Clarifying essential terminology in entrustment

With the rapid uptake of entrustable professional activties and entrustment decision-making as an approach in undergraduate and graduate education in medicine and other health professions, there is a risk of confusion in the use of new terminologies. The authors seek to clarify the use of many words related to the concept of entrustment, based on existing literature, with the aim to establish logical consistency in their use. The list of proposed definitions includes independence, autonomy, supervision, unsupervised practice, oversight, general and task-specific trustworthiness, trust, entrust(ment), entrustable professional activity, entrustment decision, entrustability, entrustment-supervision scale, retrospective and prospective entrustment-supervision scales, and entrustment-based discussion. The authors conclude that a shared understanding of the language around entrustment is critical to strengthen bridges among stages of training and practice, such as undergraduate medical education, graduate medical education, and continuing professional development. Shared language and understanding provide the foundation for consistency in interpretation and implementation across the educational continuum.

Keywords: Clinical assessment; outcomes-based; undergraduate education; postgraduate education

Competency-based medical education (CBME), a movement embraced by many education programs for health professions worldwide, is often implemented using the frameworks of milestones, competencies, and entrustment (Licari and Chambers [24]; Bok [2]; Molgaard et al. [32]; Powell and Carraccio [35]). Indeed, entrustable professional activities (EPAs), a prevailing CBME approach using the entrustment framework, have garnered substantial interest worldwide over the past decade (Chen et al. [5]; Carraccio et al. [3]; Duijn et al. [12]; Meyer et al. [29]; Molgaard et al. [31]; O'Dowd et al. [33]; Shorey et al. [41]). An unintended consequence of the rapid, widespread adoption of the entrustment framework and EPAs is the range of interpretations of the central concepts of this approach. Many publications have detailed how individuals or programs are developing EPAs, but limited attention has been paid to precisely defining concepts and principles (O'Dowd et al. [33]; Shorey et al. [41]). The result is a lack of clarity in the language describing the key concepts and principles upon which EPAs are formulated and entrustment decision-making is based (ten Cate [47]; ten Cate et al. [52]; Taylor et al. [43]; Tekian [44]; Melvin et al. [28]). Useful attempts have been made to address this confusion; however, these efforts have often limited their focus to specific areas (Englander et al. [13]; ten Cate [46], [49]; Sklar [42]; ten Cate et al. [51]; Post et al. [34]; Taylor et al. [43]; Tekian [44]). This article aims to offer a more comprehensive clarification of the entrustment construct by (1) proposing a common lexicon for the concepts of entrustment decision-making across the continuum of medical education and (2) advocating a common mental model for the context and consequences of entrustment. We believe clarifying language and more precisely defining terminology as a reference for educators, researchers, and authors contributing to this literature can reduce confusion and advance a much-needed shared mental model. We acknowledge that language is a living thing and no authors can ever have full control over its use. The reason we want to contribute to clarification around entrustment talk is our conviction that language has the power to shape education and assessment practices, and it is this underlying purpose that has motivated us for the current writing. Although this paper has the inherent limitation that the concepts and definitions reflect the authors' views, we believe it is worth noting that we are all members of the International CBME Collaborators group and are both dedicated to this topic as well as experienced in it.

Practice points

• Rapid uptake of entrustment decision-making in recent years has resulted in use of terms without agreed-upon definitions.
• Autonomy — being allowed to act of one's own volition — is given in the context of supervision; 'autonomy or being supervised' is therefore a false dichotomy.
• An individual can be trustworthy in a general sense, but not trustworthy for a specific task.
• 'Entrustable' is a feature of a professional activity, not of a trainee.
• Assessment is generally retrospective, determining what learners were able to do in the past; entrustment decisions are prospective, determining what learners are ready to do in the near future.

The entrustment framework, which considers learning and assessment through the lens of being trusted to perform activities with various levels of supervision, has several strengths. First, the entrustment framework brings care of the patient into the learning and assessment equation by considering how much supervision is, or is not, needed to provide quality care. Second, when using EPAs specifically, the entrustment framework positions clinical supervisors to observe and coach authentic activities of the profession that require the learner to integrate critical competencies. Finally, and perhaps most importantly, EPAs represent actual units of professional practice that individuals can be trusted to do with varying levels of supervision, culminating in the ability to practise without supervision and to supervise others, based on the needs of populations served (Vandewaetere et al. [58]). Thus, for educational practice purposes, one engages in a backward visioning process from the profession's EPAs to identifying and developing learning experiences that will support learners' in their progress toward becoming entrusted with those professional activities.

Fully realizing these strengths of the entrustment framework and EPAs requires a common language for definitions as well as appropriate understanding and use of concepts related to EPAs and entrustment decision-making. Table 1 summarizes proposed definitions and explanations for terms and concepts that provide the foundation for the shared mental model that we build throughout the remainder of this article.

Table 1. Proposed definitions for terms and concepts central to entrustment decision-making and entrustable professional activities.

Early work on EPAs and their related entrustment-supervision scales focused on gradually lessening supervision as care delivery improves to an ultimate level of 'independent practice' (ten Cate et al. [56]). However, the conceptualization of independence has been challenged (Schumacher et al. [39]). The word independence in an educational setting may be used to signify that an individual is not dependent on a teacher or a supervisor, but it may also be conflated with independence from co-workers. Independence should not be confused with the anachronistic notion that a medical professional is an isolated wellspring of skills or wisdom, fixing whatever ails a patient. Even a solo practitioner in a rural area will have a nurse, assistant, or other health care colleagues with whom they can collaborate. They will also have resources that they can access to seek answers to their questions; for instance, they can search the medical literature and other resources on the Internet, and they can connect with consultants and other health care providers via telephone or electronic means. These individuals are not practising independently, which would imply they are left entirely to their own knowledge, skills, abilities, and means. Thus, our preference is to avoid this word altogether and instead speak of unsupervised practice, which is consistent with how health care is actually delivered.

Often supervision is thought of as the antithesis of autonomy, as if they are at opposite ends of the same spectrum (Schumacher et al. [39]). In this framing, an individual can either be supervised or be granted autonomy. We do not support this view but rather align with Schumacher and colleagues who suggest adopting the definition of autonomy set forth in self-determination theory: acting of one's own volition (Ryan and Deci [38]; Schumacher et al. [39]). Granting this ability in the safe setting of being supervised allows individuals to progress from their current level of competence (what they can do) to demonstrating their capability (what they are able to do but have not yet done) (Fraser and Greenhalgh [15]). This not only advances their learning but also positions supervisors to push toward and observe the leading edges of individuals' performance to make determinations about their readiness for less supervision. Autonomy has a legal connotation reflecting accountability and liability, but many early-career autonomous medical professionals need or desire mentorship or coaching to strengthen their high-quality practice (Donahue et al. [10]; Duijn et al. [11]).

Supervision is generally well understood in the clinical setting as overseeing another practitioner's care and actions. However, definitions of supervision diverge, creating complexity (Lingard and Goldszmidt [25]). Clinical supervision has been defined in the British Journal of Psychology as 'the formal provision, by senior/qualified health practitioners, of an intensive relationship-based education and training that is case-focused and which supports, directs, and guides the work of colleagues (supervisees). Its function includes quality control, maintaining and facilitating the supervisees' competence and capability, and helping supervisees to work effectively' (Milne [30]). Kilminster and colleagues define supervision in medical education differently, as 'the provision of guidance and feedback on matters of personal, professional, and educational development in the context of a trainee's experience of providing safe and appropriate patient care' (Kilminster et al. [23]). Both of these definitions suffer from a lack of a more formal or hierarchical role description of supervision (literally 'over-sight'), reflected in common dictionaries. Supervision as clinical oversight is also well elaborated by the work of Kennedy and colleagues in medical education (Kennedy et al. [21]). Combining both the guidance and oversight perspectives of supervision led ten Cate to propose a new definition for clinical supervision that seems a better fit for the purpose: 'The provision of guidance and support in learning and working effectively in health care by observing and directing, when necessary, the execution of tasks or activities to ensure that they are done correctly and safely, from a position of being in charge' (ten Cate [57]).

Trustworthiness ('the quality of always being good, honest, sincere, etc. so that people can rely on you' – Oxford English Dictionary) consists of traits or behaviours that individuals demonstrate, which contribute to others trusting them. We address trust next, as trustworthiness alone is not sufficient to make entrustment decisions.

Kennedy and colleagues report four components of trustworthiness of learners in the context of medical education: (1) knowledge and skill, (2) conscientiousness (i.e. thoroughness and dependability), (3) discernment (i.e. awareness of limitations and willingness to seek help), and (4) truthfulness (Kennedy et al. [22]). Two decades before this work, in the management literature, Mayer and colleagues proposed ability, integrity, and benevolence as key components of trustworthiness (Colquitt et al. [6]). Ability, or knowledge and skill, is obviously inherent to a task, but the other characteristics that determine trustworthiness are not (ten Cate and Chen [49]). This raises the question whether general trustworthiness should be distinguished from task-specific trustworthiness. General trustworthiness can be used to refer to an individual who tends to display features such as honesty, help-seeking, and dependability. However, even this person may not possess the requisite knowledge and skill for a given task. Therefore, they may not demonstrate task-specific trustworthiness. Consider a truthful, conscientious pediatrician who is aware of their limitations. They may be seen as possessing general trustworthiness. They may also possess excellent task-specific trustworthiness for managing a child presenting with an ear infection. However, unless also trained in neurosurgery, they almost certainly have very limited, if any, knowledge and skill for resecting a pediatric brain tumor. Therefore, they lack task-specific trustworthiness for this surgical procedure. This distinction between general and task-specific trustworthiness underscores why general trustworthiness alone is necessary but not sufficient for being trusted to perform a specific task. Similarly, general trustworthiness does not equate with readiness for a specific task.

Trust is both a noun (a psychological attitude) and verb (the act of relying on someone or something) (Castelfranchi and Falcone [4]). Mayer and colleagues' definition of the noun is most cited: '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' (Mayer et al. [27]). More relevant for entrustment than the verb trust is entrusting: 'to assign the responsibility for something valued or important to a person, organization, etc.' (Oxford English Dictionary).

Entrustment is granted by a trustor to another individual (a trustee). While perceived trustworthiness is a primary factor in whether or not another individual is entrusted, the process of entrusting another is nuanced and influenced by other factors. Dijksterhuis, Hauer and others have described five groups of factors that affect entrustment decisions in the workplace: the trustee, the trustor, the task, the context, and the trustee–trustor relationship (Dijksterhuis et al. [9]; Hauer et al. [19]). Thus, a learner who may seem ready for indirect supervision (or who is entrusted with indirect supervision) by one preceptor may not be trusted by a different preceptor who is not acquainted with the trainee. The severity of illness of the patient, the difficulty and risk of the care to be provided, the time of day or night, the experience and trust propensity of the supervisor, and the busyness of the clinical environment all play a role in ad hoc decisions to entrust trainees. Entrustment may or may not happen, even independent of the learner's readiness for it. This underscores the differentiation between trust as a state of mind and entrustment as an act. Trust is necessary but not sufficient to make an entrustment decision. An individual may be trustworthy and have a positive relationship with the trustor but many not be entrusted with a task because of the context and risk involved. Similarly, an individual who demonstrates a high level of trustworthiness may not be entrusted with a specific task because of a trustor's limited inclination to trust and other considerations. These issues have led scholars to separate trust from entrustment decisions (Mayer et al. [27]; Damodaran et al. [8]; Holzhausen et al. [20]; ten Cate et al. [55]).

Complicating the relationship between trust and entrustment, supervisors differ in their trust propensity, or their willingness to trust a person regardless of the context, task, and even risk involved (ten Cate [48]). Trust propensity necessitates consideration of perceived and actual risk. Actual risk is intrinsic to the context and task. However, perceived risk may differ across supervisors and is therefore intrinsic to the trustor. There may be general agreement between actual and perceived risk when risk is high, such as with a difficult intubation of a patient in shock. However, there may be different perceptions of risk based on a preceptor's risk tolerance (determining their trust propensity) and experience level, such that one preceptor may allow learners to do things another preceptor would not let them do. Similarly, trust propensity may vary on the basis of perceived and actual benefits. For example, a preceptor who leaves the trustee alone in the hospital and goes home for the night benefits from this decision. This may also benefit the trustee who desires the opportunity to build confidence through practising with more distant supervision or it may leave an unprepared trustee exposed. In short, the tendency to entrust learners with tasks is influenced by the trustor's perceived risks and benefits of entrusting (ten Cate [48]).

As the word trustworthiness in the workplace brings to mind character traits as well as knowledge and abilities, feedback that one is not yet trustworthy is easily misunderstood because of the nuances between general and task-specific trustworthiness. To address this misunderstanding, individuals often substitute the word trustworthiness with the word entrustable when describing learners. Sometimes, this is operationalized by discussing the level of entrustability that a learner possesses. However, the word entrustable was meant to characterize activities that a learner performs and not to be assigned to the person who is performing them (ten Cate [45], [49]). This misuse has led ten Cate to argue that qualifying learners as entrustable or pre-entrustable (Association of American Medical Colleges [1]; Englander et al. [14]) not only is inaccurate but may also lead to a misinterpretation of ordinal entrustment-supervision scales, moving away from the discrete recommendations or decisions of a level of supervision and instead interpreting the scales as a continuous measure of proficiency (ten Cate [49]). These scales are not intended to be used in this manner. Furthermore, a rating between a supervisor being physically present and being not physically present is illogical. A supervisor cannot be present as a ghostly apparition. Either they are present or they are not present. For this reason, we agree with discouraging the use of scales that describe degrees of entrustability as well as referring to learners as entrustable or pre-entrustable. Rather, we advocate using the term readiness for practice.

The concept of entrustment is stable across stages of training and fields of heath care practice (i.e. readiness to practice without supervision means an individual can work without supervision for a given task whether they are a student, a resident, a veterinarian, or a dentist). However, additional levels of supervision and specificity of language used in entrustment-supervision scales may be warranted for particular stages of training and fields of practice (Rekman et al. [36]; ten Cate et al. [55]). Most often, a generic five-level entrustment-supervision scale is used: May be present but may not practise the EPA; May practise the EPA under direct, proactive supervision, with supervisor physically present in the room; May practise the EPA under indirect, reactive supervision, with supervisor not physically present but quickly available; May practise the EPA unsupervised; May act as supervisor for others for the EPA (ten Cate et al. [56]). This has been further elaborated by Chen and colleagues for undergraduate medical education; they have added additional levels at the lower end where a student may potentially not even be allowed to observe something (Chen et al. [5]).

The need for differing language for entrustment-supervision scales between fields of practice and specialties within a given field is also important to note. As Hatala and colleagues note, many internal medicine tasks are less well-circumscribed than the tasks in procedural specialties such as surgery (Hatala et al. [18]). Much learner activity in internal medicine is closely monitored but not directly observed (Hatala et al. [18]; Melvin et al. [28]). Furthermore, Weller and colleagues' perception of the core entrustment question in the minds of anesthesiology attendings is: 'Can I leave the operation theatre?' (and leave the resident in charge) (Weller et al. [59]). Finally, in the pediatric community in the United States, the language of the popular Chen entrustment-supervision scale has been adapted to meet the needs of, and align with, the mental model of entrustment and supervision in that community (Schumacher et al. [40]).

These language issues, sometimes stemming from differences in mental representations of supervision between field and specialties, cannot be resolved with a one-size-fits-all solution. We believe that the generic framework of levels of supervision should remain, but within that framework adapations are beneficial. Guidelines around how to derive contextually suitable entrustment-supervision scales cannot be offered in this paper, but educators and scholars in various fields and specialties can elaborate for their field or specialty the most effective language and the required number of supervision levels.

Entrustment-supervision scales have been used in two ways: (1) to report how much supervision, guidance, and/or assistance was needed and provided for a witnessed task, as in the case of the O-SCORE scale (with levels specified as: I had to do, I had to talk them through, etc.) (Gofton et al. [16]; Rekman et al. [37]) and (2) to recommend how much supervision a learner requires in the execution of specific activities in the future, including as-yet-undefined situations. The first is a scale that serves to report on observations that have occurred (i.e. retrospectively) and that lead to a judgment of proficiency. The second requires additional inference by the observer to render a prospective entrustment decision. Retrospective evaluations are situated in a specific context with known (albeit individually variable estimates of) risks, benefits, and outcomes, judged in hindsight. With entrustment decisions, which are inherently prospective, the contexts, risks, benefits, and outcomes are unknown. We suggest using the term entrustment-supervision scale, as we have done in this paper, to stress that this scale not only reflects how much supervision was provided in the past but also includes a recommendation for a future level of supervision, implying entrustment for the future.

Prospective entrustment decisions should be based on considering both general and task-specific trustworthiness. This accounts for a learner's general tendency to be honest, conscientious, and seek help when needed (general trustworthiness) as well as a learner's knowledge and skill for the tasks they will probably be called upon to perform (task-specific trustworthiness). This latter consideration should consider judgments about a learner's abilities to extrapolate from the known to the unknown in applying their knowledge and skills.

When considering general trustworthiness in making a prospective entrustment decision, ten cate and Chen have delineated five groups of factors on the basis of studies of learner charcteristics that enable entrustment with tasks in health care: capability, integrity, humility, reliability, and agency (ten Cate and Chen [55]). These align well with the trustworthiness constructs we discussed previously. If we consider these features when making an entrustment decision, if and when an individual is entrusted beyond their demonstrated competence, the risk involved in that decision and the future situation is mitigated by the entrustment decision leaning heavily on the individual's general trustworthiness. To further assess risk, ten Cate and Hoff have proposed including entrustment-based discussions as part of an assessment toolkit (ten Cate and Hoff [53]). Such discussions are multiple brief conversations that include discussing awareness of risks and complications as well as imposing 'what would you have done if...' questions to gauge the risks of decreased oversight.

Extending beyond strictly definitions to the potential ability to put defined terms into practice, we will end by reflecting on the fact that some entrustment recommendations cannot lead to an actual entrustment decision. For example, medical students and residents often require some level of oversight that is specified in standards, laws, or both. Even if a learner is completely competent and ready to work unsupervised, law may prohibit this if the person is not licensed or certified for a specialty. This places a limit on the scope of entrustment decisions that can be enacted during training under current legislation. If individuals are deemed ready to be entrusted with unsupervised practice, should that be permitted? Recent reports have noted the 'worrisome seniorization' of tasks in health care and consequent decrease in responsibilities for trainees (Halpern and Detsky [17]; Dacey and Nasca [7]). This may jeopardize the readiness of graduates to perform critical patient care tasks (Mattar et al. [26]; Donahue et al. [10]). However, allowing decisions about readiness to be entrusted with unsupervised practice to be made for individuals who are still in training places enormous responsibility on the medical education community to define when there is sufficient evidence for valid entrustment decisions. While lofty, we believe this is a goal that we should seek to achieve. Achieving this would also require trainees to be credentialed within a practice or hospital so that they could be permitted to work without supervision. Without taking the step of enabling enactment of entrustment decisions before graduation, is there ultimately sufficient meaning in those decisions? It is probably less risky to make entrustment decisions within the confines of a training program, where there are safety nets beyond supervisor oversight (nursing staff, colleagues, etc.) than to wait until someone goes into practice, where the possibility of a safety net lessens.

Entrustment is an assessment framework that asks clinical supervisors to make judgments about a trainee's care delivery on the basis of what they do while supervising learners in the workplace. Moreover, it brings the patient and the provision of safe, quality care into consideration. If we can agree on a shared language of entrustment, these characteristics of this construct present the opportunity to embrace a more holistic education of trainees. However, the full benefit of entrustment as a useful framework will only be realized if it is used with reasonable consistency within programs, between programs, and across the continuum of training and practice. This requires a common understanding of the language of entrustment, which we have attempted to provide in this paper.

The concept for this article originated during a two-day forum meeting of the International Competency-Based Medical Education (ICBME) Collaborators in Ottawa, Canada, July 2019.

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the article.

Prospective entrustment-supervision scale: A tool/mechanism for estimating or recommending readiness for future practice. Prospective scales include anchors such as: 'I judge (or estimate) this learner to be ready for indirect supervision,' or, if used summatively, 'From tomorrow, this learner is allowed to practise this EPA under supervisory oversight only.' Prospective entrustment is a tautology, as all entrustment decisions are prospective.