Background

In the undergraduate medical curricula, there is a need of opportunities for students to practice prescribing [1]. Prescribing encompasses a range of activities from performing a consultation, identifying a need for drug therapy, selecting and prescribing the appropriate drug to being involved in the subsequent management of the patient [1, 2]. McLellan et al. suggest that the way to teach and provide training in prescribing skills is to design interventions grounded in a real-life context, so that students can be observed and evaluated in the context of their (future) workplace [3]. Learning in the (future) workplace (Workplace Learning) is “as old as medicine itself” however does not necessarily contain specific responsibilities for students [4]. Giving students a feeling of responsibility for patient care makes their clinical experiences more ‘real’ and legitimate, and might stimulate student motivation [5]. Such enrichment of responsibility (for patient care) is thought to be an important factor to improve the training of rational prescribing skills of medical students [6]. This combination of context/workplace learning, early clinical experience, and sense of responsibility has been described as learning by doing [7].

Based on this concept, a Learner-Centered Student-run Clinic (LC-SRC) was started at the VUmc School of Medical Sciences in 2013 [8, 9]. The LC-SRC is a learner-centered project, as opposed to regular SRCs that primarily focus on providing (free) care [7]. In the LC-SRC students get the opportunity to train themselves in complex competencies such as patient communication, therapeutic reasoning, and prescribing in a real context. The LC-SRC concept and development is based on the conceptual framework of learning by doing, as an example within the more general experiential learning theory by Kolb [7, 9, 10]. Besides the experience itself, its timing and the attending responsibilities of the clinical experience are important. Experiences should be real and legitimate for optimal learning effects and involvement [4, 5, 11].

Student motivation is in general a neglected aspect in the designing of medical curricula [12]. In spite of description of the best principles for doing this [13], few initiatives consider and measure the effect of interventions catered to enhance student motivation in medical education [14], especially so in undergraduate medical curricula. The theoretical framework for this study is based on the Self-determination Theory (SDT). According to this theory, motivation can be classified into intrinsic and extrinsic types; the intrinsic motivation originates from within oneself, and extrinsic originates from external factors [15,16,17]. An example of intrinsic motivation is to learn to prescribe out of genuine interest and the desire to help one’s patients; an example of extrinsic motivation is to learn about certain drugs for an exam one has to pass. Intrinsic motivation depends on the fulfilment of three basic psychological needs, namely, autonomy, competence, and relatedness [15]. It is considered the best form of motivation to promote in-depth learning and to improve performance and competence in learning outcomes [16,19,, 18–20]. Based on SDT, an individual is never thought to be exclusive intrinsically or extrinsically motivated and motivation is different for different activities. Furthermore, both intrinsic and extrinsic types of motivation are always present in differing levels, which can be influenced. The key feature in the transformation of extrinsic to intrinsic motivation is internalization. The level of internalization differs across the SDT continuum (see Fig. 1) [21]. Internalization itself is stimulated by similar components that stimulate intrinsic motivation, being autonomy, competence, and relatedness [22]. The higher the level of internalization, the more autonomous is the motivation. Autonomous motivation is calculated as an average of the scores on score identified regulation and intrinsic motivation, whereas controlled motivation is calculated as an average of the introjected regulation and external regulation scores (see Fig. 1).

Fig. 1
figure 1

The theoretical self-determination continuum (from left, Amotivation (least autonomous) to right, Intrinsic motivation (most autonomous), and the position of the different motivational sub-types of motivation as measured with the AMS (e.g. Amotivation, External Regulation, Introjected Regulation and Identified Regulation). *Integrated Regulation is not measured in/with the AMS questionnaire and is therefore no part of the calculated Autonomous motivation subscale. The IMI was used to study intrinsic motivation in depth, see Table 1

Full size image

An early involvement in clinical practice combined with a high level of responsibility makes the LC-SRC an environment that can fulfil all three basic psychological needs for students intrinsic motivation and internalization: autonomy by giving students responsibility of patients, competence through feedback from supervisors and confidence in handling patients and relatedness through working in teams of peers, near-peers and supervisors [18]. Therefore we hypothesize that participation in LC-SRC will stimulate the intrinsic and autonomous motivation of students for this type of (pharmacotherapy) teaching and learning [12,16,, 15–17, 23].

Since the LC-SRC was designed to stimulate intrinsic and autonomous motivation, and thereby the competence of students, our research questions were:

  1. 1)

    What type of motivation do students have for this educational innovation?

  2. 2)

    Does the motivation for medical education change after participation in this innovation?

  3. 3)

    How does this innovation influence students’ proficiency in CanMEDS competencies?

Methods

Setting

The regular VUmc curriculum (6-years) consists of 3 years of preclinical education (bachelor degree), followed by 3 years of clinical education (master/medical degree). This study was performed within the extracurricular LC-SRC project of the VU University Medical Center, Amsterdam [8, 9]. The LC-SRC project for 1st to 6th year students focuses on early clinical experience with responsibility for real patients. Compared to regular clerkships, students are in the lead in the LC-SRC, they are the principal contacts for “their” patients, and are responsible for patient care including follow up. A key feature of this patient care is the proposition of a reasoned and customized (pharmacotherapy) treatment plan [9].

The students work in teams and are jointly responsible for outpatient consultations within the Department of Internal Medicine, with real patients (who have medical insurance), under supervision of an internist. Teams changed during the course of the study, based on the availability of individual students. The student teams prepare their consultations, based on the (electronic) medical record, and are encouraged to read more about the medical conditions of “their” patients. In addition, students can attend interdisciplinary discussions (e.g. radiology and microbiology), and consult nurses, administrative personnel, and medical doctors from other disciplines. Each student has a specific role and responsibility during the consultation. The third-year student leads the team, performs the consultation, and is coached by the fifth-year student. The first-year student complements with questions and makes annotations for the medical record. Patients are requested to visit the LC-SRC for follow-up after their first (regular) consultation with a resident. If appropriate, follow-up consultations to monitor treatment are planned, whenever possible, with the same student teams, to stimulate longitudinal learning. The consultations take place in the outpatient clinic of our hospital. After each consultation, the supervising internist provides students with feedback on their performance.

Participants

Participation in the LC-SRC project is extracurricular and voluntary, students were invited to apply before/during a regular lecture by sending a letter regarding their expectations and their experiences in healthcare. All students who participated in the LC-SRC pilot from March to July 2013 (n = 31) were invited to take part in the present research project. They were sent an e-survey before and after their participation. All mini-Clinical Evaluation Exercises (mini-CEX) of consultations in the LC-SRC, within this period, were included.

Measurements - motivation

Motivation was assessed with two standardized validated questionnaires, the Academic Motivation Scale (AMS) [24,25,26] and the Intrinsic Motivation Inventory (IMI) [27,28,29]. The AMS was used to differentiate between intrinsic and extrinsic motivation for studying at the medical school [24]. Intrinsic motivation for this particular project was measured using the IMI-subscales Interest, Usefulness, and Perceived Choice (Likert scale of 1–7). The Interest/Enjoyment subscale is considered the primary self-report measure of intrinsic motivation [30]. Furthermore, together the IMI-subscales are related to the three psychological needs, which are all important components within intrinsic motivation, and moreover to stimulate integration/regulation of extrinsic motivation into autonomous motivation [15, 22, 27, 28, 30]. The AMS was completed twice, pre and post participation. The IMI was completed once, post participation. Written feedback was collected by means of open questions in the post-participation questionnaire.

Measurements - competencies

The CanMEDS competencies framework [31] was chosen to evaluate clinical competencies, as both students and supervisors are used to the competences described within this framework, from their experience within the regular medical curriculum. A post-participation questionnaire was used to evaluate students’ perceived improvement in CanMEDS competencies. In order to objectively measure students’ CanMEDS competencies, these were evaluated by faculty from internal medicine, using mini-CEX, to grade and provide feedback for student teams after each consultation. These CanMEDS mini-CEX were regularly used by faculty to evaluate clinical competence of students in their regular clerkships. The seven CanMEDS competencies were scored on a 5-point Likert scale (compared to the level of a junior doctor, 3 meant achieving a junior doctor level, >3 better, <3 worse) (See Additional file 1, CanMEDS mini-CEX). Supervisors were questioned about their opinion on the improvement in medical knowledge, communication, clinical reasoning, and pharmacotherapeutic knowledge/skills of students after their participation in the LC-SRC. Figure 2 indicates which tests, questionnaires and assessments were performed at which time point.

Fig. 2
figure 2

Measurements of motivation and competence in time during the study. The pre participation questionnaire consisted of the baseline characteristics and the AMS questionnaire as shown in Table 1. The post participation questionnaire was longer and consisted of the AMS and the IMI (both provided in Table 2) and evaluation questions. These evaluation questions encompassed the students’ perceived improvement on their competencies (Fig. 4) and their reflections regarding the LC-SRC (Table 3)

Full size image

Analysis

Quantitative data were imported in SPSS (IBM, version 20.0). The AMS subscale scores were calculated (see Table 1). Autonomous motivation was calculated as the mean scores of intrinsic motivation and identified regulation subscales of the AMS. Controlled motivation was calculated as the mean of introjected regulation and external regulation [32] (see Table 1 and Fig. 1). Reliability was assessed with Cronbach’s alpha. Differences between motivation in subgroups (male vs. female and preclinical vs. clinical students), were analysed with the non-parametric Mann–Whitney-U-test given the likely non-normal distribution in these small samples. The difference between motivation before and after participation, measured with AMS (ordinal variables), was analysed with a student’s paired t-test. This parametric test was considered feasible given its high resolution (it could be considered as interval data as most subscales range from 4 to the maximum of 28), and the use of parametric tests for 5- or 7-point Likert scales [33, 34]. Interpretation of statistical significance is based on a Bonferroni correction for multiple testing. Descriptive statistics were used to report mini-CEX outcomes and student-perceived improvement in CanMEDS competencies. Pearson correlations were calculated for all motivational measures (IMI and post-participation AMS subscales) and the student-perceived improvement in CanMEDS competencies. The qualitative data (open questions, written feedback and comments) were analysed using content analysis [35]. Two authors (TS and RD) read and interpreted the students’ feedback and comments simultaneously and resolved differences through discussion and consensus. Identified themes were discussed and agreed upon within the full research team.

Table 1 Motivation questionnaires used in this study, items indicated with (R) are reverse scored items
Full size table

Ethical considerations

The institutional review board of the VU University Medical Center approved the research proposal, deeming that it did not fall under the scope of the Dutch Medical Research Involving Human Subjects Act (WMO) (ID 2013/364). Nevertheless, all stakeholders were informed about the study in advance, gave oral and/or written consent, and participated on a voluntary basis. Final responsibility for clinical decisions was at the supervisor level. The data were analysed anonymously and at a group level.

Results

During the pilot period between March and July 2013, 31 medical students (11 first-year, 10 third-year, and 10 fifth-year students) performed 31 consultations. The students performed the consultations in teams, and the individual students participated on average for two half-days (range 1–5). Cronbach’s alphas for reliability were 0.95, 0.92, and 0.59 for the IMI Interest, Usefulness, and Perceived Choice subscales, respectively (Fig. 3). The reliability score for the Usefulness subscale was higher than that reported earlier [36]. Cronbach’s alphas for the reliability of the AMS subscale of Controlled Motivation, Autonomous Motivation and Amotivation were 0.86, 0.85, and 0.69, respectively, before participation and 0.86, 0.90, and 0.81, respectively, after participation (Fig. 3). These reliability scores were consistent with those reported earlier [37, 38].

Fig. 3
figure 3

Reliability as tested with Cronbach’s alpha of used motivational scales

Full size image

Participation outcomes on motivation

Twenty-nine students, out of which 25 were females, completed the IMI questionnaire (post-participation) about intrinsic motivation to participate in the LC-SRC (response rate 93.5%). They scored a mean of 6.20 (SD 0.67) on the Interest/Enjoyment subscale. The mean scores for the Usefulness and Perceived Choice subscales were 6.02 (SD 0.81) and 5.93 (SD 0.72), respectively (Table 2). We found no significant differences in the scores of the IMI-subscales between male and female students.

Table 2 motivational scores before and after participation in the LC-SRC
Full size table

In 25 paired cases (80,6% of participants), of which 20 were female, motivation was measured with the AMS before and after participation. Intrinsic motivation for attending medical school was 5.37 (SD 0.69) before and 5.31 (SD 0.69) after LC-SRC participation (paired t-test p = 0.532). Corresponding before and after scores were 3.98 (SD 1.01) and 4.23 (SD 1.03), respectively, for Controlled Motivation (paired t-test, p = 0.055), 5.51 (SD 0.59) and 5.39 (SD 0.64) for Autonomous Motivation (paired t-test, p = 0.217), and 1.21 (SD 0.42) and 1.20 (SD 0.40) for Amotivation (paired t-test, p = 0.788) (Table 2). We found no significant difference on eventual change in motivation (before-after) within either the subgroups of male and female students or between the pre-clinical students (1st and 3rd year) and already clinical students (5th year).

Participation and competence

Twenty-seven students (response rate 87.1%) evaluated whether their CanMEDS competencies had improved (Likert scale 1–5, strongly disagree to strongly agree). Students considered that their proficiency in the CanMEDS competencies of “Collaborator”, “Communicator”, “Academic”, and “Medical expert” had improved after participation (Likert score ≥4.0). Faculty staff evaluated their clinical competence as being at a junior doctor level (Likert score 3.15 (SD 0.60) on mini-CEX at the team level) (Fig. 4). Additionally, in the post-participation questionnaire three out of four supervisors reported that in their opinion the communication skills, medical knowledge, and clinical reasoning of the participating students had improved. The fourth supervisor doubted whether these proficiencies had improved. Half of the supervisors regarded it doubtful that the pharmacotherapeutic knowledge and skills of the students had improved after their participation, and the other two did think the students’ knowledge and skills had improved.

Fig. 4
figure 4

Left Self-reported improvement in competence of participating students, scored on a Likert scale 1–5, (strongly disagree to strongly agree), and Right CanMEDS competencies of student teams, assessed by supervisors using the MiniCEX (Likert scale 1–5, strongly disagree to strongly agree; score 3 is the level of a junior doctor). Not every CanMEDS role was applicable for every consultation

Full size image

Table 3 shows the responses and comments of student participants, arranged by the categories identified during the analysis, based on the three psychological needs as described in the SDT, i.e. autonomy, competence, and relatedness. Twenty-nine students provided feedback, these responses reflected the students valued working together in teams and enjoyed the supervision (n = 14) (relatedness), they valued their roles and opportunities in contributing to real patientcare (n = 13) (autonomy), and they felt responsible and competent in the consultations, including the patient management/proposing a treatment plan (n = 11) (autonomy and competence).

Table 3 Quotations/Motivational statements by participating students (Some quotes fit into more than one theme)
Full size table

Correlations between motivation and competence improvement

Table 4 shows the Pearson correlations between motivation (subscales of IMI and AMS) with the effect sizes and perceived improvement in CanMEDS competencies. Strong positive correlations (defined as r = 0.7 to 0.9) [39] were found between the IMI subscale Interest and the IMI subscale Usefulness (r 0.724; p < 0.01), and between improvement in the CanMEDS competencies Communicator and Collaborator (r 0.713; p < 0.01). Moderate positive correlations (r 0.4 to 0.6) [39] were found between the IMI subscale Interest and improvement in the CanMEDS competencies Medical Expert and Academic (0.522; p < 0.01) (0.656 p < 0.01), and between the IMI subscale Usefulness and all self-reported improvement in the CanMEDS competencies except Collaborator (r 0.401 to 0.672). Furthermore, multiple positive correlations (significant, moderate) were found between the perceived improvement in CanMEDS competencies (see Table 4).

Table 4 Pearson correlations of IMI subscales, AMS subscales and student-reported competence. Correlations were considered ‘strong positive’ for r = 0.7 to 0.9, and ‘moderate positive’ for r = 0.4 to 0.6
Full size table

Regression analysis

A regression analysis was performed to find out whether IMI-subscales Interest, Usefulness, and Perceived Choice affected Autonomous motivation. We found that the only IMI-subscale with a significant predictive effect was Perceived Choice (R 2 = 0.309, p = 0.018), the effects of Interest (p = 0.392) and Usefulness (p = 0.708) were not significant.

Discussion

In this study we investigated the different types of motivation and the proficiency in CanMEDS competencies of the participating students in our Learner-Centered Student-run Clinic (LC-SRC). Type of motivation was measured using the Academic Motivation Scale (AMS, both pre- and post-participation) and Intrinsic Motivation Inventory (IMI, after participation). CanMEDS competencies were evaluated by faculty using a mini-clinical examination and by the students themselves using a post-participation questionnaire.

Students were intrinsically motivated to participate in the LC-SRC project, which was based on the conceptual framework of learning by doing and SDT [16, 17]. While motivation for attending the medical school in general (which was already high) did not change during the study, the students’ intrinsic motivation for the LC-SRC was positively correlated with their perceived improvement in their CanMEDS competencies medical expert and academic. The faculty involved in the assessment considered that the student teams performed at a junior doctor level.

We did not find any studies in medical education that have used the IMI, although the instrument has been used in studies of sports and dental education [28, 36]. Intrinsic motivation in our current study was comparable to the levels measured in dental students who completed a preclinical laboratory course in operative dentistry, they reported IMI subscale scores for Usefulness in two cohorts (5.9 and 6.2) [36]. In experimental manipulations, rationale, acknowledgement, and perceived choice were identified to positively influence internalization and intrinsic motivation [27].

We think that the IMI can prove useful to educationalists and curriculum developers for evaluating intrinsic motivation for medical education (projects), especially because many current educational practices (i.e., lectures and the motivation through pressure such as exams) do not stimulate intrinsic (and autonomous) motivation [18]. This is the first study to explicitly measure (intrinsic) motivation for participation in SRC projects using a validated quantitative method [7,41,, 40–42]. We found intrinsic motivation to be high, especially on the Interest/enjoyment and Usefulness subscales, which we had anticipated given the characteristics of the project [23]. Kusurkar et al. described 12 tips to stimulate intrinsic motivation in medical education, including the stimulation of autonomy (providing optimal challenges), encouraging participation, and encouraging students to accept more responsibility for their learning (and moreover for their patients) [23]. These components formed the basis of the conceptual framework of learning by doing in the LC-SRC [7, 9]. The important role of intrinsic motivation in the LC-SRC design was further confirmed by the open feedback given by the participating students, feedback which was consistent with the basic psychological needs described in the SDT (autonomy, competence, and relatedness) [16, 17]. Students referred to the SRC as an ‘instructive/educational’ place where they could function in relative ‘autonomy’, and as ‘a special opportunity for early clinical teaching and to learn and think independently about (pharmaco)therapy’.

Motivation for medical school in general did not improve, possibly because the duration (time and intensity) of this project was too short to achieve a change in motivation, and the level of motivation was already high before participation in this group. Overall, the level of autonomous motivation (AMS subscale) observed in this study seemed to be higher (pre 5.51 and post 5.39) compared to an earlier study in our institution (male 5.309, female 5.353) [38]. In this same study, the levels of Controlled motivation (male 4.464, female 3.996) were comparable to ours (pre 3.98 and post 4.23) [38]. Although this comparison suggests superior autonomous motivation in LC-SRC participants compared to regular curriculum students, these data must be interpreted with caution.

An alternative explanation for the lack of improvement of motivation is that students found the LC-SRC so exciting and challenging that they were less motivated to attend regular education such as lectures and training with fictional cases. This explanation is supported by the students’ comments, which indicated they were disappointed in their limited role in patient care during their regular clerkships.

We found no significant difference on (change in) motivation between male and female participants on the AMS and IMI subscales. Previous studies did find different motivation profiles in males and females. In these studies males tend to have higher levels of controlled motivation, and lower levels of autonomous motivation and Relative Autonomous Motivation (RAM) [38,44,, 43–45]. A possible explanation for this difference is the majority of female participants in this study and the small sample size. Therefore this study has a limited power to detect gender differences.

The students perceived that their proficiency in CanMEDS competencies improved as a result of LC-SRC participation. Nevertheless, self-reported improvement is known to be biased and is poorly correlated with other performance measures [46]. However, the faculty in this study also reported that the students’ proficiency had improved, with specific improvement in communication with patients, medical knowledge, and clinical reasoning, and that the student teams of 1st, 3rd and 5th year students performed at the level of junior doctors. The last one is especially important because competence at the level of junior doctors is expected from the students only after they finish their medical study.

In the regression analysis of the correlations between motivation and perceived competence improvement, we found the main intrinsic motivation outcome, the IMI interest subscale, was positively correlated with the students’ perceived improvement in CanMEDS competencies of medical expert and academic. The IMI-usefulness subscale which is important in self-regulation/internalization, seemed to have stronger correlations with perceived competence improvement. The latter could be expected, if a student would think his/her competences improved he/she would be more likely to consider the project useful (and thus have a higher IMI usefulness score, and vice versa). Even though expected, an earlier study in dentistry students showed no significant correlation between the IMI usefulness score and their competence (unfortunately, the IMI interest subscale was not used) [36]. An interesting finding in the regression analysis was the weak/absent correlation between the AMS intrinsic motivation and IMI subscales, this finding suggests the IMI has additional value next to the AMS. A possible explanation for this difference is the IMI measures the intrinsic motivation for a particular topic, and the AMS measures this for medical education in general.

The study had some limitations. The main limitation was the sample size and study design (no control group). Other limitations included a social desirability- and observer bias (participants could give socially desirable answers while filling out questionnaires, and faculty were not blinded while scoring students). Furthermore there was a selection bias, such that highly enthusiastic students might have applied and consequently have been selected, as discussed previously [9]. This selection might have influenced the (intrinsic) motivation, knowledge, and skills assessed during this study. The selection-bias was unavoidable, given the LC-SRC is a voluntary extracurricular activity with real patients. Therefore, the results regarding the clinical competence and (unchanged) high levels of motivation cannot be extrapolated to non-selected medical students in general. In spite of these limitations, we think that this exploratory pilot study contributes to our knowledge of motivation and learning in a student-run clinic, and helps in designing future studies. Such future research (with a control group) is needed to determine whether prolonged participation in a SRC setting improves competence and motivation for medical education.

Conclusions

Students showed a high level of intrinsic motivation to participate in the LC-SRC and perceived an improvement in their competence. Furthermore their actual clinical competence was at junior doctor level in all CanMEDS competencies. We are of the opinion these competencies can be learned best in a setting similar to the future profession such as a LC-SRC, so that students are exposed to responsibility, real patient contact, and inter and intra disciplinary collaboration that is important to stimulate students’ intrinsic motivation. The LC-SRC offers an stimulating environment according to the theoretical framework based on the SDT. Together with the observed high levels of intrinsic motivation and the qualitative comments of the students in this study, this makes the LC-SRC an attractive place for learning.