Building a doctor, one skill at a time: Rethinking clinical training through a new skills-based feedback modality

Microskills are derived from the psychology [13, 14], negotiation [15‐17], and business [18] literature and are unique in their ability to elicit targeted, relevant feedback to promote trainee development. Introduced 50 years ago, microskills (called microcounseling in psychology) transitioned the mental health education field from a theoretical to a skills-based framework. Microskills are specific and isolated actions (e.g., summarizing content back to the patient) that can be observed and repeatedly practiced by counselors, distilling complex psychological principles into understandable and repeatable skills [19]. The microskills framework remains widely used in psychology graduate programs and is the basis for numerous textbooks [20].

Decades of studies have confirmed the effectiveness of the microskills framework. Microskills-trained counselors have demonstrated improved basic and higher-level counseling skills (e.g., asking open-ended questions), established superior therapeutic alliances with patients (measured through patient satisfaction surveys), and displayed empathy better [14, 21]. These findings are summarized in several meta-analyses reviewing microskills effectiveness compared to other psychological training approaches, demonstrating statistically significant effect sizes of microskills across multiple dimensions (i.e., microcounseling is effective for teaching a variety of therapeutic communication skills) [21, 22].

The microskills framework of distilling processes into discrete, repeatable actions has been applied to other fields of study. Microskills were efficacious in teaching children problem-solving skills, adolescents interpersonal skills, and couples relationship skills [21]. Negotiation experts have distilled complicated tactics into microskills that individuals can practice, becoming a primary method for teaching negotiation today.

The idea of repeatedly engaging with skills is not new to medicine; studies have shown that residents learn and grow best when they engage in deliberate practice of highly targeted skills [23, 24]. Deliberate practice involves the attainment of expertise through (1) repeating discrete, well-defined tasks, (2) receiving immediate feedback on actions, and (3) re-engaging with the task with an intent and ideas to improve [25]. Tasks need to be specific enough so that on repeat attempts trainees can consistently improve small aspects of their performance without becoming confused by other tasks or larger goals. Each task must also be easily observable to supervisors, who can then provide targeted feedback.

To date, efforts at breaking down larger clinical functions into skills have involved trainees performing iterative and quantifiable tasks, such as reading electrocardiograms or practicing surgical skills on simulators [25, 26]. These examples focus on procedural skills in controllable circumstances with measurable outcomes (e.g., gaining speed with knot-tying). Few studies have explored deliberate practice in less controllable domains of medical training, such as patient presentations or clinical management, which involve multiple skills and cannot be easily reproduced. Perhaps the closest analogue to microskills in medicine is clinical communication skills training (CST), which teaches medical professionals verbal and nonverbal skills (i.e., interview structuring, information gathering) for patient encounters [27]. Though meta-analyses have found CST to be effective, it focuses on a specific aspect of the patient encounter (one-on-one conversations), which is a small part of a resident’s overall clinical training [27].

While the number of discrete clinical skills required for competent clinical practice may be overwhelming to trainees, organizing skills within a familiar context may overcome this challenge. Situated learning theory posits that mental schemas are formed based not only on what a learner commits to memory, but also on where and how the learning takes place (i.e., the physical and social environment, as well as the specific circumstance) [28]. Chess experts may have difficulty remembering strategies or piece positions from random boards, but they display uncanny precision when describing past matches [29]. Similarly, medical learners may struggle to learn from and practice a skill if they do not know the appropriate context for that skill and how subtle changes in a given situation may demand different actions.

As a result, the cognitive constructs trainees use to retrieve and practice skills need to be rooted not just in a clinical context (i.e., the medical circumstance in which an activity should be performed), but also in a situational context (i.e., location and time during a routine day in which the skill is utilized). If supervisors are not aligned with learners on how and where a skill should be demonstrated—both the clinical and situational contexts—feedback may suffer. As noted previously, milestones are context-independent and are meant to be applied in different settings. EPAs, on the other hand, are grounded in a clinical context but not in a specific situational context [30].

Consider how the skills within the pediatric EPA “manage patients with acute, common diagnoses” [2] can be demonstrated through different situational circumstances: patient presentations (e.g., expanding on a differential on rounds), bedside team huddles (e.g., escalating respiratory support based on expected illness trajectory), and family interactions (e.g., counseling families about testing options). Apart from referring to time/location circumstances, situational context can also delineate skills with additional descriptors: is this a presentation for a new patient or an existing patient that has been in the hospital for several days? Does the skill involve communication within the laboratory/imaging section of the presentation or the assessment section?

Microskills used for specific, actionable feedback targeting opportunities for continued growth can be developed for any specialty and any level of training. In order to demonstrate how microskills can be derived from existing assessment modalities and to emphasize the importance of incorporating user-centered design into the framework, we created a set of microskills for pediatric interns on a hospital medicine service using ACGME pediatric milestones and the American Board of Pediatrics’ (ABP) EPAs for general pediatrics [1, 2, 40]. Though we chose to use milestones and EPAs due to their widespread acceptance as assessment modalities, microskills can be derived from a variety of assessment frameworks (e.g., CanMEDS). To develop the microskills, the ACGME pediatric milestones, ABP EPAs, and ABP hospital medicine fellowship EPAs [41] were reviewed to determine those that were most appropriate for application in hospital medicine (Table S2, found in the ESM, lists the milestones and EPAs used for microskill development).

As we created microskills based upon our stated criteria, we found that some elements of existing assessment modalities were narrow enough to become microskills without significant adjustments, while others required more modifications. The EPA “ensure complete handoffs using a standardized template when patients are transitioning from one care provider to another” [2] served as a microskill almost verbatim, while the EPA “synthesizing lab tests into proper patient care” [2] was distilled into the ways an intern may engage with the skill (e.g., incorporating lab tests into presentations vs. written assessments vs. comparing patient labs with standards from an evidence-based literature review). Some EPAs required transformation from goals-based to observable skills-based actions. For example, an EPA focused on “resuscitating [the] patient and then triage to align care with severity of illness” [2] contains knowledge and management-based curricular components (e.g., distinguishes between respiratory distress and failure). However, a resident may struggle to explicitly apply this EPA to their day-to-day routine in an easily observable way. Distinguishing between respiratory distress and failure requires a specific context in which it can be applied, such as during a bedside presentation on rounds or through written notes that contain contingency planning.

As part of developing microskills mapped to milestones and EPAs, we created a user-centered framework focused on pediatric interns’ hospital medicine workday and the skills they use to navigate that day. User-centered design employs empathy to better understand the needs of the end-user and creates systems that complement the subject’s natural tendencies [42, 43]. By applying user-centered design to better understand the learner experience and through discussions with residents, we were able to imbue microskills with the relevant clinical and/or situational context, as well as construct a more naturally organized system for learners to select and practice skills. This process involved generating a user journey map—a tool used commonly in healthcare design to show how patients experience disease and interact with their care teams (see ESM, Fig. S1a) [44]. A user-centered skills map of situational context was also created based on the patient’s hospital experience and intern-patient interactions to demonstrate the possible skills a prototypical intern could employ at various stages of a patient’s hospitalization (see ESM, Fig. S1b). A clinical context map was not created, given that this context is already embedded in EPAs and would thus transfer easily onto the relevant microskills.

These user-centered maps were then combined with the extracted relevant milestones and EPAs to form situationally anchored microskills umbrellaed under a table of contents (see ESM, Fig. S2). The microskill table of contents mirrors the cognitive schema an intern uses to organize the workday, thereby lending itself to more natural microskill selection and adoption. The table of contents is an amalgamation of both the user-centered maps, incorporating unique aspects of the intern’s workday (e.g., pre-rounding, daily notes) as well as skills that surround the patient experience (e.g., delivering difficult news, family-focused presentations).

Groupings of skills that interns could situationally demonstrate to supervisors were separated into domains in the table of contents. For example, patient presentations are mentioned throughout multiple milestones [1] (e.g., PC6: make informed diagnostic and therapeutic decisions that result in optimal clinical judgment) and EPAs [2] (e.g., EPA4: manage patients with acute, common diagnoses in an ambulatory, emergency, or inpatient setting) and naturally align with both user maps. As a result, patient presentations became an independent microskill domain. However, even within patient presentations, an intern must employ different types of skills in various situational and clinical contexts, from integrating pathophysiology into everyday assessments, to creating contingency plans for acutely ill children, to relating to families in easy-to-understand language. Taking this into account, the categories within patient presentations were organized both by the features of presentations that the intern may be working on (the situational context—concise histories, tailored plans) as well as the types of patients discussed in the intern’s presentation (clinical context—new vs. multi-day vs. acutely ill patients).

An example of the many domains of competence (DoC), competencies, milestones, and microskills that inform a single EPA can be found in ESM, Fig. S3. (See ESM, Fig. S4 for an expanded example of the connections between entrustable professional activities (EPAs), milestones, and microskills).

While incorporation of microskills into practice will vary by institution, level of training, and clinical rotation, here we outline how a pediatric intern could use microskills on a hospital medicine rotation. When a new supervisor begins clinical service with the resident team, the intern would select one to three microskills (from the library of previously developed and disseminated microskills) to practice and be observed for the upcoming week. Though the intern will ultimately choose the skills he/she will practice, this should be a collaborative process with input from the supervisor and senior residents, considering skills the intern practiced in previous weeks. Throughout the week, both the intern and the supervisor know which skills are being practiced, thereby allowing the intern to refine the skills, as well as provide an opportunity for the supervisor to give real-time feedback on observations (or take notes to provide feedback at a later time). As one example, supervisors already accompany and observe trainees on clinical rounds, which enables a natural incorporation of the microskills framework. Supervisors can choose to deliver coaching in between trainees’ patient presentations or find time later in the day for more involved feedback. If supervisors find themselves pressed for time during or after rounds, they can consider integrating microskills feedback into their workflow later in the workday, such as by encouraging trainees to practice skills in follow-up conversations with patients.

Prior studies demonstrate that unless trainees have an expected forum to ask supervisors for feedback, they will instead forgo asking for feedback altogether [34]. Awareness by both the intern and supervisor of the chosen microskills at the beginning of the week increases the opportunity for feedback touchpoints and reduces the social anxiety trainees experience when asking for feedback out of context [34]. Additionally, when trainees know that the feedback they receive is based on direct observation, they are more likely to accept its accuracy and incorporate changes into practice [37]. At the end of the week, the intern, supervisor, and senior residents may review the intern’s chosen microskills, supervisor observations, and create action plans for intern skill improvement. This is also an opportunity for the supervisor to suggest new or repeat skills for the intern practice for the following week.