Personalized medicine is a move away from one-size-fits-all health care in favor of decision-making that is algorithmically optimized for each individual. Paradoxically, this individualization process requires population-level data, including genomic information and medical records. Give up your data, advocates promise, and the payoff will come in improved treatments, fewer side effects, and better preventive care. Sounds great. What, asks Barbara Prainsack in her meticulously researched and closely argued new book, could possibly go wrong?
Readers of Personalized Medicine: Empowering Patients in the 21st Century? will note the subtitle’s question mark. Without it, one might reasonably assume that this book was about personalized medicine and its potential to empower patients. It is not. Instead, the book is a thoughtful, thorough, and philosophical discussion of the many possible obstacles to the successful, equitable implementation of personalized medicine and its potential for unintended consequences.
While some of the groundwork for a system of personalized medicine has been built in recent years, including public and private databases for genomic sequence data and other health-related information, much of it remains aspirational. It’s admirable that Prainsack attempts to be proactive about identifying the potential pitfalls and weaknesses of personalized medicine; the fact that it is by necessity mainly speculation is both the book’s greatest strength and its greatest weakness.
Advocates for personalized medicine often talk about patient “empowerment.” Generally, empowerment is envisioned to mean that each individual will have access to his or her genetic and health information and web resources that will enable better decisions in and out of the medical setting. But as Prainsack points out, this sort of empowerment places a burden on patients, requiring them to take an active role in identifying and managing their health risks. What seems empowering to one individual may be exhausting to another. And as a system, it advantages those with education, time, and energy, while potentially exacerbating existing inequities, a theme Prainsack returns to often.
A system rooted in autonomy and empowerment, Prainsack suggests, is likely a system that de-emphasizes collective responsibility and leads us to ignore the public health consequences of socioeconomic inequality. Furthermore, the empowerment rhetoric can mask the commercial agenda of participants in a growing industry. As she points out, it has become difficult to make distinctions between for-profit and nonprofit, and academic
Prainsack writes about initiatives that appear participatory and consumer-driven, but that are also a means of data collection. While these services present themselves as information-giving, they are also information-taking. The internet, Prainsack reminds us, watches you.
One of the downsides of proactive commentary is that it tends to be long on evocative scariness and short on concrete examples of bad outcomes. But Prainsack effectively makes the case that the useful information collected to refine estimates of risk in order to prevent heart attacks or find cancer early could also be used to identify people whom employers or insurers might want to avoid. We tend to think of using personalized risk estimates to offer individuals preventive care, but, as Prainsack points out, they could also be used as a means of rationing, so that scarcer or more expensive resources are reserved for those more likely to respond.
The privacy implications of our growing data repositories have been discussed at length, and Prainsack, like others, notes that models based on existing legal concepts like ownership are woefully inadequate, pen-and-paper vestiges in a digital world (see “Do You Belong to You?” in the Winter 2017 issue of Genome). Discussions about privacy often remind me of elementary school parent-teacher conferences for my youngest son. Year after year, there would come a moment when the teacher turned to me with a concerned face and say, “spelling is a problem.” And we would both nod and look serious for a moment. And then the conversation would move on to something else, because there was not a damn thing to be done.
Privacy. It’s a problem.
The heart of Prainsack’s argument, and the most interesting part of the book, is the charge that personalized medicine as we have proposed it is inherently flawed because it is reductive. Personalized medicine is based on data and therefore is liable to overvalue that which can be measured and undervalue that which cannot be. What’s more, she suggests, among the various sources of data, we are prone to prefer those that seem least subjective and most high tech, whether or not that is borne out by the evidence. As Prainsack puts it:
Visions of personalized and precision medicine are based upon a tacit hierarchy of utility, with digital and computable data on top, and unstructured narrative and qualitative evidence at the bottom. This hierarchy was not brought about by public deliberation, and not even by discussions among professional experts; instead, it is a hierarchy that was brought about with the technological tools and practices that we, as a society, have decided to invest and trust in, and that corporations see as profitable.
Prainsack argues for the inclusion of what she calls “social biomarkers” — data that would capture contextual information, including environmental exposures, cultural influences, and co-morbidities, among others. She uses a number of bioethical and utilitarian arguments. Lose these “squishier” types of data from the medical record and medical care suffers. Lose them from the public record and they cease to be a part of the conversation about how to improve health outcomes.
Laura Hercher, MS, CGC, is a genetic counselor and the director of student research at the Sarah Lawrence College Joan H. Marks Graduate Program in Human Genetics.