Abstract
Summary
We evaluated a teledermatology consultation service in the education of medical trainees. The selection of cases for consultation was at the discretion of the trainees, who could contact the study team for advice about obtaining photographs and submitting the case to a web-based system. Asynchronous structured feedback was provided to trainees by an academic paediatric dermatology consultant using a web-based interface. Efficacy was evaluated by examining the trainees' self-reported competency in clinical dermatology skills before and after teleconsultation. A total of 44 trainees (31 residents and 13 medical students) completed 50 consultations. Trainees reported significant improvement (mean improvement 22%, P < 0.002) in competency in five of the six areas assessed. In addition, 88% of trainees were very satisfied with the teaching methodology (≥5 on a 7-point scale) and 86% were very likely to apply the information in their future practice (≥5 on a 7-point scale). We believe that teledermatology has great potential in the education of medical trainees.
Introduction
Providing feedback to trainees about the patients that they care for is an effective method of teaching. 1 However, in both outpatient and inpatient settings, it is difficult for generalist preceptors to provide teaching about specialty-related problems. For example, as part of an outpatient paediatric clerkship, trainees often encounter patients with dermatology conditions. However, in many of these cases the diagnosis and management is uncertain, resulting in a lost potential teaching opportunity. Telemedicine could be used to provide prompt formative educational feedback to trainees by experts who may be based elsewhere. Providing such feedback is in accordance with the principles of adult learning and would probably provide an effective learning experience for trainees. 1
The use of telemedicine in dermatology has been the subject of many studies. These studies have reported reassuring concordance between face-to-face consultation and teledermatology (range 72–84%). 2–8 Dermatology is one of the fields where telemedicine has been shown to be effective and teledermatology has been shown to reduce the need for face-to-face consultation. 9–12 Both patients and primary care physicians have reported high satisfaction with teledermatology. 13–15
While many studies have examined the effectiveness of telemedicine in patient care, very few have focused on telemedicine's role in providing specialty-related education during patient encounters. We have therefore studied the use of paediatric teledermatology for the education of medical trainees.
Methods
Any trainee (i.e. paediatric resident or third-year medical student) rotating through the general paediatric outpatient clinic of the Children's Hospital of Pittsburgh from October 2003 to April 2005 was eligible to participate in the study. Trainees were advised that a teleconsultation was voluntary, that it was to be used solely for their own education and not for patient management, and that their participation would not affect their personal evaluations. The study was approved by the appropriate ethics committee.
The selection of cases for consultation was at the discretion of trainees. Trainees could contact the study team for advice about obtaining photographs and submitting the case online. Most cases were those in which the diagnosis or management plan remained uncertain after discussion with the paediatric clinic preceptor. We obtained consent for photography from the family before capturing digital photographs of the lesion using a digital camera. The image resolution was 3072 x 2048 pixels.
Consultations
Trainees completed an online structured preconsultation questionnaire which included data from the history and physical examination as well as the trainee's diagnosis, differential diagnosis and management plan (Table 1). As part of the physical examination, trainees were asked to generate a detailed description of the lesions using dermatological terminology provided to them in dropdown menus. This structured approach was used in order to introduce and reinforce trainees to the key aspects of a dermatological examination. Specifically, trainees were asked to describe the overall patterns of the lesions (symmetric, asymmetric, acral, centrifugal, clothing-covered, sun-exposed, flexor, extensor), the regional pattern (scattered, clustered, coalesced, dermatomal, follicular), the distribution, and the primary and secondary lesion types. In addition, for each case, trainees were asked to rate their competency in six clinical skills before consultation. This information was forwarded to the dermatologist using a password protected web-based system. The terminology and meta-data structure used in the web system were based on previous work on an online dermatology image atlas. 16
Outline of the preconsultation and consultation forms (dropdown choices not shown)
When the trainees initiated a consultation they committed to a diagnosis, provided supporting evidence for their diagnosis, provided alternative hypotheses (differential diagnoses) and posed specific questions to the consultant. This accords with the principles of adult learning 1 and the ‘1 minute preceptor’. 17 The consultant then used the information provided by the trainee to formulate individualized feedback. Feedback included comments regarding the learner's ability to collect the appropriate data (history and physical examination). The dermatologist also assessed the trainee's differential diagnoses and provided trainees with steps used to arrive at the correct diagnosis (Table 1). The consultant also provided trainees with a general rule or a short phrase crystallizing a dermatology principle, e.g. ‘children with giant congenital nevi (>20 cm) are at increased risk of malignant transformation’. Once the consultation was completed (usually within 48 hours), the trainee received an email message regarding the availability of the feedback on the website. The trainees were able to review only their own cases.
Efficacy of consultations
Efficacy was measured by examining the trainees' change in perceived competency before and after the consultation. Trainees rated themselves in six competency areas: diagnosis, management, history-taking, physical examination, knowledge and differential diagnosis using a 7-point scale (e.g. from 1 = not knowledgeable to 7 = very knowledgeable) before and after consultation. Paired 2-sided t-tests were used to compare the mean change in perceived competency. The estimated sample size for the study was based on an assumed small to medium effect size (Cohen's d = 0.4), a two-tailed alpha of 0.05 and a power of 0.8. These assumptions resulted in a sample size of 52.
Results
Forty-four trainees (31 residents and 13 medical students) participated in the study (out of a total pool of approximately 150 trainees). Ten trainees submitted more than one case (range 1–3). Fifty-eight cases were submitted for consultation during the 19-month study period. Some of the most common conditions (in decreasing order of frequency) were contact dermatitis, atopic dermatitis, viral exanthema, insect bites, id reaction, pityriasis rosea, epidermal nevus, molluscum contagiosum, tinea capitis and seborrheic dermatitis.
Consultations
The mean time spent completing each consultation, as reported by the consultant, was 15.2 min (SD 1.5). The consultant was very sure or certain of the diagnosis (≥5 on the 7-point scale) 88% of the time. The consultant reported that image quality was very good (≥5 on a 7-point scale) on 88% of occasions. The trainee's diagnosis was in agreement with that of the dermatologist in 21 of the 58 submitted cases (36%).
Efficacy of consultations
The post-consultation form was not completed by trainees in eight cases (six residents and two medical students). These were excluded from the analysis of efficacy. Trainees reported significant improvement (mean improvement 22%, P < 0.002) in competency in five of the six areas assessed (Table 2). Specifically, scores improved significantly from pre- to post-consultation for the following five competencies: diagnosis (26% improvement, P < 0.001), management (16% improvement, P < 0.001), physical examination (improved 21%, P = 0.002), knowledge (improved 41%, P < 0.001) and differential diagnosis (21% improvement, P < 0.001). Trainees also reported improvement (6% improvement) in dermatology history-taking skills, but this was not significant (P = 0.08). The sample size was not large enough to allow meaningful comparison of efficacy by training level.
Self-reported change among 50 trainees in six educational competencies on a 7-point scale before and after teledermatology consultation
In addition, 88% of trainees were very satisfied with the teaching methodology (≥5 on a 7-point scale) and 86% were very likely to apply the information in their future practice (≥5 on a 7-point scale).
Discussion
To our knowledge, this is the first report of telemedicine used as a teaching tool for specialty-related content. Our experience demonstrates that a teleconsultation can be incorporated into a busy clinical setting, given some initial funding and collaboration of interested parties. Our consultation forms were designed using the principles of adult learning. The preconsultation form allowed us to teach trainees the elements of a complete dermatology history and physical examination. The consultation form provided a structure to assist the consultant to become more effective in teaching about the case. The case-based individualized feedback using a web-based interface was highly valued by trainees. Our preliminary evaluation indicates that trainees improved (mean improvement, 22%) in most of the competencies assessed.
The present study was a before/after, non-randomized evaluation. Therefore the strength of the conclusions is limited. The use of self-reporting, for example, may have introduced the Hawthorne effect, i.e. some of the reported changes in competency may have been due to the trainees' awareness of being observed as research subjects. As each learner was expected to submit only one or two consultations during the course of their primary care rotation, a standardized test in paediatric dermatology was not considered to be an appropriate evaluation method. The use of a randomized design (with a control group receiving only usual teaching) would have strengthened the conclusions. However, in the context of examining a novel teaching method, we decided to begin with the development of a feasible methodology for teleconsultation. Finally, we did not have adequate power to explore the role of training level on the study results.
Notwithstanding these limitations, teledermatology provided our trainees with an opportunity to initiate a dialogue about their own patients with an expert clinician. The web-based, structured interface prompted trainees to describe lesions in an organized and detailed manner and allowed the dermatology consultant to provide detailed feedback. We believe that teledermatology has great potential in the education of medical trainees.
Footnotes
Acknowledgements
We thank Dr Elaine Rubenstein (University of Pittsburgh, Office of Measurement and Evaluation of Teaching) for her contribution to data analysis and sample size estimation. The University of Pittsburgh provided financial support through an Innovation in Education grant.