Evaluating dermatology education in Australia: Medical students’ and junior doctors’ perceptions

Despite a high frequency of dermatological general practice and emergency department presentations, including serious and potentially fatal skin conditions, evidence suggests that undergraduate dermatological education remains sparse. Given this, there exists a need to gauge dermatology‐relevant knowledge and skill in our new doctors.


BACKGROUND
Dermatological conditions comprise 5%-8% of emergency department (ED) presentations and 15% of general practice (GP) consultations in Australia. 1,2 There are over a million Medicare services claimed each year for skin cancer, with two out of three Australians diagnosed with skin cancer by age 70. 3 Additionally, Australia has the highest age-standardised incidence and mortality rate of melanoma in the world, rates that have been increasing over the last 50 years. 4,5 Despite such alarming rates of serious and potentially fatal skin conditions, evidence suggests that dermatological education in medical schools remains sparse. A 2016 paper noted that of 17 responding medical schools in Australia, only 4 have mandatory clinical rotations in dermatology. 5 The authors report that of these attachments, the median placement time is 1.5 days. 5 Many reasons have been cited as possible explanations for this lack of clinical exposure, including clinical sites without dermatology departments, limited time to schedule dermatology with the ever-expanding medical knowledge base and a shortened timeframe with the inception of 4 years graduate medical programmes in Australia. 5,6 The British Association of Dermatology (BAD) published minimum competencies deemed to be essential for graduating junior doctors. 7 No mandated dermatological curriculum exists in Australian medical schools, and the downstream effects are concerning. Lowell et al. state the rate of correct diagnosis for dermatological presentations by non-dermatologists is less than half. 8 One study demonstrated that cellulitis is often overdiagnosed due to less familiarity with mimickers such as lipodermatosclerosis and dermatitis, resulting in inappropriate hospital admissions and management. 1,9 Given that skin conditions affect upwards of 30% of the population, and the potentially fatal consequences of misdiagnosis, there exists a need to gauge knowledge and skill in our new doctors. 10

METHODS
From March 2022 to July 2022, a prospective crosssectional survey of Australian medical students and junior doctors was performed. The survey aimed to evaluate the dermatology education provided at Australian medical schools and to gauge self-perceived confidence and objective knowledge in a variety of dermatology domains.

Population
Those eligible for inclusion were as follows: •

Analysis
Quantitative and qualitative data analysis was carried out using Prism (GraphPad Software, San Diego, California, USA) and SPSS (IBM Corporation., Armonk, New York, USA). Simple descriptive statistics, mean comparison and simple linear regression were performed to analyse data.

Cohort characteristics
A total of 1111 responses were completed, consisting of medical faculty from pre-final-phase students to PGY3+ doctors. Of the 1111 completed responses, 942 were finalphase medical students or PGY1 doctors (Table 1).

Dermatology education and exposure
On a Likert scale of 1-5, participants largely disagreed that their medical studies had prepared them for internship and beyond (μ = 1.92, 95% CI 1.869-1.973), with only 54 (5%) agreeing they felt prepared. They rated highly their belief that there should be more dermatology teaching to prepare them for internship and beyond (μ = 4.23, 95% CI 4.189-4.274). Only 17 (1%) disagreed with the statement that there should be more dermatology education in their medical programs (disagreed or strongly disagreed). Responses ranged from 2-120 h when asked about their perception of the necessary amount (in hours) of dermatology education required to prepare them for internship. Given a right-skewed data set and the use of Tukey's boxplot, 37 outliers were excluded, and the median value was used. Participants believe that approximately 45.9 h of dermatology education is required to prepare them for internship.
Characteristics such as the nature of dermatology placement and whether students had undergone formal assessment were evaluated. Of the 942 respondents, 720 (76.4%) received no dermatology placement, whilst 137 (14.5%) and 85 (9.0%) received mandated and elective clinical placement, respectively. 274 (29.1%) were formally assessed in dermatology, whilst 668 participants (70.9%) were not. Formal assessment was defined as sitting a devoted dermatology assessment in the form of an objectivestructured clinical examination, clinical examination, oral examination or viva voce-style assessment.
The method of teaching (received vs. preferred) and the timing of teaching were also reported ( responses for received medium stated they received no dermatology teaching throughout their medical degree.

Self-perceived confidence and competencies
Participants rated their confidence across a range of dermatology skills and knowledge, and their ability to recognise and manage an array of dermatology conditions on a Likert scale, ranging from 1-5 (Figures 1 and 2). Interns and medical students displayed similar trends in self-reported confidence to most dermatology-relevant skills and in diagnosing and managing dermatological conditions. Of note, there was low confidence in identifying dermatological conditions in people with skin of colour (μ = 1.69, 95% CI 1.64-1.74), handing over to a dermatology colleague (μ = 2.00, 95% CI 1.94-2.05) and in diagnosing and managing acute/emergency dermatology conditions (μ = 1.84, 95% CI 1.78-1.89). In contrast, participants felt most confident in counselling sun-protective behaviours (μ = 3.95, 95% CI 3.89-4.00) and suturing small superficial wounds (μ = 3.43, 95% CI 3.35-3.51).
The aggregated mean Likert score for each participant was calculated and analysed against a variety of factors. Those who had received any dermatology placement (mandated or elective) had a significantly higher aggregated self-reported confidence than those who had not (95% CI 0.148-0.317, p < 0.0001). Those who had been formally assessed had a significantly higher self-reported confidence than those who had not (95% CI 0.182-0.335, p < 0.0001). There was no difference in self-reported confidence between undergraduate and postgraduate participants (p = 0.207).

Objective assessment: MCQs
As seen in Figure 3, 50% correctly differentiated drug eruptions, 73% correctly identified cellulitis mimics, 63% correctly differentiated skin cancers, 72% correctly managed fungal infestations, and 29% correctly identified disorders with mucosal involvement. Figure 4 shows the proportion of participants that answered questions correctly.
The aggregate mean Likert score for self-reported confidence was analysed against the number of MCQs that F I G U R E 1 Self-reported confidence in dermatology skills and knowledge.
F I G U R E 2 Self-reported confidence in diagnosing and managing dermatology conditions. F I G U R E 3 Multiple-choice questions answered correctly. participants answered correctly. Linear regression calculated a significant (p < 0.001), very weakly positive relationship (Pearson's R 2 = 0.06) between mean Likert score and number of MCQs answered correctly Participants who reported receiving any dermatology placement (mandated or elective) answered significantly more MCQs correctly than those who had not (95% CI 0.007-0.384, p < 0.0419). Participants who had been formally assessed throughout their medical degree answered significantly more MCQs correctly than those who had not (95% CI 0.067-0.418, p = 0.007). Interns answered MCQs correctly significantly more than medical students (95% CI 0.145-0.509, p = 0.0004). Respondents who were studying, or who had studied, a postgraduate medical course answered significantly more MCQs correctly than their undergraduate counterparts (95% CI 0.026-0.346, p = 0.0227).

General comments
339 respondents left free text comments about their experience of the dermatology education provided. 69 responses reported the 'majority, if not entirety, of their clinical dermatology exposure occurred on GP placements', with heterogenous amounts of exposure and teaching. 5 responses commented that given faculties do not frequently assess dermatology, it constitutes less study priority. 218 responses supported more dermatology teaching and/or more clinical dermatology exposure in their degrees.

DISCUSSION
This study constitutes the largest evaluation of medical students' and junior doctors' self-perceived confidence in dermatology skills and pathologies with 942 eligible participants. Overall, the results demonstrate that Australia's junior workforce exhibit low confidence across a broad range of dermatology outcomes.
Junior doctor confidence in dermatological diagnosis is essential, with dermatological conditions comprising 5%-8% of ED presentations and up to 15% of GP consultations in Australia. 1,2 Despite this, only 5% felt adequately prepared by the current provision of dermatology education, and only 1% disagreed there should be more dermatology education in their medical degrees. Many medical students and interns will become GPs and ED specialists where they will encounter skin conditions daily. Whiting et al. demonstrated that GP registrars exhibit low confidence in diagnosing and managing skin conditions, with consultations containing skin conditions significantly associated with the trainee requesting in-consultation advice or assistance (Odds Ratio = 1.91). 11 Many studies have similarly sought to clarify and explore the current situation in dermatology, with results consistent with this survey's findings. Adel et al. surveyed 42 junior doctors in the United Kingdom, where confidence was assessed on a Likert scale of 1-10. 12 Junior doctors reported the least confidence in recognising and managing dermatological emergencies (μ = 3.6) and skin conditions associated with systemic diseases (μ = 4.3), whilst scoring highest in diagnosing common inflammatory conditions (μ = 5.7) and taking dermatological histories (μ = 5.7). Most respondents (91%) believed there should be monthly dermatological teaching.
Regarding clinical exposure, 76.43% reported having no dermatology placement throughout their degree. When asked about the number of hours required to provide sufficient exposure, participants reported approximately 45.9 hours (~6 days). As previously described, Australian medical students that have a clinical dermatology attachment attend approximately 1.5 days throughout their entire program. 5 This study demonstrates the importance of clinical dermatology placement, with students who had undertaken any dermatology placement significantly more confident (p < 0.0001) and able to answer more MCQs correctly (p = 0.0419).
Australian medical students' lack of confidence in dermatology is in direct contrast to other medical fields. In comparison, a survey evaluating self-perceived confidence in obstetrics and gynaecology (O&G) showed 75.8% of students reported feeling confident in recognising and managing common O&G conditions. 13 These findings suggest that including a short mandatory rotation in dermatology would increase confidence in diagnosing and treating dermatologic pathologies.
A unique feature of our study was the inclusion of knowledge MCQs, which provided two benefits: (1) permitted an objective evaluation of knowledge in the domains assessed and (2) permitted correlation analysis between various factors and objective knowledge for the domains assessed. Many studies describe that one's selfperception may be misaligned to their actual level of knowledge or competency. 14,15 Linear regression calculated a significant (p < 0.001), weakly positive relationship between the mean self-reported confidence and number of MCQs answered correctly. Whilst this infers a significant positive relationship, this suggests an increasing frequency of correctly answered MCQs in this study is explained by factor(s) other than a higher self-reported confidence.
One inherent limitation is the survey's voluntary nature, which results in self-selection. Self-selection and response bias may confound the external validity where findings would ideally be extrapolated to the entire eligible population. Additionally, the link between selfreported confidence and objective competence is unclear, with many studies citing that often individuals view themselves more favourably than objectively true. 14,15 Our study attempted to address this by including an objective assessment. More research is required to evaluate the validity of self-perceived confidence compared to objective acumen.
Whilst all interns and final-phase medical students were eligible, the accessibility of the survey was dependent on the willingness of institutions. Some universities rejected the premise of the survey, with one stating that 'a survey of this nature would not be approved'. This alone is interesting as it provides a reason that some medical schools may be aware of the shortcomings of the education they are providing. It may also have resulted in bias with distribution only in those medical schools willing to have their students' confidence assessed. Medical societies were globally willing to distribute the survey. Distribution to interns was contingent on relevant administrative hospital staff and JMO associations facilitating dissemination.
To allow participants to accurately answer questions in a standardised manner, it seemed appropriate to limit clinical exposure to that provided by dermatologists or trainees. In reality, many students will experience varying degrees of dermatology whilst on GP placements, surgical rotations, or as manifestations of diseases in internal medicine. 5 Given this, our survey may underestimate the amount of clinical dermatology students experienced. Whilst these findings demonstrate low self-perceived confidence on a national scale in dermatology, there are many potential solutions. Firstly, the installation of a national curriculum in dermatology would serve to standardise education standards and better prepare graduating doctors. In Australia, the Australasian College of Dermatologists developed undergraduate modules partnered with the University of Sydney, which were introduced to final-year medical students at the University of Western Australia. Singh et al. evaluated these modules' effectiveness, demonstrating that students who completed them felt they had a more structured curriculum, better clinical experience and had acquired skills and knowledge. 16 Widespread installation of these modules, which take approximately 20-25 h to complete, would better prepare students in dermatological competencies and standardise the curriculum. They are currently only available at 6 universities. 5 Another recommendation is the inception of dermatology-focused assessments in medical curricula. Currently, only 6 Australian universities 'required students to satisfy dermatology standards in university examination regulations'. 5 Given that students more readily engage in content that is examinable and that our study and other literature demonstrates improved learning outcomes and satisfaction when assessments are involved, the authors support at least one clinically focussed dermatology assessment across medical programs.
Finally, Australian medical students are supportive of their schools providing more dermatology exposure. Many participants believe there would be ample time for a 1-4 week attachment in dermatology throughout their course. This can come in the form of dermatology outpatient units in public hospitals or rotating through private dermatology clinics, similar to students' GP placements.

CONCLUSION
Final-phase medical students and junior doctors exhibit low confidence across an array of dermatology competencies. Participants believed they need more dermatology education, and they seldom feel prepared when exposed to dermatologic pathologies. Avenues to improve student knowledge and confidence include introducing dermatology focussed assessments, increasing accessibility to dermatology educational resources and allocating more time to clinical attachments.

FUNDING INFORMATION
This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

CONFLICTS OF INTEREST STATEMENT
There are no conflicts of interest to disclose.

ETHICAL APPROVAL
Ethics approval was obtained from the Hunter New England Local Health District Human Research Ethics Committee. Governance approval was acquired for hospitals where the JMO Manager or DPET agreed to distribute the survey to interns.