandatory coverage requirement.  Because Zetia and Omacor are the only drugs in their subgroups, Part D plans must list them, whereas the VA formulary does not.  Nevertheless, shown in Table 3, not all Part D plans have unrestricted coverage of these drugs.  The VA, which recommends Zetia as a second-line therapy, filled about 370,000 prescriptions for the drug. 

Conclusions

An objective comparison of unrestricted coverage by Medicare Part D plans to the VA National Formulary shows that the VA formulary is modestly smaller than the typical Part D plan formulary and about the same as formularies used by Kaiser Permanente.  But formulary size is not the same as access to drugs.  This small gap largely reflects the difference between integrated and non-integrated health systems and the resulting approaches to ensuring access.

The VA National Formulary is closely tied to its prescribing system.  Like a staff-model or group-model HMO, physicians in the VA system participate in creating the formulary and commit to prescribing from it when it meets their patients’ needs.  Patients in both of these integrated health care systems are less likely than other types of Part D or private health plans to receive a prescription, only to be told later that it is not covered or covered at a much higher price than a more preferred drug.  The incentive in integrated systems is for physicians to prescribe from the formulary when they can, but it is a relatively straightforward process to obtain authorization for any drug that is not on the formulary.  By contrast, physicians treating patients in non-integrated systems face a variety of formularies for the different plans in which their patients are enrolled, and they are unlikely to prescribe according to each patient’s formulary unless the plan or patient points out which drugs are preferred.  It is essential when making comparisons with regard to access to drugs between the VA and other plans to keep these differences in mind.

A full comparison of access to prescription drugs between the VA system and other health systems would require more extensive studies, such as surveys or clinical outcome studies.  Nevertheless, this comparison of Medicare Part D plan formularies to the VA formulary lends support to the conclusion that veterans maintain good access to prescription drugs through the VA National Formulary.



[1] “Comparison of Compounds on the Formularies of Medicare Prescription Drug Plans (PDPs) and the Department of Veterans Affairs Veterans Health Administration (VA) National and Regional Formularies,” prepared for the Pharmaceutical Research and Manufacturers of America by Covance Market Access Services Inc., December 2006.

[2] Greg D’Angelo, “The VA Drug Pricing Model,” The Heritage Foundation, 11 April 2007.

[3] GAO, “VA Drug Formulary: Better Oversight is Required, but Veterans Are Getting Needed Drugs,” Report to the Ranking Member, Senate Committee on Veteran’s Affairs, January 2001.  Donna Young, “VA’s 10-Year Journey to One Formulary Concludes,” American Journal of Health-System Pharmacy, 64 (15 March 2007): 578-580. 

[4] Marc C. Geraci, “Drug Class Review Oral Bisphosphonates in the Treatment of Osteoporosis,” VHA Pharmacy Benefits Management Strategic Healthcare Group and the Medical Advisory Panel, September 2003.

[5] GAO, “VA Drug Formulary: Better Oversight is Required, but Veterans Are Getting Needed Drugs.”

[6] Institute of Medicine, Description and Analysis of the VA National Formulary (Washington: National Academic Press, June 2000).

[7] IOM, Description and Analysis of the VA National Formulary.

[8] Michael Valentino, “Overview of the VA Pharmacy Benefits Management Strategic Health Care Group (PBM),” presentation to the American Enterprise Institute, 19 January 2007.

[9] IOM, Description and Analysis of the VA National Formulary.

[10] Peter A. Glassman et al., “Physician Perceptions of a National Formulary,” American Journal of Managed Care 7:3 (March 2001), pp. 241-251.

[11] Peter A. Glassman et al., “Physician Satisfaction with Formulary Policies: Is It Access to Formulary or Nonformulary Drugs that Matters Most?” American Journal of Managed Care 10:3 (March 2004), pp. 209-216.

[12] “VA Drug Formulary: Better Oversight is Required, but Veterans are Getting Needed Drugs.”

[13] Frank R. Lichtenberg, “Older Drugs, Shorter Lives? An Examination of the Health Effects of the Veterans Health Administration Formulary,” Center for Medical Progress at the Manhattan Institute, October 2005.

[14] Michael Valentino, “Overview of the VA Pharmacy Benefits Management Strategic Health Care Group.”

[15] For more details on our sample of drugs, see Jack Hoadley et al., “An In-Depth Examination of Formularies and Other Features of Medicare Drug Plans,” Henry J. Kaiser Family Foundation, April 2006, and Jack Hoadley et al., “Benefit Design and Formularies of Medicare Drug Plans: A Comparison of 2006 and 2007 Offerings – A First Look,” Henry J. Kaiser Family Foundation, November 2006.

[16] Data on whether a drug is on the VA National Formulary and its restrictions were obtained from the VA’s website (January 2007 version).  Information on the prescription volume was provided to the authors by the VA.  The VA’s formulary lists all drugs by chemical name; thus we assumed that when both a generic and a brand-name version of a particular chemical exist, only the generic version is listed.

[17] When a drug has a quantity limit, we do not treat that as a restriction.  Some plans designate large numbers of drugs with quantity limits, apparently to restrict the dispensing of prescriptions of more than 30 days.

[18] Deroy Murdock, “VA Program No Model for Helping Americans Buy Medications,” Deseret News, 31 December 2006.

[19] “Ezetimibe (Zetia®) for Nonformulary Use,” VHA Pharmacy Benefits Management Strategic Healthcare Group and the Medical Advisory Panel, updated January 2007.

[20] After this analysis was completed, changes were made to the VA National Formulary’s treatment of statins in response to FDA approval of generic versions of Pravachol and Zocor. 

[21] http://www.bestbuydrugs.org/drugreport_DR_Statins.shtml