While the promise of pharmacogenomics might seem obvious, it has yet to be widely adopted in healthcare. This is due in part to the cutting-edge nature of the field; the fact that for many medications, multiple genes affect metabolism; and the impact of environmental factors. In addition, when prescribing medications, physicians must still consider issues such as drug-drug interactions, cost, and guideline recommendations.

Operationally, there remains a great deal of difficulty in receiving, processing, managing, and sharing pharmacogenomic testing results. On the data-sharing front, a significant barrier is the lack of a universally accepted terminology for documenting test findings. Unique terms associated with genetic testing, such as the genotype CYP2D6*1/*5, have not been standardized to easily pass from one IT system to another. Then there is the matter of data formats. When a physician does order a pharmacogenomics panel, the results often come back as narrative stored in a PDF. While such documents can be stored and retrieved from EHRs, the information is not always readily visible and is not computable as part of an electronic decision support system.

Finally, the relative “newness” of pharmacogenomics – the first FDA approval of a pharmacogenomics test only took place in 2005 – means that many clinicians currently in practice weren’t exposed to the field while in medical school. As a result, these providers may be less likely to use genetic data for clinical decision-making, and instead may look to other resources they know from experience to be reliable.