Advertisement

Financing a Cure for Diabetes in a Multipayer Environment

Open ArchivePublished:May 11, 2016DOI:https://doi.org/10.1016/j.jval.2016.03.1859

      Abstract

      Background

      Financing medical breakthroughs or cures is becoming increasingly challenging in the current fiscal environment.

      Objectives

      In this paper, we develop the precise conditions needed for a financing mechanism, HealthCoin, to work between a private payer and Medicare, to incentivize the former to invest in breakthrough therapies or cures in the US.

      Methods

      We illustrate the valuation of such a currency for a cure of Type 2 diabetes.

      Results

      We show that without a HealthCoin, a private payer does not invest in the cure, a small fraction of the patients live up to age 65, Medicare pays for the full price of the cure at age 65 and incurs net loss in returns over the elderly ages, and the manufacturer only sells cures for those who reach age 65. In contrast, a HealthCoin is feasible in this market, incentivizing the private payer to invest in the cure during the non-elderly ages and leading to Pareto improvements for all three stakeholders.

      Conclusions

      Discussions around the applicability of HealthCoin for breakthrough therapies on the horizon, such as gene therapies for blindness and hemophilia B, and the feasibility of instituting such payments through new legislations or demonstration projects could be of great value.

      Keywords

      Introduction

      There are growing concerns about the financing of technological breakthroughs in medicine that provide not just incremental health benefits but substantial leaps in patients’ health in the form of cures. The recent introduction of several new products for treating hepatitis C infections, which have been shown to cure 80% to 95% of hepatitis C infections, irrespective of age, highlights these issues [

      Loftus P. Senate Committee is investigating pricing of hepatitis C drug. Wall Street J July 11, 2014. Available from: http://online.wsj.com/articles/senate-finance-committee-is-investigating-pricing-of-hepatitis-c-drug-1405109206. [Accessed August 11, 2014].

      ]. The pending development of targeted treatments in cancer and emerging health technologies such as regenerative medicine and gene therapy indicates that the concerns around financing these technologies are likely to persist [
      • Mason C.
      • Dunnill P.
      The strong financial case for regenerative medicine and the regen industry.
      ].
      The primary question for these kinds of high value/curative treatments does not seem to be pricing per se, but the fact that health care systems have and will struggle to manage the short-term budget impact of treating large patient populations in a short amount of time. For example, recent analysis shows that the potential cost to Medicare of covering sofosbuvir (Solvaldi®) alone, a treatment and potential cure for hepatitis C infection, is a 3% to 8% increase in federal Part D outlays and Part D premiums, where Part D represents the prescription drug benefits plan paid for by US Medicare for the elderly and administrated by managed care plans [
      • Neuman T.
      • Hoadley J.
      • Cubanski J.
      The cost of a cure: Medicare’s role in treating hepatitis C.
      ]. Similar concerns are expressed for Medi-Cal, California’s Medicaid Program for low-income population about financing stem-cell therapies [

      Fulton BD, Felton MC, Pareja C, et al. Coverage, cost-control mechanism and financial risk-sharing alternative of high-cost healthcare technologies. 2009. Available from: 〈http://www.cirm.ca.gov/sites/default/files/files/about_cirm/UC-Berkeley%20final%20report%20to%20CIRM%2010.08.09.pdf〉. [Accessed October 18, 2014].

      ]. In recognition of the challenges that may be pending, recent discussions have highlighted the need to develop credit markets that will be able to amortize the costs of financing a cure over a longer time period, with or without risk sharing [

      Philipson T, von Eschenbach AC. Medical breakthroughs and credit market. Forbes July 9, 2014. Available from: http://www.forbes.com/sites/tomasphilipson/2014/07/09/medical-breakthroughs-and-credit-markets/. [Accessed August 2, 2014].

      ,
      • Gottlieb S.
      • Carino T.
      Establishing new payment provisions for the high cost of curing disease.
      ,
      • Brennan T.A.
      • Wilson J.M.
      The special case of gene therapy pricing.
      ,
      • Edlin R.
      • Hall P.
      • Wallner K.
      • McCabe C.
      Sharing risk between payer and provider by leasing health technologies: an affordable and effective reimbursement strategy for innovating technology.
      ].
      Although access to capital is an important consideration in solving for the budget impact question, it alone is likely not enough to solve the problem of financing of cures in the United States. This is because of the fact that in a multipayer system, like that in the United States, a cure is valued by all private and public payers, but no one payer can exclude others from appropriating some of the value associated with a cure. Consequently, because no one payer can appropriate the full value of providing treatment, none may be willing to bear the full costs of the cure—a classic “free-rider problem” [
      • Samuelson P.A.
      The theory of public expenditure.
      ], which leads to the underprovision of cures and underinvestment in innovation. In a recent commentary, we proposed a solution to this problem by recommending the development of a tradable new currency, HealthCoin, that would convert the incremental consequences produced (but not the upfront costs incurred) by curative treatments to a common numeraire, such as health stock units, life-years equivalents, or some other metric [
      • Basu A.
      Financing cures in the United States.
      ].
      As in the case of any capital stock, HealthCoins would depreciate over time. At any point in time, HealthCoins could be traded with real dollars in the marketplace among private and public payers. Because Medicare is the largest public insurance system in the United States covering elderly individuals and would often be the recipient of a substantial part of the value generated because of investment in cures in early life, it would be expected that the US government would back this currency with necessary assurances. Medicare or its managed care plans would also buy the present value of the HealthCoins as patients become eligible for Medicare, much in line with the spirit of social impact bonds that are being used across the world to pay for investments that produce returns to public sector [

      Azemati H, Belinsky M, Gillette R, et al. Social impact bonds: lessons learned so far. Available from: http://www.frbsf.org/community-development/files/social-impact-bonds-lessons-learned.pdf. [Accessed August 11, 2014].

      ].
      In this article, we illustrate how an evidence base can be developed to issue and appropriately value such HealthCoins from multiple perspectives. Specifically, we study the hypothetical case in which a cure for type 2 diabetes becomes available. We estimate what impacts on patient’s health and health care costs such a cure would produce and study how these impacts would be distributed across different stakeholders and over time. Using these estimates, we then illustrate how issuance of HealthCoins could be informed.

      Type 2 diabetes

      Type 2 diabetes or diabetes mellitus is a collection of metabolic diseases that results in an abnormally high level of glucose in the blood. Diabetes is a chronic lifelong condition, and it also increases the risk of cardiovascular disease, kidney disease, blindness due to diabetic retinopathy, and so on. Several burden-of-disease estimates for type 2 diabetes are available [
      American Diabetes Association
      Economic costs of diabetes in the U.S. in 2007.
      ,
      • Dall T.M.
      • Zhang Y.
      • Chen Y.J.
      • et al.
      The economic burden of diabetes.
      ]. The annual medical and indirect costs associated with the prevalent cohort of diabetes are approximately $218 billion in the United States alone [

      Azemati H, Belinsky M, Gillette R, et al. Social impact bonds: lessons learned so far. Available from: http://www.frbsf.org/community-development/files/social-impact-bonds-lessons-learned.pdf. [Accessed August 11, 2014].

      ]. Therefore, a cure for the disease would certainly be a welcome technological breakthrough. It is likely that the cost of such a cure would be significant—several orders of magnitude greater than the cost we have observed for hepatitis C infections. However, given current concerns about health care financing, it is unclear whether there are sufficient economic incentives in place to encourage the development of such a treatment.
      To understand the incentives in the current market and in a market with a HealthCoin currency in place, and whether such incentives are sustainable over a longer period, we focus on the incident population of type 2 diabetes. In the United States, there are 1.34 million incident annual cases, aged 18 to 64 years, of type 2 diabetes [

      Centers for Disease Control and Prevention. Diabetes public health resource. Available from: http://www.cdc.gov/diabetes/statistics/age/fig1.htm. [Accessed September 4, 2015].

      ]. They form our target population for this exercise, although their health care burden will be assessed throughout their lives.

      Methods

      Theoretical Setup

      Our analyses reflect a problem of financing for a cure for type 2 diabetes. Two assumptions are made about this cure. First, we assume that the cure applies to all ages of patients; that is, its effectiveness, which is set at 100%, does not vary over age. Second, we assume that the effect of the cure is persistent; that is, a patient with diabetes, when cured, does not redevelop diabetes in his or her lifetime. We discuss the implications of these assumptions in the last section. We approach the problem of financing from two sides: a private payer responsible for coverage of health care for nonelderly adults (18–64-year-olds) and a public payer (e.g., Medicare) responsible for coverage of health care for the elderly (65+-year-olds). Throughout this article, we assume that the private payer’s decision making follows maximizing the net monetary benefits (NMBs) [
      • Stinnett A.A.
      • Mullahy J.
      Net health benefits: a new framework for the analysis of uncertainty in cost-effectiveness analysis.
      ] criterion from its own perspective. Unlike US Medicare, as described below, a private payer can restrict coverage for a drug/ medical technology on the basis of its costs. Because an NMB criterion requires the use of a threshold to covert health gains (in our case, life-years) to dollar amount, we use a broad range of threshold values to illustrate our points. Discussion about the threshold is added in the last section. To embed the analysis in the US context, we assume that decision making by Medicare follows the law laid out before it, which is to provide coverage to all health care interventions that are effective and necessary [
      • Neumann P.J.
      • Chambers J.D.
      Medicare’s enduring struggle to define “reasonable and necessary” care.
      ]. Specifically, Medicare is prohibited by Congress to look at costs in making coverage decisions, and therefore, it cannot reject any expensive therapy as long as it is deemed to be effective [
      • Chambers J.D.
      • Neumann P.J.
      • Buxton M.J.
      Does Medicare have an implicit cost-effectiveness threshold?.
      ]. However, we do study the welfare implications for Medicare’s coverage decision for the elderly population using the NMB criterion from a Medicare’s perspective, even though Medicare’s decision does not follow the NMB criterion.
      Note that throughout the article, our use of the term “welfare” is meant to represent the value accruing to a generic health care sector based on the NMB criteria. However, this value is based on some societal willingness to pay for an added year of life.
      The goal of this analysis was to try to value a HealthCoin that Medicare would be willing to pay a private payer for a beneficiary who is transitioning to Medicare at the age of 65 years, if the private payer had invested in a cure for diabetes before age 65 years. The market for such a HealthCoin, as a currency of exchange between Medicare (public) and the private payer, would clear only if the value of such a HealthCoin for a 65-year-old “cured” diabetes patient meets the threshold NMB criteria from both the Medicare and the private payer perspectives. Although Medicare decision making does not follow NMB criterion per se, Medicare is still expected to accept dominant strategies where it receives the same or additional amount of benefits but at lower costs. We develop the necessary threshold from both payers’ perspectives and study the implications. If the market for a HealthCoin does not clear, we study the implications for the private and the public payers and the manufacturer. All our calculation on prices and HealthCoin values reflect the present value calculation based on 3% discount rate, where present reflects that point of cure for incident disease for 18 to 64-year-olds and also the point when the private payer pays for the cure.

      Price of a Cure for Diabetes

      From a manufacturer’s point of view, pricing for a cure of diabetes is based on the average incremental net monetary benefits (INMBs) in the entire population (inclusive of all ages) that the cure can produce [
      • Berndt E.R.
      • McGuire T.
      • Newhouse J.P.
      A primer on the economics of prescription pharmaceutical pricing in health insurance markets.
      ]. Because a treatment labeled as a “cure” or “breakthrough” may carry some additional premium to it, we vary the threshold to value life-years over a range of $100,000/life-years to $500,000/life-year. The higher end of the range reflects the cost-effectiveness of some of the newer oncolytic drugs and other breakthroughs that have been covered by Medicare [
      • Neumann P.J.
      • Rosen A.B.
      • Weinstein M.C.
      Medicare and cost-effectiveness analysis.
      ]. This does not signify that Medicare is making decisions on the basis of such threshold, but reflects that such valuations are acceptable for Medicare’s point of view.
      For a monopoly manufacturer enjoying patent protection, price for such a breakthrough would certainly be higher than the marginal cost of the product. Under the assumption that the marginal cost is negligible, the average NMBs reflect the average per beneficiary surplus in the entire population, not just the nonelderly [
      • Jena A.B.
      • Philipson T.J.
      Cost-effectiveness analysis and innovation.
      ]. Therefore, equating the price with the average surplus corresponds to the efficient provision of the cure from a dynamic sense of efficiency [
      • Berndt E.R.
      • McGuire T.
      • Newhouse J.P.
      A primer on the economics of prescription pharmaceutical pricing in health insurance markets.
      ]. In this population, the average surplus corresponds to total incremental value, expressed in INMBs, generated because of the cure for diabetes for the entire incident population of 18- to 85-year-olds (denoted by EINBALL, Lifetime). For the purpose of this analysis, we assume that both the private payer and the Medicare face the same price. Let it be denoted by PCure (= EINBALL, Lifetime).

      Medicare Perspective

      As explained above, Medicare is assumed to pay for the cure, but the welfare implication of that decision is expressed in NMB terms. From the Medicare perspective, let the total incremental value, expressed in INMBs, of a cure for diabetes for a 65-year-old would depend on the life expectancy gain for the 65-year-old and the net saving in health care costs that would happen over the patient’s lifetime due to the cure. Let this be denoted as INMB65,Medicare. Note that the net saving in health care costs produced by the cure would include the potential increase in health care consumption due to increased survival from the cure (a.k.a. future costs [
      • Meltzer D.
      Accounting for future costs in medical cost-effectiveness analysis.
      ]).
      The key aspect to note is that any HealthCoin that Medicare may be willing to pay the private payer for a cured diabetes patient at age 65 years, in order to incentivize the private payer to invest in the cure earlier than age 65 years, must be valued less than PCure. In such a scenario, Medicare would save costs without sacrificing any health benefits. Such a transaction would always be welfare enhancing from Medicare’s perspective irrespective of whether PCure was less or greater than INMB65,Medicare. Even when PCure > INMB65,Medicare and, by paying for the cure, Medicare is incurring a welfare loss, transaction with a HealthCoin valued such that PCure > HealthCoin > INMB65,Medicare would enhance welfare by minimizing the welfare loss but by not eliminating it.
      In a nutshell, Medicare would be willing to pay for a HealthCoin, if it is valued such that HealthCoinMedicarePCure.

      Private Payer Perspective

      The value of a HealthCoin that clears the market between Medicare and a private payer not only has to satisfy the conditions described above from the Medicare’s perspective but also needs to meet certain requirements from the private payer perspective. Specifically, because a private payer would invest in a cure for nonelderly (NE) patients at the age of diagnosis for diabetes, the life time INMBs of a cure (INMBNE,Lifetime) would be different for patients diagnosed at different (nonelderly) ages. The weighted average INMBNE, Lifetime across all nonelderly adults would produce an expected lifetime value of cure for diabetes for all nonelderly adults, or EINMBNE, Lifetime. The portion of these benefits that would accrue to the private payer, EINMBNE, Private, (i.e., until before they turn 65 years, at various time in the future) could be carved out and compared with the price of cure to establish the threshold for the value of a HealthCoin from the private payer perspective.
      That is if PCure < EINMBNE, Private, then a private payer would have an incentive to invest in a cure for diabetes among nonelderly patients by directly purchasing the cure from the market. If PCure > EINMBNE, Private, then the private payer will not have an incentive to purchase cures from the market unless it accrues back the HealthCoin from Medicare as patients turn 65 years. Because only a fraction of patients (say, S) receiving the cure in nonelderly ages would survive to age 65 years, from a private payer’s perspective the HealthCoin should be valued such that
      HealthCoin(PCureEINMBNE,Private)/S

      Valuation of a HealthCoin for Cure for Diabetes at Age 65 Years

      From the above discussions, it is apparent that for the market for HealthCoins to clear between the private payer and Medicare, it must satisfy the following:
      PCureHealthCoin(PCureEINMBNE,Private)/S
      This range of feasible values for a HealthCoin would vary with the threshold used to value life- years. Keeping S constant, if increasing the threshold life-year value increases PCure at a faster rate than EINMBNE, Private, then the range decreases. This is because, since S ≤ 1, the right-hand bound increases faster than the left-hand bound in the above equation. Intuitively, it indicates that life-year gains with cure are more concentrated in elderly years and the gains from cure during the nonelderly ages are small. Hence, as the value of these life-years increases, the private payer demands a proportionately higher HealthCoin value from the public paper with respect to the price of the cure.

      Empirical Analysis for Valuation of a HealthCoin for the Cure of Type 2 Diabetes: Methods

      Our analysis consists of four steps. All calculations are based on net present value estimates with a 3% discount rate.
      • 1.
        We develop a probabilistic model to estimate the health care costs and life-years associated with incident diabetes population aged 18 to 85 years. We obtain age-specific attributable costs of diabetes using Medical Expenditure Panel Survey data. Next, we calculate the life tables for the general population if there were no diabetes by subtracting age-specific diabetes-specific mortality rates [
        • Mulnier H.E.
        • Seaman H.E.
        • Raleigh V.S.
        • et al.
        Mortality in people with type 2 diabetes in the UK.
        ] from overall mortality rates [
        • Arias E.
        United States life tables 2010.
        ], weighted by the age-specific prevalence of diabetes. The Centers for Disease Control and Prevention did similar calculations for older life tables [
        • Arias E.
        • Heron M.
        • Tejada-Vera B.
        United States life tables eliminating certain causes of death 1999-2001.
        ]. The difference in survival between the general and diabetes-free life tables was used to obtain estimates of diabetes-attributable life-years. Using these methods, we develop the estimates for the net present value (at 3% discount) of expected life-years and health care costs of one incident population of 18- to 85-year-old patients with diabetes and the counterfactual estimates had they been instantaneously cured of diabetes over the next 50 years. Fifty years was chosen to capture benefits over the lifetime of the patients with 3% discount rate. Moreover, it is a round number that would also allow the 18-year-olds reach Medicare. Therefore, we compare the expected lifetime costs and benefits of the incident cohort of patients with diabetes under status quo and the expected costs and benefits for the same cohort if they were cured of diabetes at incidence. We convert these differences to incremental net benefits using a range of threshold values to obtain estimates of EINBALL, Lifetime = PCure.
      • 2.
        We then look only at the incident population of 18- to 64-year-old patients and compute similar estimates as in (1), that is, EINBNE, Lifetime. We decompose these estimates to those that accrue over the nonelderly years (EINBNE, Private) versus those that accrue over the elderly years. We estimate the proportion of this cohort (S) whose members would live to age 65 years if they are cured of diabetes at incidence. We then propagated the incident cohort of patient with diabetes over time but applying age-specific and general-population (with no diabetes)-specific mortality rates to obtain the proportion of “cured” patient who would reach 65 years of age. We compare these returns to the price of cure to assess whether the private payer would have any incentive to pay for the cure.
      • 3.
        Next, on the basis of theoretical arguments discussed above, we estimate the potential range of values for HealthCoin that would clear the market for the HealthCoin between Medicare and the private payer and incentivize the private payer to invest in the cure among the nonelderly population. These valuations were based on net present values of NMBs at 3% discount rates.
      • 4.
        Using the midpoint of this range of value of HealthCoin, we illustrate the consequences, from the perspectives of Medicare, private payer, and the manufacturer, of a market without such a HealthCoin (i.e., current scenario) versus where such a financial instrument exists.
      Discussion follows in the next section.

      Results

      Table 1 presents estimates of EINBALL, Lifetime over the range of threshold values. At $100K/life-year, PCure is estimated to be $229,183 per patient. At $500K/life-year, it is $1,137,183.
      Table 1Expected per patient returns from curing diabetes in the entire incident cohort (18–85-y-olds) in 2014 over next 50 y
      With diabetesWithout diabetesDifference: (Without – With)
      E(LYs)E(Cost)E(LYs)E(Cost)ΔE(LYs)ΔE(Cost)
      8.64$117,26310.92$115,0802.27$2,183
      Incremental NMB:
      @$100K/LY$229,183
      @$200K/LY$456,183
      @$300K/LY$683,183
      @$400K/LY$910,183
      @$500K/LY$1,137,183
      Note. NMB = ΔE(LYs) × threshold − ΔE(Cost); Discount rate: 3%.
      LY, life-year; NMB, net monetary benefit.
      Table 2 presents estimates of EINMBNE, Lifetime and EINMBNE, Private (detailed composition of these estimates can be found in Appendix Table 1). It also shows that at any threshold, EINMBNE, Private < PCure, implying that the private payer would not have the incentive to invest in the cure among the nonelderly. Similarly, even though Medicare would always pay for this cure, the return it gets from this investment among 65-year-olds during the elderly years are less than the price of cure, resulting in welfare loss.
      Table 2Returns to private and public payers in curing diabetes among 18–64-y-old incident cohort at the age of incidence by a private payer
      Per patient returns during the entire 50 y
      Difference: (Without – With)ThresholdINMB (EINMBNE, Lifetime)Price (Table 1)Net value
      ΔE(LYs)ΔE(Cost)
      2.49−$4287@$100K/LY$253,287$229,183$24,104
      2.49−$4287@$200K/LY$502,287$456,183$46,104
      2.49−$4287@$300K/LY$751,287$683,183$68,104
      2.49−$4287@$400K/LY$1,000,287$910,183$90,104
      2.49−$4287@$500K/LY$1,249,287$1,137,183$112,104
      Per Patient returns during the nonelderly period
      Difference: (Without – With)ThresholdINMB (EINMBNE, Private)Price (Table 1)Net value
      ΔE(LYs)ΔE(Cost)
      0.72−$30,198@$100K/LY$102,879$229,183−$126,304
      0.72−$30,198@$200K/LY$175,559$456,183−$280,624
      0.72−$30,198@$300K/LY$248,240$683,183−$434,943
      0.72−$30,198@$400K/LY$320,391$910,183−$589,792
      0.72−$30,198@$500K/LY$393,601$1,137,183−$743,582
      Per patient returns during elderly period for those turning 65-y-old
      Difference: (Without – With)ThresholdINMBPrice (Table 1)Net value
      ΔE(LYs)ΔE(Cost)
      1.61$1659@$100K/LY$159,341$229,183−$69,842
      1.61$1659@$200K/LY$320,341$456,183−$135,842
      1.61$1659@$300K/LY$481,341$683,183−$201,842
      1.61$1659@$400K/LY$642,341$910,183−$267,842
      1.61$1659@$500K/LY$803,341$1,137,183−$333,842
      INMB, incremental net monetary benefit; LY, life-year.
      We estimate S, the proportion of nonelderly incident diabetes cohort members who would reach age 65 years if their diabetes were cured, to be 0.654 (876,000/134 million).
      Because (PCure − EINMBNE, Private)/S is less than PCure (Table 3), it indicates that a HealthCoin valuation is feasible. However, as Table 2 also shows, as the threshold value for a life-year increases, the price of cure increases at a faster rate than does EINMBNE, Private. This would indicate that at a higher value of the threshold, the range of feasible values for HealthCoin decreases. Table 3 presents the range of values between PCure and (PCure − EINMBNE, Private)/S that would suffice for a HealthCoin for a cure for diabetes at age 65 years. The midpoint of this range is chosen for further illustration. For example, at the threshold value of $100K/life-year, the HealthCoin would be valued at $211,154.
      Table 3HealthCoin valuation
      ThresholdPCure ($)EINMBNE, Private ($)(PCureEINMBNE, Private)/S ($)HealthCoin values satisfying Equation 1 ($)Midrange value for HealthCoin ($)
      @$100K/LY229,183102,879193,125193,125–229,183211,154
      @$200K/LY456,183175,559429,089429,089–456,183442,636
      @$300K/LY683,183248,240665,050665,050–683,183674,117
      @$400K/LY910,183320,391901,823901,823–910,183906,003
      @$500K/LY1,137,183393,6011,136,9761,136,976–1,137,1831,137,079
      Note. Equation 1: PCure ≥ HealthCoin ≥ (PCure − EINMBNE, Private)/S; S was estimated as follows: S: Proportion of nonelderly incident patients with diabetes who would survive to age 65 y if their diabetes was cured = 0.876.
      LY, life-year.
      Table 4 presents the incentives and returns to the private payer, Medicare, and the manufacturer without and with the HealthCoin financial instrument. For illustration, we use the lowest threshold of $100K/life-year. Estimates from other thresholds are available in Appendix Table 2. Without the HealthCoin, the private payer will not have the incentive to pay for cure for the 1.34 million incident nonelderly diabetes cases (as shown in Table 2). Without the cure, 507,000 of these patients will reach age 65 years. At that point, Medicare will pay for the cure at $229,183, incurring costs of $116 billion, which accrues to the manufacturer as revenue. The returns to this investment, over the remaining life, will be $81 billion, producing a welfare loss.
      Table 4Population returns to different stakeholders with and without a HealthCoin at the threshold of $100K/LY
      Without HealthCoin
      No. of 18–64-y-old incident diabetes casesCure applied at incidencePrivate payer costs of cureTotal population return for private payerNo. of patients who reached age 65 y with diabetesCure applied at age 65 yPublic payer costsTotal population return for public payer
      1.34 million000507000507000507,000 × $229,183= $116 billion507,000 × $159,341=$81 billion
      With HealthCoin
      No. of 18–64-y-old incident diabetes casesCure applied at incidencePrivate payer costs of cureTotal population return for private payerNo. of patients who reached age 65 y with diabetes curedCure applied at age 65 yPublic payer costsTotal population return for public payer
      1.34 million1.34 million1.34 million ×$229,183– 876,000 ×$211,154= $122 billion1.34 million × $102,879= $138 billion876,0000876,000 ×$211,154= $185 billion507,000 ×$159,341 + (876,000 – 507,000) ×$328,043
      From Table 1: The expected NMB obtained for cured patients during elderly ages: 4.20 LY × $100K/LY − $91957 = $328,043.
      = $202 billion
      LY, life-year; NMB, net monetary benefit.
      low asterisk From Table 1: The expected NMB obtained for cured patients during elderly ages: 4.20 LY × $100K/LY − $91957 = $328,043.
      In contrast, when a HealthCoin instrument is present, the net costs to the private payers for paying for the cure among nonelderly patients is $122 billion (Table 4), as the payer is going to recuperate the HealthCoin, valued at $211,154, for the 876,000 cured patients who reach the age of 65 years. The private payer is also going to receive benefits of $102,879 per cured patient during the nonelderly years (Table 2), amounting to $138 billion. Therefore, the net value to the payer in paying for the cure is positive. The manufacturer would sell 1.34 million units of cure, amassing a revenue of $307 billion. From the Medicare’s perspective, it would now pay the value of the HealthCoin for 876,000 patients, totaling $185 billion. However, in return, it would generate a total value of $202 billion, which comprises of the $81 billion generated among 507,000 patients who would have been cured at age 65 years even without the HealthCoin and the additional 369,000 patients who reach age 65 years with their diabetes cured and live their expected life expectancies.
      As the threshold value of the life-year increases (Appendix Table 2), we found that the HealthCoin increases the net returns to the public payer while the net returns to the private payer decreases and at $500K/life-year, the private payer becomes indifferent to investing even with the HealthCoin.

      Discussion

      Breakthroughs in health care often come at high prices and large budget impacts for payers. A disaggregated system of health payers, as that found in the United States, naturally leads to underinvestment in these cures, even when a robust credit market exists for financing. This is because even though the price of cure would likely reflect the lifelong benefits of a cure, no one payer would recuperate all those benefits and, therefore, does not have the incentive to pay for the cure. New financial instruments are needed to solve this underinvestment. One such concept, HealthCoin, devised in the nature of a social impact bond, was proposed to solve this underinvestment [
      • Samuelson P.A.
      The theory of public expenditure.
      ].
      In this article, we illustrate the problem of underinvestment for a cure for type 2 diabetes by a private payer and show how a HealthCoin currency could be devised by Medicare and valued, so that it can incentivize a private payer to invest and make all stakeholders better off. In sum, we showed that without a HealthCoin, the private payer does not invest, Medicare pays for cure at age 65 years resulting in net welfare loss, and the manufacturer sells 507,000 units annually. With the HealthCoin, the private payer pays for the cure as it leads to net gain, Medicare pays HealthCoin and achieves net gains, and the manufacturer sells 1.34 million units annually. Thus, HealthCoin could lead to Pareto improvement for all three stakeholders.
      This stylized example has several limitations. It assumes that the cure is permanent and that it applies equally to all ages. Variations to the effectiveness of the cure could easily be incorporated in such a framework. For example, if only a constant fraction of patients with diabetes at any age would be cured, but the insurer and providers cannot identify who they are, the HealthCoin implications presented here would carry through, because all estimates, including the price and the HealthCoin valuation, would likely adjust downward to reflect this ambiguity. If the insurer and providers can identify those who would benefit from the cure, the same per patient numbers presented here would apply but to a smaller population. Limited durability of the cure effect would require more elaborate modeling on the probability of relapse and the costs of care for these patients on relapse. Because of the plethora of possible variations, which one can think about the attributes of such breakthrough therapies, these calculations are better served when applied to a real-world breakthrough therapy.
      Any exchange of a health currency would require some way to monetize health. We have illustrated this monetization over a range of values and showed the feasibility of the HealthCoin. The challenge, however, will remain on the private and public payers negotiating to agree on a valuation of health unit. Any such deal, and the failure to reach a deal, would implicitly highlight the bounds on the value stakeholders place on a unit of health, which, in turn, could lead to broader discussions in this area.
      We illustrated the HealthCoin valuation methods using the incident cohort of patients with diabetes to show that such a financing mechanism would work in an equilibrium going forward. Curing the prevalent cohort, which could be viewed as a one-time event, may not be addressed with HealthCoin, because most patients are elderly. Financing for these people could be directly addressed through accessing credit markets and certain discounts to the payers.
      In the United States, the feasibility of introducing such a health currency primarily relies on Congress, because new legislation would be required to allow Medicare the capability of issuing such a currency. From Medicare’s perspective, as we show in our illustrated example, issuing a HealthCoin could be welfare enhancing especially for cures that are in demand in nonelderly years and those that will continue to produce benefits through the lives of the patients and especially after age 65 years. Once such a HealthCoin is issued, such currency could readily be traded between the private payers so that all of them are incentivized to invest. Additional empirical work is needed to demonstrate such a possibility. Even more important will be to explore valuation of a HealthCoin around financing a realistic and upcoming breakthrough therapy and, the consequences of the failure to do so, to convince policymakers that the time for innovative financing models in health is already here. Such work can also pave the way for demonstration projects where the impact of such HealthCoin availability could be tested with Medicare waivers under the Center for Medicare & Medicaid Innovation demonstration projects. We hope that this work can foster more analyses in this line to develop a robust body of evidence around the importance of the concept of a HealthCoin.

      Acknowledgments

      We thank three anonymous reviewers for their comments on an earlier version of this article.
      Source of financial support: This study was conducted using funds provided by Global Health & Value Innovation Center, Pfizer, Inc., and a consortium of 12 biomedical life sciences companies to the University of Washington.

      Appendix

      Appendix Table 1, Appendix Table 2
      Appendix Table 1Expected per patient lifetime returns from curing diabetes in the entire incident cohort 18–64-y-olds in 2014 over next 50 y
      Per patient returns during the entire 50 y
      With diabetesWithout diabetesDifference: (Without – With)
      E(Life-years)E(Cost)E(Life-years)E(Cost)ΔE(Life-years)ΔE(Cost)
      10.09$125,87912.58$121,5912.49−$4287
      Per patient returns during the nonelderly period
      With diabetesWithout diabetesDifference: (Without – With)
      E(Life-years)E(Cost)E(Life-years)E(Cost)ΔE(Life-years)ΔE(Cost)
      9.11$91,6899.83$61,4900.72−$30,198
      Per patient returns during elderly period for those turning 65-y-old
      With diabetesWithout diabetesDifference: (Without – With)
      E(Life-years)E(Cost)E(Life-years)E(Cost)ΔE(Life-years)ΔE(Cost)
      2.59$90,2984.20$91,9571.61$1659
      Appendix Table 2Population annual returns to different stakeholders with and without a HealthCoin at different threshold values
      Threshold valueWithout a HealthCoinWith a HealthCoin
      Private payer net returnsPublic payer net returnsUnit sales for manufacturerPrivate payer net returnsPublic payer net returnsUnit sales for manufacturer
      $100K/LY0−$35 billion507,000$16 billion$17 billion1.34 million
      $200K/LY0−$162 billion507,000$12 billion$51 billion1.34 million
      $300K/LY0−$244 billion507,000$8 billion$85 billion1.34 million
      $400K/LY0−$326 billion507,000$3 billion$118 billion1.34 million
      $500K/LY0−$407 billion507,000$0 billion$152 billion1.34 million

      References

      1. Loftus P. Senate Committee is investigating pricing of hepatitis C drug. Wall Street J July 11, 2014. Available from: http://online.wsj.com/articles/senate-finance-committee-is-investigating-pricing-of-hepatitis-c-drug-1405109206. [Accessed August 11, 2014].

        • Mason C.
        • Dunnill P.
        The strong financial case for regenerative medicine and the regen industry.
        Regen Med. 2008; 3: 351-363
        • Neuman T.
        • Hoadley J.
        • Cubanski J.
        The cost of a cure: Medicare’s role in treating hepatitis C.
        Health Affairs Blog. June 5, 2014; ([Accessed October 18, 2014])
      2. Fulton BD, Felton MC, Pareja C, et al. Coverage, cost-control mechanism and financial risk-sharing alternative of high-cost healthcare technologies. 2009. Available from: 〈http://www.cirm.ca.gov/sites/default/files/files/about_cirm/UC-Berkeley%20final%20report%20to%20CIRM%2010.08.09.pdf〉. [Accessed October 18, 2014].

      3. Philipson T, von Eschenbach AC. Medical breakthroughs and credit market. Forbes July 9, 2014. Available from: http://www.forbes.com/sites/tomasphilipson/2014/07/09/medical-breakthroughs-and-credit-markets/. [Accessed August 2, 2014].

        • Gottlieb S.
        • Carino T.
        Establishing new payment provisions for the high cost of curing disease.
        AEI Res. July 2014; ([Accessed August 2, 2014])
        • Brennan T.A.
        • Wilson J.M.
        The special case of gene therapy pricing.
        Nat Biotechnol. 2014; 32: 874-876
        • Edlin R.
        • Hall P.
        • Wallner K.
        • McCabe C.
        Sharing risk between payer and provider by leasing health technologies: an affordable and effective reimbursement strategy for innovating technology.
        Value Health. 2014; 17: 438-444
        • Samuelson P.A.
        The theory of public expenditure.
        Rev Econ Stat. 1954; 36: 386-389
        • Basu A.
        Financing cures in the United States.
        Exp Rev Pharmacoecon Outcomes Res. 2015; 15: 1-4
      4. Azemati H, Belinsky M, Gillette R, et al. Social impact bonds: lessons learned so far. Available from: http://www.frbsf.org/community-development/files/social-impact-bonds-lessons-learned.pdf. [Accessed August 11, 2014].

        • American Diabetes Association
        Economic costs of diabetes in the U.S. in 2007.
        Diabetes Care. 2008; 31 (576–15)
        • Dall T.M.
        • Zhang Y.
        • Chen Y.J.
        • et al.
        The economic burden of diabetes.
        Health Aff. 2010; 29: 297-303
      5. Centers for Disease Control and Prevention. Diabetes public health resource. Available from: http://www.cdc.gov/diabetes/statistics/age/fig1.htm. [Accessed September 4, 2015].

        • Stinnett A.A.
        • Mullahy J.
        Net health benefits: a new framework for the analysis of uncertainty in cost-effectiveness analysis.
        Med Decis Making. 1998; 18: S68-S80
        • Neumann P.J.
        • Chambers J.D.
        Medicare’s enduring struggle to define “reasonable and necessary” care.
        N Engl J Med. 2012; 367: 1775-1777
        • Chambers J.D.
        • Neumann P.J.
        • Buxton M.J.
        Does Medicare have an implicit cost-effectiveness threshold?.
        Med Decis Making. 2010; 30: E14-E27
        • Berndt E.R.
        • McGuire T.
        • Newhouse J.P.
        A primer on the economics of prescription pharmaceutical pricing in health insurance markets.
        Forum Health Econ Policy. 2011; 14: 1-28
        • Neumann P.J.
        • Rosen A.B.
        • Weinstein M.C.
        Medicare and cost-effectiveness analysis.
        N Engl J Med. 2005; 353: 1516-1522
        • Jena A.B.
        • Philipson T.J.
        Cost-effectiveness analysis and innovation.
        J Health Econ. 2008; 27: 1224-1236
        • Meltzer D.
        Accounting for future costs in medical cost-effectiveness analysis.
        J Health Econ. 1997; 16: 33-64
        • Mulnier H.E.
        • Seaman H.E.
        • Raleigh V.S.
        • et al.
        Mortality in people with type 2 diabetes in the UK.
        Diab Med. 2006; 23: 516-521
        • Arias E.
        United States life tables 2010.
        Natl Vital Stat Rep. 2014; 63: 1-63
        • Arias E.
        • Heron M.
        • Tejada-Vera B.
        United States life tables eliminating certain causes of death 1999-2001.
        Natl Vital Stat Rep. 2013; 61: 1-129