Insurance Risk Classification in an Era of Genomics: Is a Rational Discrimination Policy Rational?

摘要: There is continuing societal debate about whether insurers should be able to collect, access, or use genetic test results when considering applications setting premium and coverage levels.1 This centers around deeply rooted beliefs over the privacy personal nature of information on one hand financial necessities economic considerations insurance industry other. Insurers argue access applicants’ essential for industry’s security. However, public distrust companies, coupled with anecdotal evidence individuals unable secure insurance, led calls barring from and, in context health realization this goal.2 In 2008, Congress passed Genetic Information Nondiscrimination Act (GINA), which bars covered employers collecting using information.3 Other insurers, such as life, long-term care, disability are exempt law.4 Since GINA’s passage, continued suggestions have been raised expand legislation these other insurances, but date regulation has limited, variable, confined state level.5 It remains an open question how by circumscribed. Across globe, countries similarly struggle balance concerns. many countries, especially those that universal national care systems, focused types insurances excluded GINA, income, critical illness insurance.6 a survey international policy approaches addressing insurer information, Yann Joly colleagues identified six intersecting overlapping often employed.7 Three relate specifically can all, some, no results—labeled status quo, rational discrimination, prohibitive approaches, respectively.8 Under quo approach, set their own standards tests they will gather use.9 approach premised fact overarching goal market assess risks order offer commensurate risk level economically viable.10 Additionally, since companies business attracting customers, incentive accurately possible.11 For two reasons, allows create rules characteristics,12 implicitly trusting available data responsibly. contrast, under barred accessing results.13 Such prioritizes social arguments unfair, may violate concerns, discourage undertaking medically recommended testing. The discrimination stands middle ground between approaches. Here, allowed subset meet established scientific, clinical, actuarial significance.14 Given current scientific knowledge, actually quite small. Overwhelmingly, do not provide helpful information. Despite promises, clinical understanding genetics is, most part, lackluster at best its ability predict risk.15 Although some identify high conditions cancer neurodegenerative diseases, small overall information.16 Much arising vulnerable misinterpretation17 arguably effect aggregate health18 markets.19 Therefore, narrows only criteria prediction. A was implemented United Kingdom early 2000s, versions proposed States Australia.20 Australia policies place independent body mediator decisions, external committee must approve before permitted. Article examines merits drawbacks address results. argues adopted necessary baseline protection against misuse results, while allowing met sufficient evidence. Part II begins overview theoretical underpinnings insurance—social fairness fairness—and theories employed U.S. law. Social focuses solidarity; ensuring paramount goal. fairness, also called fair seeks treat equal equally unequal unequally. Part III applies frameworks discusses greater depth three results—a quo. Part IV motivate classification. Risk classification process gathering referred “characteristics,” applicant, assessing characteristics affect risk, placing applicants into classes dictate levels. Historically, used relatively number characteristics, age, gender, occupation, individuals’ assign premiums; however, time, became more refined increasing medical exams lifestyle factored assessment.21 The technical realm imbued subjectivity. Thus, foreshadows areas where ensure appropriate met. following currently Precision medicine advancing technologies further possible challenging existing regulatory bounds data. Some argued because it insurers’ interests avoid forces obviate need additional oversight. VI responds arguing insufficient check system due burden placed consumers, competitive pressures face narrowly refine capture low-risk differences markets markets, failure fear discrimination. Finding provides balanced mechanism VII delineates determining relevance States, Kingdom, Australia. proposes five inquiry any determination As discussed VIII, compromise does all concerns insurers. Part, maintains positions beneficial both. A balances needs public. regulator quipped, “The math correlation so lovely. Because blind fair. can’t biased doesn’t know who you are, you.”22 Assessing highly determination, statistical exact science—it blind, certainly guaranteed fair.23 Instead, underlying assumptions, evidence, inputs go calculations outcome. lack predictive precision useful parents weighing subjective patients alternative treatments, generally feasible A potential playing field across nationally, enhance efficiency system, increase trust classification, dampen fears grappling genetic-test