Opportunities for Improving the Cost-Effectiveness of
`Antihypertensive Treatment
`
`WILLIAM B. STASON, M.D., M.S.
`Boston,
`Massachusetts
`
`Health,
`Public
`of
`School
`Harvard
`the
`From
`Administration
`Veterans
`Roxbury
`BrocktoniWest
`qequests
`Boston, Massachusetts.
`Medical
`Center,
`6.
`for
`reprints
`should
`be addressed
`to Dr. William
`Stason,
`Veterans
`Administration
`Medical
`Center,
`14900
`V.F.W.
`Parkway,
`West Roxbury,
`Massachu-
`setts
`02132.
`
`important public health problem, both
`is an extremely
`Hypertension
`medically and economically. The cost burden of treatment may sig-
`nificantly compromise care for the individual patient, while in aggre-
`gate the direct costs of antihypertensive
`treatment
`in the United
`States approach $8 billlon a year. Improved
`insurance coverage and
`efforts
`to control
`the costs of antihypettensive
`treatment
`are
`needed. Efforts to reduce the costs of care, with minimal or no re-
`duction
`in its quality, should
`focus on the following:
`(1) limiting
`treatment
`to patients with sustained diastolic hypertension;
`(2) im-
`proving
`the efficiency of the delivery process; and (3) emphasizing
`“low-cost
`prescribing
`strategfes.”
`Tfie uncertainty
`that
`remains
`over the risk-benefit
`ratio of pharmacologic
`treatment
`for patients
`with very mild hypertension
`(90 to 94 mm Hg diastolic)
`raises addi-
`tional questions. Even if treatment of mild hypertension
`is effective,
`it is without doubt
`less cost-effective
`than treatment of moderate
`and severe’hypertension.
`Is this cost worthwhile? Such trade-offs of
`cost and benefits wilf increasingly have to be confronted
`in the face
`of limited health care resources.
`
`Policy makers, employers, and patients alike are ever more frequently
`expressing concerns over the high and increasing costs of health care
`and, perhaps more importantly, over whether these high costs are worth-
`while in terms of resulting health benefits., Each group has its own views
`on the appropriate level of costs and on measures of benefit that are most
`valued. The message, nonetheless, is a consistent and undeniable one.
`Particularly germane to this discussion is evidence that the cost burden
`of hypertension care may have deleterious effects on some hypertensive
`patients. This conclusion is suggested by the results of two surveys of
`physicians and patients conducted recently by the Gallup. Organization
`[1,2], and is further supported by the Rand Health Insurance Study [3].
`Yet another study, a survey of hypertensive patients in Georgia [4], has
`indicated that the cost burden of hypertension care is particularly onerous
`in patients who are poor and have moderate to severe hypertension.
`Two primary options exist that may reduce the cost burden for the
`patient with hypertension. One is to reduce the costs of care, and the
`other is to improve insurance coverage. There is no question that we
`should pursue the goal of expanded insurance coverage for medications,
`office visits, and essential laboratory tests for patients with moderate or
`severe hypertension. More controversial, but equally important, however,
`is the simultaneous need to explore opportunities for controlling the costs
`of care. Some physicians argue that trade-offs between cost and quality
`are inevitable, and that cost-control initiatives (those that are already in
`
`December
`
`31, 1996
`
`The American
`
`Journal
`
`of Medicine
`
`Volume
`
`61 (suppl
`
`6C)
`
`45
`
`LOWER DRUG PRICES FOR CONSUMERS, LLC
`Exhibit 1030-1
`IPR2016-00379
`
`

`
`SYMPOSIUM ON ANTIHYPERTENSIVE
`
`TREATMENT-STASON
`
`go-
`95
`305
`94
`104
`MU
`Treat only
`
`Scg
`Treat
`detection and
`of hypertension
`rigwe 1. Cost-effectiveness
`treatment
`(1984 dollars)
`for sustained diastolic hypertension
`or screening
`and
`treatment
`of all persons with diastolic
`blood pressure of 95 mm Hg or above. Full adherence
`(dark
`crossed-hatched
`areas)
`and expected
`adherence
`(light
`cross-hatched
`areas) are shown. Reproduced
`with permis-
`sion
`from
`[5].
`
`place and those being contemplated) will necessarily re-
`duce the quality of care. But is this really so?
`In support of the argument for greater cost control, this
`article will first review the magnitude of the hypertension
`problem from an economic perspective; second, highlight
`several findings from our study of the cost-effectiveness of
`hypertension detection and treatment; and finally, suggest
`some opportunities for reducing the costs of care.
`COST OF ANTIHYPERTENSIVE
`TREATMENT
`The cost of treating cardiovascular disease in the United
`States in 1986 has been estimated by the American Heart
`Association to be $78.6 billion. This figure includes medi-
`cal and long-term care expenses and indirect costs due to
`reduced productivity from disability, but it omits the very
`real costs associated with premature deaths from stroke
`and myocardial icfafctions. Perhaps as much as a third of
`the total costs ofpa;diovascular disease may be attributed
`indirectly to hyperfens[on through associated, accelerated
`rates of stroke and other cardiovascular events.
`At an avergge annual trelltment cost of $400 per pa-
`tient, nearly .$O’billiop ig being spent directly each year on
`
`in the United States, and
`the treatment of hypertension
`the pharmaceutical
`industry estimates that $3.7 billion
`was spent worldwide on antihypertensive medications in
`1985. No matter how they are examined, the economic
`consequences of hypertension are enormous. This fact
`underscores the importance of vigorously pursuing oppor-
`tunities to reduce the adverse health impacts of hyperten-
`sion while at the same time using the resources devoted
`to its treatment wisely.
`
`OF ANTIHYPERTENSIVE
`
`COST-EFFECTIVENESS
`TREATMENT
`Cost-effectiveness analysis permits one to examine the
`relationship between the costs and benefits of treatment,
`thereby providing a yardstick by which to compare one
`health care program with another or to compare the treat-
`ment of hypertension with the treatment of other condi-
`tions, such as cancer or coronary artery disease. In our
`studies, we have expressed cost-effectiveness ratios in
`terms of dollars per quality-adjusted years of life, an im-
`portant but complicated concept that encompasses both
`patient preferences for the quality of their lives and a con-
`sideration for the length of those lives. A lower cost-effec-
`tiveness ratio indicates relatively greater cost-effective-
`ness.
`If resources are limited, the cost-effectiveness argu-
`ment would suggest that funds should be allocated first to
`the most cost-effective program and then to progressively
`less cost-effective programs until resources are ex-
`hausted. Obviously, many
`influences other than cost-
`effectiveness considerations will affect the decisions that
`are actually made.
`To examine the cost-effectiveness of treatment for hy-
`pertension, we used the results of the Framingham Heart
`Study [5] to estimate the morbidity and mortality benefits
`of blood pressure control. Costs included the direct costs
`of antihypertensive
`treatment
`less the savings that re-
`sulted from preventing strokes or myocardial infarctions
`caused by untreated hypertension.
`Three conclusions from our study are particularly rele-
`vant to this symposium. First, we found that antihyperten-
`sive treatment does not save medical care costs. Only 22
`percent of the treatment costs for moderate hypertension
`and 15 percent of those for mild hypertension were offset
`by savings from the cardiovascular events prevented. The
`cost of treatment, therefore, can only be justified in human
`terms: in lives saved and in reduced disability.
`Second, it was found that the cost-effectiveness of anti-
`hypertensive treatment is directly related to the pretreat-
`1, the
`ment level of blood pressure. As shown in Figure
`cost-effectiveness ratios ranged from about $12,000 per
`quality-adjusted year of life gained for patients with pre-
`treatment pressures in the moderate range (pretreatment
`diastolic blood pressure 105 mm Hg and above) to more
`than $60,000 per quality-adjusted year of life in those with
`very mild hypertension (90 to 94 mm Hg diastolic).
`
`46
`
`December
`
`31, 1986
`
`The American
`
`Journal
`
`of Medicine
`
`Volume
`
`81 (suppl 6C)
`
`LOWER DRUG PRICES FOR CONSUMERS, LLC
`Exhibit 1030-2
`IPR2016-00379
`
`

`
`SYMPOSIUM ON ANTIHYPERTENSIVE
`
`TREATMENT-STASON
`
`TABLE I
`
`Clinic Characteristics Associated with
`Ambulatory Care Costs or Treated
`Diastolic Blood Pressure in Veterans
`Administration
`Hypertension
`Clinics
`
`Variable
`
`15 minutes
`
`Direction of
`Association’
`
`cost
`-
`+
`NS
`-
`+
`NS
`NS
`+
`+
`-
`+
`
`DBP
`-
`NS
`-
`-
`-
`-
`-
`NS
`NS
`NS
`-
`
`Clinic waiting
`Length
`of visit
`Time
`spent
`counseling
`Clinic
`director
`satisfied
`Clinic
`director
`feels
`supported
`Hypertension
`therapist
`has one supervisor
`Hypertension
`therapist
`feels
`supported
`Hypertension
`therapist
`cultures
`throat
`Clinic
`director
`treats
`patients
`Visits
`per FTE
`Visits
`per patient
`
`time
`
`less
`
`than
`
`year
`
`Third, it was found that the problem of maintaining life-
`long adherence to treatment markedly reduces both the
`effectiveness and the cost-effectiveness of care. Many
`factors influence adherence, among which are side ef-
`fects of the medication, the convenience of treatment, and
`the costs of treatment.
`OPPORTUNITIES TO REDUCE COSTS
`Opportunities to reduce the costs of antihypertensive care
`revolve around decisions concerning who is treated and
`the drug treatment prescribed, as well as the efficiency
`and effectiveness of the care delivery process. A few ex-
`amples will serve to highlight some possibilities.
`Cost savings, without sacrifices in the quality of care,
`almost certainly would result if pharmacologic treatment
`were reserved for those patients with sustained elevations
`of blood pressure documented at multiple office visits. As
`many as a third of all patients with elevated blood pres-
`sure at one office visit will have normal blood pressure at
`subsequent visits [6]. Although there may be a risk of sus-
`tained hypertension developing in such “labile hyperten-
`sive” patients, the benefits of antihypertensive treatment
`in such patients have never been demonstrated. The Gal-
`lup survey, reported at this symposium, indicates that as
`many as one third of all physicians proceed with drug
`treatment for patients with diastolic blood pressures in the
`90 to 99 mm Hg range at the first office visit. If this finding
`is representative of the United States as a whole, as many
`as 10 percent of the people being treated for hypertension
`(one third times one third), or well over one million Ameri-
`cans, are receiving antihypertensive
`treatment need-
`lessly. The potential for cost savings is obvious.
`Uncertainties that persist over whether the benefits ex-
`ceed the risks of treatment in the large number of patients
`with very mild hypertension (90 to 94 mm Hg diastolic)
`suggest
`further potential
`for cost savings. Particularly
`when quality-of-life issues are considered, conservatism
`in offering pharmacologic treatment to these patients may
`be warranted on risk-benefit grounds,
`let alone cost-
`benefit ones.
`Improving the efficiency of delivery of antihypettensive
`care provides still another opportunity to reduce costs. A
`study we performed on the cost-effectiveness of care in
`more than 3,000 patients who were under the care of a
`network of United States Veterans Administration clinics
`provides some interesting insights in this regard (Table I).
`After controlling for patient characteristics, we found that
`shorter and less frequent clinic visits, shorter clinic waiting
`times, and greater provider productivity (measured as the
`number of yearly visits per full-time staff position) were all
`independently associated with lower costs. These findings
`are not surprising, but they do serve to identify clinic char-
`acteristics that are subject to managerial control.
`Other more subjective variables, such as the level of job
`satisfaction of the clinic director, his or her perception of
`support by superiors, and the degree of responsibility for
`
`blood pressure;
`
`DBP = diastolic
`equivalent.
`regression
`of a multiple
`‘Results
`characteristics.
`Associations
`
`NS = not significant;
`
`FTE =
`
`full-time
`
`that also controls
`analysis
`are at p ~0.05.
`
`for patient
`
`TABLE ii
`
`Average Costs of Treatment in Veterans
`Administration Hypettenslon Clinics (1981
`dollars)
`
`Visits
`(step
`Medications
`tests
`Laboratory
`Total
`ambulatory
`
`l-38
`
`percent;
`
`step 2-41
`
`percent)
`
`costs
`
`per patient-year
`
`$158
`$118
`$ 49
`$325
`
`to non-physician
`clinical decisions that was delegated
`hypertension therapists, were also important predictors of
`costs. Several of these clinic characteristics were also
`associated with better blood pressure control. “Shorter
`clinic waiting time”
`is especially interesting, because it
`appears to be related both to better blood pressure control
`and reduced costs.
`The costs of antihypertensive treatment in these Veter-
`ans Administration clinics (Table ii) averaged $325 per
`patient-year of care in 1981 dollars, or approximately
`$435 in 1986 dollars. Nearly half of these costs (49 per-
`cent) were for office visits, and more than one third (36
`percent) were for medications. Medication costs signifi-
`cantly underestimate
`those for the average patient with
`hypertension, because the Veterans Administration ob-
`tains drugs more cheaply than could an individual patient.
`Opportunities to reduce the costs of antihypertensive
`care through judicious selection of medication regimens
`are especially appealing. Costs vary widely among
`classes of drugs and between generic and trade name
`alternatives. Moreover, some antihypertensive drugs may
`require the additional use of relatively expensive potas-
`sium supplements. Hydrochlorothiazide costs about $15
`
`December 31,1986
`
`The American Journal of Madlclna
`
`Volume 81 (suppl BC)
`
`47
`
`LOWER DRUG PRICES FOR CONSUMERS, LLC
`Exhibit 1030-3
`IPR2016-00379
`
`

`
`SYMPOSIUM ON ANTIHYPERTENSIVE
`
`TREATMENT-STASON
`
`TABLE
`
`III
`
`from a “Cost-
`Potential Cost Savings
`Saving”
`Treatment
`Strategy
`
`Regimen
`
`Cost per Year*
`(dollars)
`
`chloride
`of patients
`
`51
`
`97
`138
`422
`
`125
`162
`175
`402
`10 percent,
`
`plus
`
`plus
`
`blocker
`
`regimen
`
`changes+
`
`plus potassium
`diuretic
`A. Generic
`in the 20 percent
`supplements
`hypokalemia
`develops
`in whom
`B. Potassium-sparing
`diuretic
`(Dyazide)
`C. Generic
`beta
`(propranolol)
`D. Captopril
`Cost
`per year with
`A
`B
`C
`D
`1986,
`August
`on Red Book Update,
`‘Based
`pharmacy
`fee per 100 units dispensed.
`$2.00
`rates of 20
`reaction
`drug
`+Assumes
`regimen
`changes
`due
`to adverse
`and 8 percent
`for
`percent
`for diuretics,
`13 percent
`for beta blockers,
`captopril.
`The
`therapy
`of patients
`experiencing
`side effects
`with cap-
`topril
`is changed
`to regimen
`A. The
`therapy
`of patients
`receiving
`regi-
`mens A, B, or C who experience
`side effects
`is changed
`to captopril.
`
`men D (captopril). These figures (comparing regimens A
`and D) would extrapolate to cost savings of $277 million
`dollars per million patients treated. Unless the additional
`costs of captopril (or calcium channel blockers) can be
`rationalized on the basis of better blood pressure control
`or improved quality of life (and currently available evi-
`dence does not adequately support either circumstance),
`the argument is convincing from a cost-effective perspec-
`tive that diuretics should remain the first line of treatment.
`
`TRADE-OFFS
`
`OF COST AND QUALITY
`
`Finally, trade-offs of cost and quality need to be consid-
`ered in situations where marginal benefits are purchased,
`but at a relatively higher price. One such trade-off involves
`the pharmacologic
`treatment of patients with very mild
`hypertension. Even if treatment
`is effective in these pa-
`tients, it is undoubtedly less cost-effective than the treat-
`ment of more severe hypertension. From a policy point of
`view, the question is whether treatment of very mild hy-
`pertension is a good use of resources. The results of the
`British Medical Research Council Working Party study of
`the treatment of mild hypertension (90 to 109 mm Hg dia-
`stolic after three screenings) highlight this dilemma [7]. In
`this trial, stroke incidence was reduced by 60 percent in
`treated patients, but coronary events and mortality were
`not significantly different
`in the
`treated and placebo
`groups. The decrease in stroke rates in the treated group
`was equivalent to preventing one stroke annually for every
`850 patients treated.
`If treatment costs were $400 per
`year, each stroke prevented would cost $340,000. Is this a
`good value? Opinions will vary.
`
`COMMENTS
`
`is an extremely
`There is no question that hypertension
`important public health problem, both medically and eco-
`nomically. Medications are available that can safely and
`effectively lower blood pressure and reduce the risk of
`mortality and morbidity. The challenge to physicians, pa-
`tients, and policy makers alike is to use-or
`facilitate the
`use of-treatment
`in a manner that will balance expected
`health benefits and the cost burden to the individual pa-
`tient and to society.
`Opportunities do exist for reducing the costs of treat-
`ment with minimal or no sacrifice in the quality of care.
`Among these alternatives, two deserve our special atten-
`tion: limiting treatment to patients who are most likely to
`benefit from it and emphasizing the use of the least costly
`medications compatible with successful blood pressure
`control and good side-effect profiles. At a societal level,
`full insurance coverage for patients with moderate or se-
`vere hypertension is a goal that needs to be actively pur-
`sued as well.
`
`per year of treatment, potassium supplements $85 to
`$225, potassium-sparing
`diuretics about $100, beta
`blockers $200, and angiotensinase inhibitors $400 at cur-
`rent prices. Calcium channel blockers have not yet been
`approved by the Food and Drug Administration for use as
`antihypertensive drugs, but their costs would be even
`higher than angiotensinase
`inhibitors because of multiple
`daily dose requirements.
`A reasonable cost-saving strategy would be to start
`treatment with a diuretic and to proceed to more expen-
`sive medications only if side effects develop. Side effects
`sufficient to require discontinuation of medications occur
`in 20 to 30 percent of all patients [7,8]. A recent study that
`compared methyldopa, propranolol, and captopril found
`withdrawal rates due to adverse reactions of 20, 13, and 8
`percent, respectively [9]. This latter study, unfortunately,
`did not include a direct comparison with diuretics, the
`most commonly prescribed step-one drug in antihyperten-
`sive treatment. This cost savings, therefore, would proba-
`bly be successful in nearly 80 percent of the patients re-
`ceiving monotherapy. The same principles could be used
`to select the multidrug regimens, when they are required
`for adequate blood pressure control.
`A simplified example will illustrate the potential cost
`III, four first-
`savings of such a strategy. As seen in Table
`step treatment regimens could be envisioned for patients
`with hypertension. Assuming the adverse drug reaction
`rates reported by Croog et al [9] (and a relatively pessi-
`mistic 20 percent rate of severe side effects for diuretics
`alone), the cost per year of treatment would range from
`$125 under regimen A (generic diuretic plus supplemental
`potassium in 20 percent of patients) to $402 under regi-
`
`48
`
`December
`
`31, 1988
`
`The American
`
`Journal
`
`of Medicine
`
`Volume
`
`81 (suppl 8C)
`
`LOWER DRUG PRICES FOR CONSUMERS, LLC
`Exhibit 1030-4
`IPR2016-00379
`
`

`
`Organization,
`for
`hypertension.
`
`Inc.:
`
`who
`of physicians
`Survey
`New York: Gallup
`Organization,
`
`REFERENCES
`6.
`provide
`Inc.,
`
`1. Gallup
`care
`1988.
`2. Gallup Organization,
`are
`on medications.
`1988.
`reduced
`free care
`GA, et al: How
`EB, Brook RH, Goldberg
`Keeler
`hypertension
`in the Health
`Insurance
`Experiment.
`JAMA
`1985;
`254:
`1928-1931.
`Shulman WB, Martinez
`obstacle
`to hypertension
`76: 1105-l
`108.
`5. Weinstein
`MC, Stason WB: Cost-effectiveness
`to prevent
`or
`treat
`coronary
`heart
`disease.
`Health
`1985; 6: 41-63.
`
`3.
`
`4.
`
`Inc.: Survey
`New
`York:
`
`patients
`of hypertension
`Gallup
`Organization,
`
`who
`Inc.,
`
`B, Brogan
`therapy.
`
`cost
`D, et al: Financial
`Am
`J Public Health
`
`as an
`1988;
`
`interventions
`of
`Ann Rev Public
`
`SYMPOSIUM ON ANTIHYPERTENSIVE
`
`TREATMENT-STASON
`
`treatment,
`of awareness,
`the status
`and
`prevalence
`Hypertension
`the Sub-
`of
`Final
`report
`States.
`in
`the United
`and
`control
`on Definition
`and Prevalence
`of
`the Joint National
`committee
`on Detection,
`Evaluation,
`and Treatment
`of High
`Committee
`Blood
`Pressure,
`1984.
`United
`States
`Government
`Printing
`Office,
`1985;
`1-16.
`Medical
`Party:
`Working
`7. Medical
`Research
`Council
`of mild hypertension:
`Council
`Trial of
`treatment
`97-104.
`sults
`Br Med
`J 1985;
`291:
`JP, et al: Long-term
`Curb
`JD, Borhani
`NO, Blaszkowski
`lance
`for
`adverse
`effects
`of antihypertensive
`drugs.
`1985;
`253: 3263-3268.
`of antihyper-
`Croog
`SH, Levine
`S, Testa MA, et al: The effects
`tensive
`therapy
`on
`the quality
`of life. N Engl J Med 1986;
`314:
`1657-l
`664.
`
`8.
`
`9.
`
`Research
`principal
`re-
`
`surveil-
`JAMA
`
`December
`
`31, 1966
`
`The American
`
`Journal
`
`of Medicine
`
`Volume
`
`61
`
`(suppl
`
`6C)
`
`49
`
`LOWER DRUG PRICES FOR CONSUMERS, LLC
`Exhibit 1030-5
`IPR2016-00379