AI-Powered Personalized Onboarding & Early Training for Family Medicine PGY1s Transitioning from a UME to GME Learning Environment at Bayhealth
Introduction:
The transition from undergraduate medical education (UME) to graduate medical
education (GME) represents one of the most critical and challenging periods in a
physician's training journey. While marking a significant step towards independent
practice, the steep increase in clinical responsibility, cognitive load, and workplace
demands placed upon new interns frequently leads to substantial stress, anxiety, and a
well-documented increase in medical errors, often termed the "July effect"
[https://pmc.ncbi.nlm.nih.gov/articles/PMC8842721/]. Furthermore, this abrupt transition
is a known contributor to resident burnout, which carries significant consequences for
physician well-being, workforce retention, and potentially patient care quality
[https://www.ama-assn.org/medical-residents/medical-resident-wellness/pgy-2s-see-27higher-rate-burnout-interns-heres-why].
Compounding these challenges is the inherent variability among incoming interns.
Despite completing standardized UME curricula and licensing exams, residents begin
their PGY-1 year with diverse strengths, weaknesses, learning preferences, and levels
of preparedness for specific clinical tasks and the unwritten curriculum of their new
environment [https://pubmed.ncbi.nlm.nih.gov/37983405/]. Traditional residency
onboarding processes at institutions like Bayhealth often employ a standardized, onesize-fits-all approach. While providing essential orientation, these methods are typically
insufficient to address the unique learning needs of individual residents or proactively
identify those who may be struggling academically or emotionally during the crucial
initial months. This reactive model misses valuable opportunities for early intervention
and personalized support, potentially delaying competency development and
exacerbating transition-related stress.
There is a clear need for innovative, scalable solutions that move beyond traditional
onboarding paradigms to offer personalized, data-driven support tailored to the
individual needs of each incoming resident. To address this critical gap, we propose the
development and implementation of the AI-Powered Personalized Onboarding &
Early Training (AI-POET) platform at Bayhealth for the Family Medicine residency
program. AI-POET is designed to leverage existing UME performance data and early
GME assessments, utilizing artificial intelligence to generate individualized learning
plans, identify potential academic or well-being challenges proactively, and connect
interns with targeted institutional resources, ranging from specific educational modules
and simulation scenarios to faculty mentors and wellness programs.
This project directly aligns with the imperative to transform GME by fostering safer and
more effective transitions, enhancing resident well-being, and accelerating the
development of practice readiness through personalized competency development. By
providing proactive, individualized support from day one, AI-POET represents a novel,
technology-enhanced approach to mitigate the known risks of the UME-GME transition
and cultivate a more supportive and effective learning environment. This proposal
outlines the design, implementation, and evaluation plan for this innovative platform,
detailing its potential to significantly improve the PGY-1 experience and create a
scalable model for reimagining residency onboarding.
Specific Aims:
The overall goal of this project is to transform the challenging transition from
undergraduate medical education (UME) to graduate medical education (GME) by
developing, implementing, and evaluating an innovative AI-Powered Personalized
Onboarding & Early Training (AI-POET) platform for incoming interns. We hypothesize
that providing proactive, data-driven, personalized guidance and resource connections
will improve intern well-being, accelerate competency development, and create a safer,
more effective transition into residency. To achieve this goal, we will pursue the
following specific aims:
Aim 1: Develop and iteratively refine the AI-POET platform tailored to the needs of
incoming Family Medicine interns at Bayhealth Medical Center.
We will develop predictive algorithms utilizing selected UME performance data
(USMLE scores) and early GME assessments to identify potential individual
learning needs and well-being risks.
We will build a user-friendly, secure platform featuring an intern dashboard
displaying personalized insights and recommendations, and a curated, mapped
library of institutional resources.
Aim 2: Implement and integrate the AI-POET platform into the onboarding
process for the PGY-1 Family Medicine resident cohort starting July 2026.
We will deploy the AI-POST platform for use by all incoming PGY-1 residents in
the Family Medicine program at Bayhealth.
We will develop and deliver comprehensive training materials and sessions for
interns on utilizing the platform effectively and for faculty advisors/mentors on
interpreting insights (if applicable to the chosen model).
We will establish clear workflows for platform integration within the existing
residency onboarding structure and provide ongoing technical support during the
pilot implementation period (the first 6-12 months of internship).
Aim 3: Evaluate the impact of the AI-POET platform on the PGY-1 transition
experience, resident well-being, and early competency acquisition.
We will compare the AI-POET cohort to the previous year’s cohort on key metrics
including validated measures of stress, burnout, and self-reported confidence.
We will analyze GME performance data to assess any differences in the
trajectory of competency acquisition.
Expected Outcomes: Successful completion of these aims will result in a validated,
innovative platform designed to ease the UME-GME transition. We expect to
demonstrate measurable improvements in intern well-being and potentially accelerated
competency development. This project will provide a model for leveraging technology to
personalize support in GME and offer valuable insights into best practices for datadriven residency onboarding, contributing to the broader goals of transforming medical
education.