Volume 16, Issue 2

Personalized medicine is best viewed from a broad perspective of trying to use information about a patient to improve care. While “personalized medicine” often emphasizes the value of genetic information, traditional clinical approaches to personalizing care based on patient phenotype, provider and system-level factors should not be neglected. As these diverse approaches to personalization are examined, tools such as cost-effectiveness analysis can provide important insights into the value of these approaches, strategies for their implementation and dissemination, and priorities for future research. Such analyses are likely to be most insightful if they recognize that patient and provider behaviors are essential determinants of the value of treatments and that patient factors in particular may have large effects on the value of treatments and the need for interventions to improve decision making. These comments suggest three major areas of opportunity for economic analyses of personalized medicine: (1) traditional clinical approaches to personalized medicine, (2) multi-perspective studies of the benefits and costs of personalized medicine, and (3) the role of behavior in the value of personalized medicine.

The objective of this paper is to provide an overview of economic evaluation of personalized medicine, focusing particularly on the use of cost-effectiveness analysis and other methods of valuation. We draw on insights from the literature and our work at the University of California, San Francisco Center for Translational and Policy Research on Personalized Medicine (TRANSPERS). We begin with a discussion of why personalized medicine is of interest and challenges to adoption, whether personalized medicine is different enough to require different evaluation approaches, and what is known about the economics of personalized medicine. We then discuss insights from TRANSPERS research and six areas for future research: Develop and Apply Multiple Methods of Assessing Value Identify Key Factors in Determining the Value of Personalized Medicine Use Real World Perspectives in Economic Analyses Consider Patient Heterogeneity and Diverse Populations in Economic Analyses Prepare for Upcoming Challenges of Assessing Value of Emerging Technologies Incorporate Behavioral Economics into Value Assessments

Personalized prevention uses family history and predictive genetic testing to identify people at high risk of serious diseases. The availability of predictive genetic tests is a newer and still-developing phenomenon. Many observers see tremendous potential for personalized prevention to improve public health. At the same time, the emergence of these new markets raises familiar health policy concerns about costs, cost-effectiveness, and health disparities. This paper first discusses an economic framework for the analysis of personalized prevention. On the demand side, consumers use personalized prevention as a form of information that allows them to make better choices about prevention, including medical care and health behaviors like diet and exercise. On the supply side, an interplay of complex market forces and regulations will determine the prices, advertising, and insurance coverage of predictive genetic tests. Beyond the question of whether health insurance will cover the costs of predictive genetic tests, there is a great deal of concern about whether consumers’ use of genetic tests might place them at risk of genetic discrimination or might lead to adverse selection. The paper also reports descriptive analysis of data from the 2000, 2005, and 2010 National Health Interview Surveys on the use of predictive genetic tests. The empirical analysis documents large socioeconomic status-related disparities in consumers having heard of genetic tests: for example, consumers with less schooling, Blacks, and Hispanics were substantially less likely to have heard of genetic tests. Evidence from other empirical studies provides little evidence that genetic testing leads to genetic discrimination in insurance markets. There is more evidence suggesting adverse selection, where genetic testing leads consumers to purchase long-term care insurance. The paper concludes with some preliminary thoughts about important directions for future research. The goal of the paper is to review relevant research to help develop an economic approach and social science research agenda into the determinants and consequences of genetic tests for prevention.

The world of patient-centered outcomes research (PCOR) seems to bridge the previously disjointed worlds of comparative effectiveness research (CER) and personalized medicine (PM). Indeed, theoretical reasoning on how information on medical quality should inform decision making, both at the individual and the policy level, reveals that personalized information on the value of medical products is critical for improving decision making at all levels. However, challenges to generating, evaluating and translating evidence that might lead to personalization need to be critically assessed. In this paper, I discuss two different concepts of personalized medicine – passive personalization (PPM) and active personalization (APM) that are important to distinguish in order to invest efficiently in PCOR and develop objective evidence on the value of personalization that will aid in its translation. APM constitutes the process of actively seeking identifiers, which can be genotypical, phenotypical or even environmental, that can be used to differentiate between the marginal benefits of treatment across patients. In contrast, PPM involves a passive approach to personalization where, in the absence of explicit research to discover identifiers, patients and physicians “learn by doing” mostly due to the repeated use of similar products on similar patients. Benchmarking the current state of PPM sets the bar to which the expected value of any new APM agenda should be evaluated. Exploring processes that enable PPM in practice can help discover new APM agendas, such as those based on developing predictive algorithms based on clinical, phenotypical and preference data, which may be more efficient that trying to develop expensive genetic tests. It can also identify scenarios or subgroups of patients where genomic research would be most valuable since alternative prediction algorithms were difficult to develop in those settings. Two clinical scenarios are discussed where PPM was explored through novel econometric methods. Related discussions around exploring PPM processes, multi-dimensionality of outcomes, and a balanced agenda for future research on personalization follow.

Personalized medicine – the targeting of therapies to individuals on the basis of their biological, clinical, or genetic characteristics – is thought to have the potential to transform health care. While much emphasis has been placed on the value of personalized therapies, less attention has been paid to the value generated by the diagnostic tests that direct patients to those targeted treatments. This paper presents a framework derived from information economics for assessing the value of diagnostics. We demonstrate, via a case study, that the social value of such diagnostics can be very large, both by avoiding unnecessary treatment and by identifying patients who otherwise would not get treated. Despite the potential social benefits, diagnostic development has been discouraged by cost-based, rather than value-based, reimbursement.