» Back to Table of Contents

Director's Letter

The old saying “what’s sauce for the goose is sauce for the gander” doesn’t apply in medicine. Every clinician knows that treatments that work for one patient may not work for another.¹ Treating chronic diseases such as hypertension or depression often involves multiple rounds of trial and error to find a drug that provides the greatest benefit and fewest side effects. The growth of predictive analytics has improved our ability to estimate benefits or harms based on demographic or clinical data while pharmacogenomic research has uncovered physiologic mechanisms underlying differences in drug metabolism, drug effectiveness, and side effects. Nowhere has the progress been more visible than in the development of drugs targeted to underlying genetic abnormalities in specific cancers. A future where doctors can prescribe the right treatment to the right patient at the right time is no longer a distant reality.

That said, claims about precision medicine can get ahead of solid evidence of clinical value, abetted by commercial interest in selling genomic tests or electronic health record-based decision tools. The solution to this is better research. As more of these tests and tools come to market several critical questions arise that neither manufacturers nor regulatory agencies are positioned to address. First, where are precision medicine approaches, and specifically searches for genomic determinants, most useful for guiding therapy? Many factors contribute to variable responses to therapy, and not all of them are biologic/physiologic factors. Especially for patients with chronic diseases, individual factors such as health priorities, health literacy, and trust in the health system may have an important effect on a patient’s ability to adhere to specific treatments, and none of these factors are addressed by a genomic test. Second, when is evidence sufficient to scale any of these approaches across a health system? For genomic tests, we need more studies like that of Oslin et al. (which is described in this issue) that measure the extent to which precision medicine tests alter treatment decisions, and whether those decisions lead to better outcomes. Finally, for interventions ready for wider adoption, how do we best equip clinicians to use them in practice? This last question suggests an area ripe for implementation science. Advances in precision medicine are exciting, but we will need health services researchers to implement these advances into practice so they can benefit patients.²

References

1. Kravitz RL, Duan N, Braslow J. “Evidence-based Medicine, Heterogeneity of Treatment Effects, and the Trouble with Averages,” Milbank Quarterly 2004;82(4):661-87.
2. Atkins D, Makridis CA, Alterovitz G, Ramoni R, Clancy C. “Developing and Implementing Predictive Models in a Learning Healthcare System: Traditional and Artificial