Determinants of undergraduate medical students’ satisfaction with clinical supervision: A cohort study in a longitudinally structured sixth year clinical placement

A clinical supervisor acts as a gatekeeper by carefully steering students’ access to tasks to ensure the quality of professional services provided by trainees to patients, as a trainer to develop their competence, and as a mentor to support learning in the workplace [1]. Although the workplace is theoretically an authentic learning environment for complex professional skills [2], it does not automatically provide an ideal learning setting for medical students, and learning experiences vary markedly between students and clinics [3, 4]. A curricular structure with defined learning objectives for participation in the clinical team and patient care under close supervision can promote learning in the clinical workplace [1]. Nevertheless, even in best-practice workplace learning programs students express concerns about making mistakes, staff neglect, and poor adjustment to the clinical setting [5]. Only about one third of undergraduate medical students (UGMS) report being very satisfied with their clinical supervision, around 40% report achieving their desired progress during the placement, and only about 50% would recommend their placement department to a friend [6]. From the program providers’ perspective, this lack of learners’ positive subjective reactions should be a cause for concern. An association between positive subjective reactions to working or learning and favorable employee outcomes [7‐9] has been established. These outcomes include less exhaustion, fewer turnover intentions [7, 9], increased job engagement [7], or interest in a clinical field [10]. Even the absolute clinical workload remains unidentified as “the core issue” in the lack of well-being among obstetrics and gynecology trainees and faculty physicians. Instead, “work satisfaction influences well-being more than workplace environment” [8, p. 4]. As such, evaluating subjective reactions beyond objective learning outputs, behavior changes, and long-term outcomes has been an important pillar of training program evaluation since Kirkpatrick suggested this fourfold approach for program evaluation in 1959 [11].

To improve satisfaction with supervision, placement duration [12], schedule feasibility, learning material quality, examination fairness [13], and opportunities for participation in clinical activities [14] have been investigated as determinants of satisfaction. Still, supervisor engagement contributes the most to student satisfaction with placement [15]. Good supervisors can alleviate students’ concerns and foster their learning and sense of belonging [16, 17]. Consequently, parallels between patient-centered care and learner-centered supervision have been drawn [13, 18]. Thus, we propose developing a quantitative measure expressing students’ experiences of how their supervision needs were met based on an empirically tested theory from motivation psychology.

According to self-determination theory (SDT; [19]), a person needs to experience competence, autonomy, and social relatedness in their social environment to be motivated, satisfied, and to perform well. A central assertion of SDT, as applied to clinical supervision, is that supporting students’ basic psychological needs, here coined learner-centered supervision, promotes motivation, learning, and satisfaction. To expand our understanding of the mechanisms of learner-centered supervision and satisfaction in the workplace, we will use the lens of SDT to introduce a measure to quantify students’ learning experience, the supervision deficit index (SDI). The index is based on supervisory activities that students indicate as being most helpful for their learning. The discrepancy between desired and received activities reflects the individually experienced supervision deficit.

Consequently, we will address the following question: How do students’ perceptions of a deficit in learner-centered supervision contribute to explaining general satisfaction with supervision? Based on SDT, it is hypothesized that experiencing a deficit negatively impacts general satisfaction.

Young and inexperienced UGMS are in need of high-quality supervision [16] during their clinical placement, as their study behavior is strongly influenced by supervision quality. To help define high-quality supervision, we draw on the SDT basic psychological needs concept. Meeting students’ supervision needs is an important aspect of supervision quality, as seeing one’s needs met strongly impacts one’s satisfaction. The extent and quality of supervisor behavior require consideration when evaluating students’ subjective supervision experiences [30]. Expanding this idea, we present an SDI to represent the extent to which students’ supervision needs were met in a single measure. With the currently implemented choice procedure, a reliability of 0.735 was observed, with satisfactory psychometric quality for most of the 26 items.

Following the SDT’s basic psychological needs concept, experiencing learner-centered supervision was confirmed as the main predictor of satisfaction with clinical placement. General clinical experience, clinical field, hospital size, and gender did not contribute to satisfaction above the SDI. Our results empirically reinforce previous opinions [18]. Structuring a teaching encounter with a student as one would structure a patient encounter, with a strong focus on personalizing the encounter, inviting a shared presence, engaging students, checking their understanding, and building trust is the key to the “quality of interactions between residents and medical students” [31, p. 348]. Consistent with our findings are results on favorable resident teacher attributes [32]: Besides “having a strong knowledge base,” “tailoring teaching to learner’s level”, and being “approachable” are the top favored teacher attributes indicated by over 70% of students out of five Canadian medical schools. Our findings corroborate those reporting that fulfilment of medical students’ basic psychological needs reduces their stress [33].

On average, UGMS favored the same seven supervisory activities in both clinical fields, where five are training activities, one is a gatekeeping and one a mentoring activity. This result seems to corroborate the reporting that satisfaction of competence, but not autonomy or relatedness, predicted increased students’ resilience [34]. About half of the UGMS favored the two training activities, “discuss individual patients” and “discuss away from the bedside”, which postgraduate medical students and their supervisors also indicated as the most helpful [21]. The other five activities were regarded as most helpful by an even smaller share of UGMS in our study and also regarded as less helpful during postgraduate training.

As such, the following practical implications for clinical supervisors and program providers can be derived: (a) students’ supervision needs are generally similar between internal medicine and surgery placements, despite subtle differences, but show high variability between students. Although a one size fits all approach to supervision is nonexistent, identifying opportunities for participation in the daily clinical routines and allowing attending UGMS to experience the three most popular training activities regularly during their stay is a good strategy to prepare oneself as a clinical supervisor. (b) Student factors such as their prior general clinical experience and gender, or external factors, such as hospital size and clinical field, do not explain satisfaction above the SDI in this study. Still, according to a previous study students’ supervision expectations seem to be shaped by prior experiences [35]. A successful supervision approach requires both parties to understand each other’s expectations, needs and resources and to respect boundaries. Supervisors engaging in good student-onboarding practice explore students’ wishes and expectations. They also should openly emphasize their responsibility of gatekeeping to provide training opportunities for students and their struggle to balance patients’ and students’ needs [1]. It may be encouraging for supervisors to know that approximately two third of UGMS report their supervisors to engage fully in gatekeeping activities such as ensuring student and patient safety [35]. In addition, outlining how students can and should actively contribute to the supervisory process, given the boundaries of the respective setting, contributes to understanding each other. (c) Program providers seeking to evaluate UGMS satisfaction with their placement might use a group level SDI to monitor students’ perceptions of learner-centered supervision while further gaining insights into students’ supervision preferences and chances thereof. This approach would be beneficial for evaluating the effect of faculty development on clinical supervisors, such as sharing and refining best-practice examples.

Privacy protection precluded data collection at the department level in the hospitals. Data could also only be collected after all three parts of the clinical placement had been completed. Despite tailoring the paper-based survey procedure to the time available, many students did not follow the instructions to provide five favorites. Thus, more unique missingness patterns with 0 or perfect scores emerged, for which person location parameters could neither be estimated nor extrapolated. Optimizing and shortening the set of activities will facilitate the completion of the choice procedure as instructed. Future studies might consider using our item quality results to shorten the list of activities and modify or abandon the choice procedure to reduce the number of observed missingness patterns. Studies should either control for sequence effects when collecting data after completing all parts of the clinical placement or strive to collect data on supervisory experience shortly after each part of the training. Providing the survey online may contribute to all suggestions for improvement.

Providing learner-centered supervision promotes students’ motivation and thus their learning [8]. With the SDI, we presented in principle how to reliably and validly quantify the lack of learner-centered supervision, which might be a risk factor for falling short in learning. Using the index, we showed that addressing and fulfilling students’ supervision needs is more important for student satisfaction with their clinical placement than context variables such as hospital size or their previous general clinical experience. We replicated the results on UGMS, who, like postgraduate trainees, preferred competence support above autonomy and relatedness support in workplace-based learning. With the SDI, the quantitative evaluation of interventions targeting improving students’ and supervisors’ satisfaction with supervision is possible. Clinical supervisors and program providers might be interested in considering how their students and potentially future colleagues experience supervision above evaluating the students’ learning output [11], as those experiences influence their career choice [10, 36].