INTEGRATING HEALTH INNOVATION POLICY.

• Author: Sachs, Rachel E.

Table of Contents I. INTRODUCTION 58 II. CONCEPTUALIZING FRAGMENTATION IN HEALTH LAW 63 A. Defining the Scope of Fragmentation 63 B. Assessing the Impacts of Fragmentation 67 III. FRAGMENTATION OVER TIME 69 A. Examples of Time-Based Fragmentation 70 1. Fragmentation in Insurance Structure 70 2. Time-Based Fragmentation's Impact on Access to Treatments 73 B. Time-Based Fragmentation's Effects on Innovation Policy 76 IV. FRAGMENTATION BY BENEFIT STRUCTURE 78 A. Examples of Benefit Structure Fragmentation 78 1. Fragmentation in Medicare Design 79 2. Fragmentation in Financing by Private Insurers 80 B. Benefit Structure Fragmentation's Effects on Innovation Policy 81 V. FRAGMENTATION IN POLICYMAKING 84 A. Examples of Policymaking Fragmentation 85 1. Fragmentation in Administrative Agencies 86 2. Fragmentation in Congressional Committees 91 B. Policymaking Fragmentation's Effects on Innovation Policy 93 VI. IDENTIFYING OPPORTUNITIES FOR INTEGRATION 97 A. Framing the Approach to Integration 98 B. Articulating Potential Solutions 102 1. Systemic Integration-Encouraging Reforms 103 2. Narrowly Tailored Integration-Encouraging Reforms 105 a. Fragmentation Over Time 106 b. Fragmentation by Benefit Structure 108 c. Fragmentation by Policymaker 110 VII. CONCLUSION 113 I. INTRODUCTION

In the last few years, scientists have brought to market a series of incredible new pharmaceutical products for once-untreatable illnesses. These new gene therapies have the ability to alter a patient's own genetic material to help treat their disease in a single session, rather than through a drug that a patient may take for the rest of their life. Already, patients with rare inherited blindness or children with spinal muscular atrophy are eligible to receive these newly approved pharmaceuticals. (1) One-time treatments for devastating conditions like hemophilia, sickle-cell disease, and other inherited conditions are expected to be approved in the next few years. (2)

Although these new products represent genuine scientific breakthroughs, they come with a price tag to match. The first such one-time product to come to market, Luxturna, treats a small number of patients with a particular genetic form of blindness for $850,000. (3) The second approved product, Zolgensma, a one-time treatment for children with spinal muscular atrophy, is the most expensive drug ever marketed, with a list price of $2.1 million. (4) These prices represent floors, not ceilings--companies are already publicly suggesting that future gene therapy products will launch at even higher prices. (5)

Several features of these new treatments give insurers pause as they decide whether and under what circumstances to provide coverage for these pharmaceuticals. First, the unpredictability of these expenditures, given the rarity of some of these conditions, means that some insurers will have patients who are eligible to receive these treatments, but others will not. (6) As a result, insurers face actuarial risk problems in planning their expenditures. Second, the new treatments' high price tags can have a material impact on insurer spending, affecting budgeting as well as increasing patients' costs in the future. Third, particularly for state Medicaid programs, money spent on these pharmaceuticals is money that must be found elsewhere in the budget--money that cannot be spent on treatments for other patients, or on education or infrastructure. (7)

But another, more uniquely American dynamic also threatens patient access to these new treatments. In the United States, patients rotate on and off of different insurance plans frequently, a phenomenon referred to in the literature as "churn." (8) As a result, an insurer knows that in a few months' or years' time, their patients with costly rare diseases may enroll in another insurer's plan or may lose their coverage entirely. As a result, the insurer does not want to pay millions of dollars even for a highly cost-effective one-time treatment today, if the health benefits of that treatment will accrue to a future insurer. (9) Insurers therefore have incentives to make it difficult for patients to access these new medications.

This fragmentation of health insurance, in which patients may move in and out of private insurance, Medicaid eligibility, and uninsurance before becoming eligible for Medicare, is just one example of the ways in which our health care system is fragmented. Patients may be treated by a range of specialists who coordinate poorly, if at all, meaning that patients may be subject to duplicative, wasteful tests. Health care delivery is fragmented between hospitals, ambulatory clinics, physicians' offices, and other settings, creating concerns about the financial motives of providers to offer care in particular types of settings. Electronic health record ("EHR") systems often cannot communicate with each other, potentially leading to adverse events for patients whose providers are not aware of all of their relevant health information. (10)

The existing health policy and legal literature has identified these problems of health care fragmentation and argued that they drive up costs, lower health care quality, and may harm patients. (11) Some of these forms of fragmentation may be justified by other benefits (such as physician specialization), and in those cases scholars and policymakers have proposed reforms that attempt to mitigate the potential harms of fragmentation but not to eliminate the fragmentation itself. (12) In other cases, scholars have argued that fragmentation has few if any redeeming qualities (such as for EHRs that cannot share information) and have proposed reforms that would eliminate the fragmentation entirely. (13)

Yet scholars have not recognized the impact of health care fragmentation on incentives for innovation into new health care technologies. This Article therefore makes two primary contributions. First, it identifies and describes two new forms of health care fragmentation that have not been articulated in the literature: fragmentation by benefit structure and fragmentation by policymaker. For instance, patients do not only move in and out of different insurance plans over time. Even within those plans, insurance is often fragmented by type of product and is delivered separately for medical services and for pharmaceuticals.

Second, this Article applies the overarching framework of fragmentation both to questions of health care cost and quality and also to questions of innovation policy. The above-described fragmentation over time, as patients move in and out of different insurance plans, certainly does impact patients' access to care. But in so doing, it also impacts pharmaceutical companies' incentives to bring new drugs to market. In practice, insurance reimbursement functions to reward innovator firms for their discoveries. As a result, insurer decisions to provide reimbursement for a class of technologies expand the potential returns for companies in that area and drive research and development investment. By contrast, when an insurer aims to avoid covering a new class of products, that decision decreases potential returns and incentives to innovate in that class. When combined with fragmentation, this dynamic introduces a problematic incentive into our health care system, in which private market signals may distort innovative activity away from a class of therapies that would be highly socially valuable. Similarly, fragmentation by benefit structure and by policymaker have impacts on companies' innovation-related decisions.

Although these three forms of fragmentation--over time, by benefit structure, and by policymaker--all play out in different practical ways, their impacts on innovation incentives share a core deficiency. In each case, a particular actor (typically an insurer) would bear all the costs of making a particular coverage decision, but would reap few if any of the benefits. Those externalities would redound to other actors, including future insurers. As such, because these actors cannot internalize both the full costs and benefits of providing access to certain types of therapies, they will provide less access to those therapies relative to what might be socially advantageous, discouraging companies from investing in those kinds of products.

This Article proceeds in five Parts. Part II defines the concept of fragmentation and provides examples of the ways in which the concept has been deployed to describe a range of circumstances throughout our health care system. Many of these examples of fragmentation are explicit creatures of law, but others are driven more by economic incentives that arise under our existing legal and policy structures. Part II also details the impacts that the literature has previously identified as resulting from fragmentation, in terms of health care costs, quality, and patient outcomes.

Parts III, IV, and V examine particular forms of fragmentation in detail and articulate the impacts they are likely to have on pharmaceutical innovation. Part III begins with the above-described fragmentation over time, in which patients cycle on and off of different insurance programs (and may even lose insurance entirely), in a way that discourages insurers from providing coverage for certain products or services. Although the existing literature has previously identified the phenomenon of fragmentation over time, scholars have not considered the ways in which this form of fragmentation may create challenges for innovation policy. Because fragmentation over time encourages insurers to delay providing one-time treatments or preventive care, in hopes that an insured patient may soon become another insurer's financial responsibility, it may discourage the development of those types of products.

Parts IV and V consider forms of fragmentation that have not been articulated in the literature, but that come into view when fragmentation is examined through an...