Teledermatology for preoperative consultation
According to a survey from the American College of Mohs Surgery, 67% of Mohs surgeons reported completing preoperative consultation with patients [
30]. Consulting prior to surgery establishes an opportunity for the patient and surgeon to discuss the relative risks and benefits of the procedure. This appointment fosters patient comfort and better understanding of the surgical procedure [
31]. Because this consultation is mainly informative, live video teledermatology can be implemented as a convenient and economic alternative. A retrospective analysis of Mohs surgery preoperative appointments within the Veterans Health Administration found that teledermatology consultations saved an average of 162.7 min, 144.5 miles, and $60.00 per person in average travel costs and significantly decreased consent failure rates for the surgery [
32]. Teledermatology also decreased the wait time before surgery and increased the proportion of lesions treated within a 60-day period [
32]. Telemedicine thus seems to be an effective modality for MMS preoperative consults and may improve the efficiency of subspecialty care.
Teledermatology for post-surgical follow up
Due to the low risk of surgical site infection post Mohs surgery [
33], the primary focus of a postoperative follow-up visit is to assess scar formation, wound healing, patient satisfaction, and answer any patient questions [
33]. A randomized controlled trial of 90 patients undergoing Mohs surgery reported that patients preferred to receive wound care instructions via text messages for future visits and 91% of them found the service to be “helpful” or “very helpful.” [
34] Thus, smartphone applications or postoperative text messaging may be beneficial in avoiding complications after dermatologic surgery and maximizing wound healing [
35]. Postprocedural care and positive patient experience in the post-operative period is essential to dermatologic surgery outcomes, and telemedicine shows great potential for promoting optimal recovery in this period.
Patient satisfaction and barriers
For many patients who cannot drive or easily get to the dermatology clinic, teledermatology offers a relatively simple solution to gain access to care from their preferred setting. In particular, Whited et al. reported that most patients were confident in dermatologists using images during telemedicine visits to reach a diagnosis and viewed teledermatology consultations as more convenient than in-person appointments [
42]. In two investigations in which patients submitted their own pictures either with or without dermoscopy, they noted being satisfied with the practicality and convenience [
43] and reported willingness to pay out-of-pocket amounts for teledermatology services [
44]. Other telehealth benefits appreciated by patients include more consistent skin monitoring, reduced waiting times, and enhanced privacy and comfort [
5].
In addition to its many benefits, teledermatology holds possible barriers to its implementation in dermatology practices. As specified by a database study by Fogel et al., no reported cases with final findings of medical malpractice resulting from faulty management or negligence associated with teledermatology were found [
40]. However, there may be non-reported malpractice cases, such as claims that are still being processed or were settled preceding a court decision, in addition to claims that have not yet been brought to light due to the common legal delay after a malpractice case has been established [
40]. Dermatologists should be mindful of the potential malpractice risk that comes with the use of telemedicine and may want to limit the scope of their virtual visits to less concerning dermatologic conditions or expedite in-person visits for patients with atypical lesions [
40]. Moreover, there is always a chance of acquiring low quality clinical, dermoscopic, or patient self-captured images that can make diagnosing via teledermatology unfeasible, as well as patients missing concerning lesions during skin self-examination [
40]. The application of telemedicine to Mohs Surgery can also be challenging and poses its own limitations. Incorporating telemedicine into preoperative and postoperative surgical workflow could limit schedule flexibility for both the surgeon and the patient [
45]. Another concern is the physical exam shortcomings imposed by teledermatology such that certain lesion characteristics like invasion depth may be difficult to accurately examine [
45]. Patient skill and training with new technology may also be suboptimal given the initial learning curve and that the majority of skin cancer patients are elderly [
45]. Certainly, the benefits of utilizing telemedicine must be measured against associated drawbacks and gauged with the unique framework of each practice in mind.
Pediatric teledermatology
Beyond adult populations, teledermatology has arisen as a viable solution for pediatric patients obtaining dermatologic care in an environment rife with impediments to care. These barriers are multifactorial and are influenced by aspects related to a patient’s geographic location, caregivers and patients needing to miss work or school, caregivers needing to coordinate childcare for other children, lengthy wait times, insurance, and overall deficit of practicing pediatric dermatologists. Teledermatology programs present a promising solution to these problems and are most commonly utilized for evaluating conditions like atopic dermatitis, benign nevi, infantile hemangioma, inflammatory dermatoses, molluscum contagiosum, acne, and verruca vulgaris [
46‐
48]. Studies have demonstrated a range of diagnostic concordance of pediatric telehealth with in-person evaluation between 70.1% and 89%, validating the success of these virtual visits for many dermatologic diseases [
46].
Besides its diagnostic success, the utilization of telehealth for pediatric dermatology visits has been generally well-received among pediatric patients, parents, and dermatologists. Various papers have described high satisfaction rates ranging from 77 to 98.4% for this modality of dermatologic care [
49‐
52]. Pediatric dermatologists, specifically have also expressed overall positive feedback regarding the application of telemedicine, with one study relaying that 90% of providers thought it would increase access to care, 77% predicted it could be time-efficient, and 69% believed it could be leveraged to sufficiently manage disease [
51]. Nevertheless, physicians still have apprehensions about the quality of teledermatological care due to beliefs that more mistakes could be made, various technological issues during encounters, inability to perform in-office procedures, lack of examination by palpation, and challenges associated with assessing an actively moving child through video and keeping him or her engaged [
46,
50]. Notably, the experiences of nurse practictioners and physician assistants with telehealth in pediatric patients has yet to be investigated and would be beneficial to explore in future studies.
Future directions
With the rising popularity of teledermatology, many emerging technologies are being put into practice to enhance its experience. For instance, reflectance confocal microscopy (RCM), is a noninvasive examination technique that enables high-resolution in vivo assessment of skin lesions and shows promise for telemedicine settings [
4]. Use of RCM in research and clinical environments has yielded high reliability and sensitivity in diagnosing both nonpigmented and pigmented lesions, but specificity greater depends on an individual’s’ training and skill level [
53]. With the start of integration of RCM into clinical practice, educational programs for this technique are being further advanced to avoid misdiagnosis in both in-person and virtual settings. The application of RCM to telehealth learning platforms may complement the increasing knowledge base of clinical experts in addition to helping connect patients to trained providers [
53].
Another technology with great potential for improving teledermatology efficiency is artificial intelligence (AI). Based on a meta-analysis of 70 studies, the accuracy of computer-aided diagnosis of melanoma was discovered to be comparable to that of clinical experts [
54]. The dependability of AI may thus make it a complementary service in virtual consults by tracking lesions over time and expanding the differential diagnosis list [
55]. Furthermore, the use of artificial intelligence as a telemedicine triage tool for patients with potentially malignant skin lesions can prioritize those who need expedited care [
56]. In a study evaluating the application of AI in telemedicine triage and diagnosis of cutaneous lesions, Majidian et al. found no significant difference in diagnostic accuracy between a group of three dermatologists and AI using the software’s first three differential diagnoses [
56].
However, the integration of two elaborate fields like dermatology and computational intelligence begets foundational challenges. For image-based technologies, data quality, quanitity, and diversity are conceivably the most influential factors of the model’s performance [
57]. Efforts are underway to construct substantial open-source, representative, and continuously amended datasets that will be accessible by AI developers [
58]. Another obstacle for AI development and implementation is the “black box” characteristic of modern machine learning such that the algorithm cannot explain its decicions-making rationale [
59]. Explainable artifical intelligence (XAI) is an expanding field that has been proposed to overcome this limitation and commonly utilizes a “post hoc” approach for interpretation after the result is acquired, with more recent models adding attention visualization to the process [
58]. Multimodal techniques incorporating inherently interpretable models, fine-grained structural heatmaps, and prototypical explanations are also being advanced for the realm of skin cancer recognition [
60].
From a legal perspective, the lack of standardized explainability by these AI algorithms presents ongoing issues regarding potential unforeseen and perplexing failure forms. Medical AI may also be trained using subpar techniques, with incomplete data, or under improper conditions possibly leading to patient injury. Despite no case law on liability encompassing medical AI, physicians must take care to avoid medical malpractice liability by providing competent specialty-standard care and considering all accessible resources. Thus, under current legal codes, standard of care is intrumental to liability for medical AI [
61]. Accompanying the Food and Drug Administration’s consideration of clinical decisions offered by machine intelligence as AI guided, a physician is held liable only when he or she does not comply with the legal standard of care and a patient injury follows [
61,
62]. Accordingly, physicians should constantly gauge how to interpret AI recommendations and encourage administrative and legal efforts to develop guidelines for employing AI in specific clinical domains of need.
Overall, AI can still serve as a beneficial tool in recognizing lesions that require further workup and possible biopsy, especially for primary care providers or less experienced specialist providers [
56]. Although AI is beneficial in triaging disease into broader categories, the expertise and contextual knowledge of dermatologists will always be needed for meticulous diagnosis, management recommendations, and atypical cases. Considering the rapid progression of technological developments, familiarization with the fundamental operations of AI along with its potential applications and current disadvantages will be crucial for the future of dermatology.