Imagine this: You attend on an inpatient medicine unit and assess a resident’s ability to interpret an electrocardiogram (ECG) in a patient presenting with chest pain. As part of this assessment, you are asked to report the degree to which you trust the trainee to do this task, unsupervised. What thoughts run through your mind as you make this decision? Do you consider the trainee’s board score or how long you have worked with them? Do you weigh the context of your decision or the complexity of the case? Or do you bring something else to the table, hidden even from you: Your own propensity to trust, biases, and beliefs? We believe a supervisor’s ability to trust is a crucial element in accurately assessing the competency of our trainees.
Increasingly, Entrustable Professional Activities (EPAs) have been used to leverage competency-based assessment in graduate trainees. Indeed, some professional societies have initiated the groundwork to use EPA assessment as a basis for high stakes decisions, including readiness for graduation and board certification. EPAs show promise by creating moments of objectivity grounded in our daily work. Did the resident obtain an ECG in that patient with chest pain and identify ST changes? Did they evaluate and manage the patient with new hyponatremia? Despite the general thrust towards EPA assessment in graduate medical education (GME), current validity evidence suggests we are not as objective as we believe.
Trust is the vehicle for trainee autonomy and the basis of entrustment; yet trust is a complex construct. Mayer, et al, previously defined trust as “the willingness of a party to be vulnerable to the actions of another party based on the expectation that the other will perform a particular action important to the trustor, irrespective of the ability to monitor or control that other party.”1 In ancient Greek, the word for trust—πίστις (pistis)—can be interpreted as faith. The Christian New Testament describes trust as “the substance of things hoped for, the evidence of things not seen” (Bible, Hebrews 11:1).2 The very nature of trust is belief in a reality for which we do not have direct evidence. Therefore, to trust a trainee to perform tasks on their own requires some degree of faith on the part of the supervisor.
Our medical system is designed to assume trust, and educators have long accepted some amount of risk to afford trainee autonomy with an eye towards growth.3 Existing data suggests the system works for most trainees. Most residents meet their milestones and successfully graduate; however, EPAs add a different weight to our assessments in which they ask us to report the degree we trust our trainees. It has been suggested that such trust is a potent vehicle for trainee growth.4 Yet, as supervising physicians we must also ensure patient safety and proper care. Every patient needs a safe follow-up plan, but not every patient needs a workup for a pheochromocytoma. How much do we trust our graduate trainees? Do we have faith in them? How does our own discomfort with letting go affect trainee progression towards autonomy? Importantly, how much does it bias our assessment?
Previous conceptualizations of supervisor-learner trust describe factors such as the clinical context, nature of the task, trainee competence, supervisor bias, and the underlying relationship between both parties; however, the relative weight of these variables has not been closely examined.4 Existing evidence suggests that a substantial amount of variability in EPA assessment may come from the assessors themselves, and qualitative studies have shown that supervising attendings’ internal processes do indeed guide entrustment decisions—a “reflexive trust” grounded in prior experiences with trainees, internal rules, and personal biases.4 While many such studies have emphasized faculty development as an avenue to mitigate this variability, these strategies do not closely examine faculty members’ internal trust processes, their propensity to trust, or their implicit biases. In assessing a trainee, the question is not only, “How trustworthy is the trainee?” but also, “How comfortable am I with trusting others?” Awareness of our proclivity to trust others (or not), our biases, and our prior experience become crucial for accurate evaluation of trainees.
In 1995, Mayer, et al, described a model of trust that argued measuring a supervisor’s willingness to be vulnerable to the actions of others was key to understanding trust decisions.1 Their work further clarified an individual’s willingness to trust correlated with whether they ultimately entrusted tasks to others. In Mayer’s model, a supervisor’s propensity to trust (considered to be a relatively stable personality characteristic of an individual) impacts all aspects of the trust process, including the trustors perceptions of the trustee’s ability, benevolence, and integrity. This is deeply subjective territory and makes every evaluation vulnerable to a supervisor’s “comfort.” We propose that a close examination of frontline assessors’ willingness to trust is critical to accurate EPA assessment. Currently, there are no validated metrics for this purpose, but previously defined scales in social science literature might serve as a useful start.5 Faculty development efforts should focus not only on EPA frameworks but also allow participants to reflect on their own biases towards trusting trainees and the process of arriving at trust. This is a delicate issue. Why do some preceptors trust more easily than others? Examining our own biases with authenticity opens ourselves to our own insecurities, our own failures, and our own moments of untrustworthiness. At our intuition, we created an exercise of faculty self-reflection adapted from Frazier, et al, 20135 to initiate these conversations. A downloadable version of this document can be found at https://tinyurl.com/3ta86y25.
There must be a balance between trust and autonomy. Endowing trust upon a trainee is essential for their growth and sense of identity as a developing physician. Their autonomy must, of course, be balanced by patient safety. As educators we have attempted to objectify this practice: chart stimulated recall, didactics, direct observation, modeling, mentorship, narrative medicine—the list goes on. However, EPAs add a newer and potentially deeper facet to the assessment paradigm. Informally, we have been doing EPAs for years. We have allowed our residents to enter the room without us—perhaps by some instinct that we did not recognize at the time. Only in recent years have we attempted formally to proscribe EPAs as objective exercises in vetting autonomy. Because EPAs are not as objective as we believe, supervisors must be aware of their own comfort with trust. They must reflect on how willing they are to be vulnerable to the actions of others to truly ground their assessment in objectivity. They must—at times—have pistis. Faith.