Best Practices in Pharmacogenetic Implementation
Insight based on The University of Florida (UF) Health Precision Medicine Program (PMP) whose initial focus was on pharmacogenetic testing
Personalized medicine and genetic testing

The University of Florida (UF) Health Precision Medicine Program (PMP) launched in 2012 with an initial focus on pharmacogenetic testing. The first gene-drug pair implemented into practice was CYP2C19-clopidogrel in the setting of percutaneous coronary intervention. Since then, the program has expanded to support implementation of the following gene-drug pairs: TPMT/NUDT15-thiopurines, CYP2D6/CYP2C19-antidepressants, CYP2D6-opioids, CYP2C19-proton pump inhibitors (PPIs), and CYP2C9-nonsteriodal anti-inflammatory drugs. Each implementation occurred in response to a physician request for pharmacogenetic testing to help inform prescribing decisions. Panel-based testing became available in 2019 and covers nine pharmacogenes, including those listed above.

With each implementation, several key best practices are followed. First, it is critical that there is strong evidence to support genetic associations with drug effectiveness and/or safety. Importantly, it is not necessary for there to be evidence of improved clinical outcomes with the use of genotype information in prescribing decisions. Rather, genotype is considered to be another important factor, similar to age or serum creatinine, to consider when optimizing drug therapy for a given patient. The Clinical Pharmacogenetics Implementation Consortium (CPIC) evaluates and summarizes the pharmacogenetic evidence.1 For gene-drug pairs with a high-level of evidence, CPIC provides recommendations for how to integrate genotype results into prescribing decisions. Thus, for institutions newly implementing pharmacogenetic testing, CPIC guidelines are a great resource for evidence evaluation.

Another best practice followed at UF Health is related to the selection of pharmacogenetic variants for testing. There is significant variability in allele frequencies across ancestry groups, and many clinically relevant alleles are common in one ancestry group but rare in others. Our program strives to include any functional variant with appreciable frequency (e.g., ≥1%) in at least one population. For example, for the CYP2D6 gene, the decreased-function *17 and *29 alleles occur frequently in those of African ancestry. Therefore, even though they are uncommon in other populations, *17 and *29 are included on the UF Health CYP2D6 genotype assay. A helpful resource for selecting pharmacogenetic variants for testing comes from the Association of Molecular Pathology, which publishes recommendations for which alleles to include on clinical testing panels.2-4

A third best practice is that genotype results are entered into the electronic health record (EHR) as discrete data. This allows for building automated alerts within the EHR to guide prescribing decisions at the point of care. In the event that a provider orders a drug for a patient, and the patient has a genotype associated with reduced effectiveness or increased risk of adverse effects to the drug ordered, an alert will automatically appear to notify the provider of such. For example, if a provider orders tramadol for a patient with a CYP2D6 genotype associated with reduced tramadol effectiveness, an alert will appear to warn the prescriber that the patient may get little to no pain relief with tramadol.

A final best practice is that clear clinical decision support is provided to assist prescribers with applying genotype results to drug selection and dosing. Genotype-guided recommendations are provided through the automated alerts described above and are consistent with CPIC guideline recommendations. Referring to the example above, the alert would include recommendations to avoid tramadol and either consider a non-opioid analgesic or an opioid that is not affected by CYP2D6 metabolizer status, with specific examples provided. Clinical pharmacists are available for consultation as needed. 


  1.  Relling MV, Klein TE. CPIC: Clinical Pharmacogenetics Implementation Consortium of the Pharmacogenomics Research Network. Clin Pharmacol Ther 2011;89:464-7.
  2. Pratt VM, Cavallari LH, Del Tredici AL, Gaedigk A, Hachad H, Ji Y, Kalman LV, Ly RC, Moyer AM, Scott SA, van Schaik RHN, Whirl-Carrillo M, Weck KE. Recommendations for Clinical CYP2D6 Genotyping Allele Selection: A Joint Consensus Recommendation of the Association for Molecular Pathology, College of American Pathologists, Dutch Pharmacogenetics Working Group of the Royal Dutch Pharmacists Association, and the European Society for Pharmacogenomics and Personalized Therapy. J Mol Diagn 2021;23:1047-64.
  3. Pratt VM, Cavallari LH, Del Tredici AL, Hachad H, Ji Y, Moyer AM, Scott SA, Whirl-Carrillo M, Weck KE. Recommendations for Clinical CYP2C9 Genotyping Allele Selection: A Joint Recommendation of the Association for Molecular Pathology and College of American Pathologists. J Mol Diagn 2019;21:746-55.
  4. Pratt VM, Del Tredici AL, Hachad H, Ji Y, Kalman LV, Scott SA, Weck KE. Recommendations for Clinical CYP2C19 Genotyping Allele Selection: A Report of the Association for Molecular Pathology. J Mol Diagn 2018;20:269-76.