Barriers to clinical implementation (references in square brackets)

1. Should testing be performed?

Lack of evidence of clinical validity/utility of pharmacogenomic testing, including a lack of validated, pharmacogenomic-guided, treatment algorithms [34,35,36,37,38,39,40,41] Lack of evidence demonstrating cost-effectiveness of pharmacogenomic testing (and consequent impact on cost to a public healthcare system, private health insurance companies, and out-of-pocket patient costs) [34,37,38,40,41,42,43,44,45] Lack of expertise amongst prescribing clinicians to determine whether a pharmacogenomic test is appropriate [35,37,42,46] Lack of recommendations from professional organizations, or changes to health policy, to support clinicians in determining whether a pharmacogenomic test is appropriate [37,40,45] Discrepancies between pharmacogenomic guidelines of different organizations [37,47] Perceived or actual financial conflicts of interest for authors of research/guidelines supporting the utility of pharmacogenomic testing Ambiguity in how to clinically apply pharmacogenomic biomarker information in drug labels [37,41,48] Lack of physician acceptance of pharmacogenomic testing [35,37,39,41,44,45,49] Lack of patient acceptance of pharmacogenomic testing (concerns regarding privacy, genetic discrimination, cost, etc.) [41,44,50,51,52]

2. Challenges to integration

Logistics of, and regulatory requirements for, performing pharmacogenomic testing in the clinical setting, including selection of genomic testing platform [34,35,38,44] Logistics of integrating pharmacogenomic test results into the electronic health record [34,35,37,38,44,45] Logistics of incorporating pharmacogenomic testing into the clinical workflow [35,41,42,43,44,49] Lack of standardized report formats for pharmacogenomic test results and inconsistency in practices for data storage and retrieval [35,37,38,42,53] Logistical and ethical issues regarding stewardship of pharmacogenomic test results, including responsibility for re-analyzing results in light of new evidence [37,42,43,44] Complexity in pharmacogenomic test results, with attendant difficulties in interpretation—including the complex architecture of pharmacogenes such as CYP2D6, and the potential for interactions with other prescribed medications known to impact enzyme function [34,37,41,45,49] Ambiguity in pharmacogenomic test results, such as variants of uncertain significance, particularly for non-White populations (for which there is a paucity of reference data), and the lack of evidence regarding how to combine results from multiple pharmacogenes [34,37,39,54] Lack of expertise amongst prescribing clinicians to interpret and manage a pharmacogenomic test result [34,35,37,38,42,43,46] Lack of support for clinicians to interpret and manage pharmacogenomic test results, such as inadequate information in drug labelling, pharmacogenomic practice guidelines, or decision support infrastructure [34,35,37,40,45] Discrepancies between pharmacogenomic guidelines of different organizations [37,48] Perceived or actual financial conflicts of interest for authors of research/guidelines supporting the utility of pharmacogenomic testing Lack of physician acceptance of pharmacogenomic recommendations [35,37,49] Lack of ongoing patient engagement with pharmacogenomic testing (for communication of results) [44] Identification of increased disease risks incidental to pharmacogenomic testing (e.g., BRCA1/BRCA2 germline genetic test results to guide treatment also confer increased disease risk) [35,37]