Effective communication between doctors and their patients is recognised as a critical goal because of the wide-ranging benefits that have been shown to accrue to both participants from such interactions, including improvements in doctor and patient satisfaction and various measures of patient outcomes [1]. Programs of communication skills training (CST) and the assessment thereof are widely established as important components of medical curricula, particularly since effective communication is not necessarily an inherent skill or determined by personality, but can be both taught and learned [2]. Didactic methods of CST are now generally augmented by experiential learning approaches to CST, with medical programs tending to employ real patients, simulated patients (SPs), and student roleplay (RP) in various combinations for this purpose; the mix of techniques is determined by factors including the educational stage and needs of the student and resource availability [3]. A considerable amount of research has attempted to determine the optimal approach to CST, although objective comparisons of SP and RP methods are somewhat lacking [1]. The available evidence suggests that these methods are equivalent for teaching certain communication skills [4-6], although RP may be superior to SP in fostering empathy for patient perspectives because RP inherently requires students to assume the patient perspective [4]. In addition, RP is considered more cost-effective as a consequence of the student’s role as both trainer and learner [7]. While students do value the RP method and it can be educationally useful [8], it may not be preferred by participants and may limit learning opportunities [8]. Alternatively, a well person playing the role of a patient for the purpose of students learning clinical skills in controlled circumstances (namely, an SP) offers a potentially more acceptable approach to CST by providing students with objective feedback on their performance [1]. SP programs are favoured by both learners and faculty because they expose students to unfamiliar people and authentic, unbiased interactions with appropriately intense emotional elements in a safe and supported learning environment [9]. It should be noted that SP programs require significant administrative and financial investment, with trained professional medical actors receiving more than AUD 80 per hour (EUR 55, USD 60) [10]. As a result, SP programs utilising volunteers have been proposed as a more sustainable means to achieve the same teaching outcomes, and have been shown to be equally effective educationally to RP [11].

The assessment of communication skills involves real patients as well as SPs, and is generally delivered through objective structured clinical examinations (OSCEs) [12]. OSCEs have been established as the clinical assessment of choice in medicine, nursing, exercise physiology, physiotherapy, and other allied health programs and commonly examine multiple skillsets simultaneously. As such, OSCEs can generate scores for generic communication skills (how the student forms and maintains a relationship with the patient and understands their perspective) and clinical communication skills (medical history-taking).

Given the available evidence regarding the optimum CST approach, our institution (the University of New South Wales) implemented an inexpensive volunteer SP program in 2015 that replaced our pre-existing student RP program and aimed to deliver improved pre-clinical CST. Two years after the SP program was first implemented, the present study set out to determine whether this volunteer SP method translated into improved OSCE performance in communication skills compared to our previous RP strategy. An OSCE at the end of year 2 forms a barrier to progression into the clinical phase of our program. Due to the relative reliability and validity of OSCEs as an assessment method, these circumstances provided the opportunity to compare 2 forms of CST as objectively as possible. To our knowledge, this is the first study of its kind in an Australian undergraduate medical program.

The results showed that the 2016 cohort was not statistically significantly different from the 2014 cohort in terms of the results of selection tools, with the exception of UMAT scores, which were lower for the 2016 cohort than for the 2014 cohort (62.61 versus 64.39). However, OSCE scores grouped by domain and overall score were significantly higher for the 2016 cohort compared to the 2014 cohort (Table 3). The effect sizes of the differences were moderate (0.37-0.44). Further, univariate analysis of variance was performed to identify EMMs by cohort, while controlling for the impact of scores on the admission tools. The results show that the impact (Cohen’s d) was slightly decreased but remained within the low-moderate effect size range (0.26-0.34), with statistical significance (P<0.01) (Table 4). The raw data are available in Supplement 1.

Our volunteer SP program for CST had a significant positive impact on students’ generic and clinical communication performance on OSCEs compared to the previous and commonly employed RP approach. Interestingly, this difference between RP and SP cohorts was maintained in the physical examination/procedural skills domain. A simple, transparent statistical analysis indicated that despite the 2016 cohort scoring lower on the UMAT (P=0.014) and ATAR (not significant), they performed better on the OSCE, a finding that further strengthens our assertion regarding the impact of SP. Our findings can be attributed with reasonable confidence to our SP program intervention alone, thus advancing our understanding of optimal methods for CST in preclinical medical students, an important area of medical education research [1]. These results appear to refute other work that has thus far suggested either equivalence of RP and SP [13] or superiority of the former strategy [4]. Importantly, the transition from RP to volunteer SP CST has been a cost-neutral process at our institution, in contrast to previous comparative analyses of professional SP and RP CST that have demonstrated inferior cost-effectiveness of the SP method [7]. While the authors acknowledge the established benefits of increased empathy and understanding of the patient’s perspective through RP [4], we argue that there are many clear advantages to using volunteer SPs that are not captured by structured observer rating forms. For example, the mean age of our SP pool (57.7 years old) is such that many are very experienced real patients who can draw on countless interactions with healthcare professionals to inform both their SP persona and their impressions of the student’s performance. This in turn imbues their feedback with an authentic patient perspective, while RP can only provide an ‘appreciation’ of that perspective (i.e., empathy) as a construct entirely dependent on the student’s own perceptions and biases. Additionally, students have reported that SP feedback is more instructive and provides specific indicators that they can act on to improve their communication skills, while the interaction itself tends to be more engaging due to heightened emotional intensity. Learners also appreciate the opportunity to practice their skills with SPs before encountering real patients for the first time [14]. Finally, we have noticed that SP feedback is more oriented towards non-technical skills than peer feedback, a focus that we consider to be the most effective use of contact teaching time for CST. In contrast, students are generally familiar with the role of the doctor and medical terminology, and therefore are not as challenged in the RP setting when engaging with peers to use patient-appropriate language and behaviour.

The unexpected result of improved OSCE scores in the domain of physical examination/procedural skills for the SP cohort may be explained in terms of the benefits that effective communication skills can bring to overall clinical performance. In particular, enhanced consultation effectiveness and efficiency, higher satisfaction, and an improved therapeutic relationship may be experienced by both patient and doctor in this setting [1]. Whilst performing physical examinations and conducting procedures on patients may at first glance appear to be related purely to technical proficiency, these tasks do require patient compliance and cooperation. Hence, good generic communication skills (including non-verbal communication skills) can be expected to facilitate improved performance beyond this domain, through establishing a strong rapport that enables physical examinations to be conducted sensitively and fluently.

A limitation of this study may lie in the equivalency of the 2 cohort groups according to demographic information. However, it should be noted that there were no significant differences according to characteristics such as age (P=0.204) and ATAR scores (P=0.59) between the 2 cohorts. Whilst the UMAT score (P=0.014) showed a significant difference between the 2 groups, as mentioned above, the 2016 cohort, which scored lower on the UMAT (P=0.014), performed better on the OSCE.

In conclusion, we have presented evidence to suggest that CST using volunteer SPs produces superior results in generic communication, clinical communication, and physical examination/procedural skills when compared to RP CST. Our volunteer SP program created no additional financial burden compared to the previous RP approach, an important consideration for any institution for whom paid professional actors would outstrip the resources available for CST. These findings add a counterpoint to some studies in the current literature and suggest that a volunteer SP program can contribute significantly to CST programs among junior medical students.