Characteristics of initial pharmacogenetic testing and subsequent physician referral behavior in youth mental health.

Abstract

While prior studies have focused on the clinical utility of pharmacogenetic (PGx) testing, less is known about factors that influence physicians' continued use of PGx testing in routine practice. Using data from the PGx-SParK trial, we examined whether the actionability of PGx test results from a physician's initial referral predicted subsequent PGx referral behavior. Youth aged 6-24 years referred for PGx testing in Western Canada between October 2020 and November 2025 were included. For each physician, an index PGx referral was identified and linked to patient characteristics, testing turnaround time, and CYP2B6, CYP2C19, and CYP2D6 results. Actionability for psychotropic prescribing was defined based on current PGx guidelines and categorized as current, future, or none. The number of subsequent referrals per physician was modeled using generalized linear models with a Poisson distribution corrected for overdispersion and adjusted for follow-up time and relevant covariates. Among 371 physician index referrals, 90% of tested youth had at least one current (22%) or future (68%) actionable result. Over a mean follow-up of 29 months, 52% of physicians made at least one subsequent PGx referral. Neither current nor future actionability of the initial PGx test results was associated with subsequent referral behavior. However, patient age (older) and physician speciality (psychiatrists) were associated with greater odds of a subsequent referral. These findings suggest that sustained physician engagement with PGx testing in youth mental healthcare may be driven more by implementation context and physician-level factors than by the clinical actionability of test results.