Impact of AI Scribes on Physician Productivity and Workflow in Medical Oncology.

Abstract

Purpose: Artificial intelligence (AI) scribes are being rapidly adopted in oncology, yet their real-world impact on physician productivity, workflow, and workplace satisfaction is understudied. We evaluated the quantitative and qualitative effects of an AI scribe with electronic medical record (EMR) integration in a multisite community oncology practice.

Methods: This single-center, multisite study enrolled 22 medical oncologists and three primary care physicians randomly assigned to receiving training and exposure to the Knowtex AI scribe in an initial phase or control phase. Physician billing data were collected for 6 months before the onboarding of intervention arm physicians and for 16 weeks afterward-within the 2024 fiscal year. Surveys administered at baseline and week 8 postexposure assessed documentation burden, quality, EMR experience, and overall documentation satisfaction.

Results: All enrolled providers adopted the Knowtex AI scribe during their respective study phase. Intervention arm physicians exhibited an increase in mean visits (β = 3.35, SE = 1.08, P = .010) and mean total billing per working day (β = $433.61 in US dollars, SE = $134.82, P = .007). There was no change in the number of diagnostic codes per visit. Physicians' perceptions of EMR challenges markedly decreased after implementing the AI scribe (P < .001) with no significant change in the quality of documentation. Physicians reported higher satisfaction with documentation workflow, reduced in-clinic documentation hours, and increased time with patients.

Conclusion: Adoption of an AI scribe with EMR integration in oncology clinics may enhance certain billing metrics and positively shift physician perceptions of documentation processes and EMR challenges. These findings highlight potential benefits and limitations of AI scribes in improving physician productivity and satisfaction.