Artificial intelligence (AI) can potentially increase the utility of genetic information by contextualizing it along with other patient variables to inform clinical decision-making. One such example is the development of AI-enabled predictive pharmacogenomic tools to assist with medication selection for mental health disorders. These tools leverage pharmacogenomic information to prioritize medications and dosages that will be most effective, accelerating a process that is often challenging for patients and physicians due to disorder complexity and patient variability. However, there is a lack of data characterizing the perspectives of physicians treating patients with mental health concerns regarding AI-enabled pharmacogenomic tools. We present our study involving case-based interviews with 42 physicians (21 family medicine practitioners and 21 psychiatrists) from an academic health system. Interviews were transcribed and subsequently analyzed using qualitative analysis methods. Our findings suggest that there may be differences in how clinical specialties view the value of AI-enabled pharmacogenomics tools. Family medicine practitioners showed optimism about the potential for these tools to assist them in caring for their patients. Psychiatrists were generally more skeptical of the underlying science and applicability of pharmacogenomics independent of AI support. Physicians also described benefits and risks of these tools in clinical practice. We contextualize our findings to debates over the value of precision medicine when pharmacogenomic techniques are introduced to the clinic, highlighting issues related to physician adoption of patient genetic information in clinical decision-making and appropriate knowledge to leverage these technologies. We further discuss the nuance added by AI.
Authors: Austin Stroud, Mayo Clinic; Susan Curtis, Mayo Clinic; Isabel Weir, University of Virginia; Jeremiah Stout, Mayo Clinic; Joel Pacyna,Mayo Clini; Journey Wise, Mayo Clinic; Barbara Barry, Mayo Clinic; Arjun Athreya, Mayo Clinic; William Bobo, Mayo Clinic; Richard Sharp, Mayo Clinic
