2.5 Retrospective documentation and validation
For the retrospective data analysis and validation as part of this study, the clinical pharmacologist (SR), the clinical pharmacist (DN) and a pharmacist in training (AR) reviewed all available original medical records, referral letters, pharmacotherapy prescriptions and laboratory results. Patient characteristics and clinical factors including current pharmacotherapy, laboratory results and medical history were extracted and compiled in a study database. Comedications were also categorized according to their potential for moderate or strong inhibition of cytochrome P450 enzymes 2C19 and 2D6 according to the mediQ-database (www.mediq.ch).
All clinical recommendations from the reports were validated and categorized as appropriate.
First, in patients where a specific drug-gene pair was the indication of PGx testing, we documented if the test result of the related gene led to a recommendation to change therapy with the drug that triggered PGx testing.
Second, current comedication and results for all 16 genes of the PGx panel were analysed for any additional clinically relevant drug-gene interactions.
Third, for all subjects including those with a screening indication, we documented if any PGx variants were detected that related to a drug-gene pair with “actionable”, “recommended” or “required” classification according to PHARMGKB. Such variants were presented in our PGx reports as potentially relevant for future medication and further discussed in the individual clinical context of tested subjects.
Drug-gene pairs, their classification of clinical relevance according to PHARMGKB, and the assignment of genotypes to according phenotypes are presented in supplementary Table S2 .