Invited Commentary
Health Informatics
December 18, 2024
Characterizing Residents’ Clinical Experiences—A Step Toward Precision Education
Jesse Burk-Rafel, Carolyn B. Drake, Daniel J. Sartori
JAMA Netw Open. 2024;7(12):e2450774. doi:10.1001/jamanetworkopen.2024.50774
Experiential learning from patient care activities is the primary means by which residents develop the skills that shape their future practice. However, these clinical learning opportunities are typically not well characterized, limiting educators’ ability to precisely tailor the experiential curriculum.
Elsewhere in JAMA Network Open, Lam and colleagues1 provide a detailed characterization of internal medicine residents’ experiences admitting patients overnight within a single Canadian residency program’s 5 teaching hospitals over a 10-year period. The authors developed a bespoke clinical-educational database that captures rich data from the electronic health record (EHR) for each included clinical encounter and subsequently attributes each to a specific admitting resident. Leveraging this database, they identified significant resident, hospital, and temporal variation across 6 domains, including both admission volumes and multiple patient and encounter characteristics. This work contributes in important ways to better understanding the considerable variability that exists in residency training, including unpacking what might be warranted or unwarranted variability. The authors’ findings1 are consistent with those of recent publications also using EHR data to characterize the wide variations in clinical content exposure that exist at both the program and individual resident levels.2,3 Their research1 also adds new dimensions to our understanding of such variation by evaluating additional factors, including patient acuity, complexity, and social determinants of health.
Practice habits established during residency imprint durably, driving future practice patterns. However, it is not yet clear how variations in individual residents’ experiences affect their developing practice habits and their progression toward competence. In the Lam et al1 cohort, some residents admitted 3 times as many patients with heart failure as did their peers. This resurfaces questions raised in previous research3: is volume of exposure to a diagnostic area consistently associated with improved care in that area? How might this variability impact residents’ global development of diagnostic and management reasoning? Could shift-level patient volume, acuity, and complexity be used to assess cognitive load and optimize the clinical learning environment for patient care and resident well-being? How do contextual characteristics of an encounter, such as patient demographic characteristics, influence a resident’s learning yield and ability to apply gains in knowledge and skills to future patients? The complex relationships between experience and competence remain underexplored and warrant further study; answering these questions will be essential for advancing precision education.
Precision education is an emerging educational approach that seeks to use data-driven insights to create precise educational interventions tailored to each trainee’s needs and goals, accelerating their path to competency and potentially improving patient care outcomes.4,5 In this model, meaningful data about each trainee are the fundamental currency for generating insights. Problematically, widely utilized measures of trainee performance such as standardized examination scores and faculty rater–based assessments, lack density and do not directly capture clinical exposure, limiting our ability to evaluate the relationships between residents’ actual learning opportunities and their clinical skill development. By contrast, EHR data can support novel, temporally dense, automated assessments of trainees, characterizing individual clinical experiences and linking them to trainee-attributable care patterns and patient outcomes.6
Clinical experience data are an essential prerequisite to tailored experiential learning opportunities. As Lam et al1suggest, EHR-derived insights regarding case-mix exposures could allow for the creation of individualized study plans meant to target trainees’ specific gaps. Perhaps more radically, patients could be intentionally and dynamically assigned to trainees based on their prior experience managing certain diagnoses, while trainees lacking exposure to higher-acuity scenarios—or who have only minimal experience caring for patients with certain social determinants of health—might benefit from adjusting their clinical rotations across a program’s teaching sites. Residency programs can also use aggregate exposure data to drive programwide curricular change. As has been previously described,3the identification of low aggregate-level exposure to specific diagnoses at a particular training site can allow program leadership to create dedicated rotations on subspecialty units as well as targeted didactics and simulations to address experiential blind spots. These insights can be visualized through resident- and faculty-facing dashboards to facilitate individual self-reflection and to enhance the accuracy and specificity of faculty-led coaching.
Routinizing the use of resident-attributed patient care data from the EHR is a key next step for both research and education. However, several challenges lay ahead as the collection and application of these data become more widespread. Accurately attributing patient encounters to specific residents is complex, and even accurate attribution does not always imply meaningful contribution to each patient’s diagnosis or management. More sophisticated disease-specific approaches to defining residents’ contributions to patient care have recently been developed and are currently being validated6,7; these will complement existing clinical experience data and provide a more complete picture of trainees’ contribution to patient outcomes. In addition, it is not yet clear what constitutes overexposure or underexposure for a given patient condition or characteristic or what features of an encounter drive its value as a learning opportunity. Though normative referencing of one resident’s exposures relative to their peers is convenient, there are numerous contextual factors that shape how residents learn from experience, and much more needs to be learned about the relationships between experience and outcomes before evidence-based targets can be defined. Finally, residents’ dual role as both learners and employees adds complexity to the task of tailoring their clinical experiences, as adjusting clinical assignments is logistically challenging. Innovative and flexible approaches to precision interventions will be necessary to feasibly individualize training curricula.
Residency training is the period in which physicians cultivate expertise in tending to patients’ clinical and social needs. With this research, Lam and colleagues1 help to build the foundation for further exploration of exactly how residents’ clinical experiences shape their trajectory toward expertise, influencing their future practice patterns and patient outcomes. This and other ongoing research build an essential evidence base for how to move residency training from a fixed, one-size-fits-all approach to a dynamic, precise set of curricula tailored to each learner’s needs. We encourage other training programs to similarly embrace the use of clinical experience data to advance the paradigm of precision education.
