Assessing the Application of Large Language Models in Generating Dermatologic Patient Education Materials According to Reading Level: Qualitative Study.

BACKGROUND: Dermatologic patient education materials (PEMs) are often written above the national average seventh- to eighth-grade reading level. ChatGPT-3.5, GPT-4, DermGPT, and DocsGPT are large language models (LLMs) that are responsive to user prompts. Our project assesses their use in generating dermatologic PEMs at specified reading levels. OBJECTIVE: This study aims to assess the ability of select LLMs to generate PEMs for common and rare dermatologic conditions at unspecified and specified reading levels. Further, the study aims to assess the preservation of meaning across such LLM-generated PEMs, as assessed by dermatology resident trainees. METHODS: The Flesch-Kincaid reading level (FKRL) of current American Academy of Dermatology PEMs was evaluated for 4 common (atopic dermatitis, acne vulgaris, psoriasis, and herpes zoster) and 4 rare (epidermolysis bullosa, bullous pemphigoid, lamellar ichthyosis, and lichen planus) dermatologic conditions. We prompted ChatGPT-3.5, GPT-4, DermGPT, and DocsGPT to "Create a patient education handout about [condition] at a [FKRL]" to iteratively generate 10 PEMs per condition at unspecified fifth- and seventh-grade FKRLs, evaluated with Microsoft Word readability statistics. The preservation of meaning across LLMs was assessed by 2 dermatology resident trainees. RESULTS: The current American Academy of Dermatology PEMs had an average (SD) FKRL of 9.35 (1.26) and 9.50 (2.3) for common and rare diseases, respectively. For common diseases, the FKRLs of LLM-produced PEMs ranged between 9.8 and 11.21 (unspecified prompt), between 4.22 and 7.43 (fifth-grade prompt), and between 5.98 and 7.28 (seventh-grade prompt). For rare diseases, the FKRLs of LLM-produced PEMs ranged between 9.85 and 11.45 (unspecified prompt), between 4.22 and 7.43 (fifth-grade prompt), and between 5.98 and 7.28 (seventh-grade prompt). At the fifth-grade reading level, GPT-4 was better at producing PEMs for both common and rare conditions than ChatGPT-3.5 (P=.001 and P=.01, respectively), DermGPT (P<.001 and P=.03, respectively), and DocsGPT (P<.001 and P=.02, respectively). At the seventh-grade reading level, no significant difference was found between ChatGPT-3.5, GPT-4, DocsGPT, or DermGPT in producing PEMs for common conditions (all P>.05); however, for rare conditions, ChatGPT-3.5 and DocsGPT outperformed GPT-4 (P=.003 and P<.001, respectively). The preservation of meaning analysis revealed that for common conditions, DermGPT ranked the highest for overall ease of reading, patient understandability, and accuracy (14.75/15, 98%); for rare conditions, handouts generated by GPT-4 ranked the highest (14.5/15, 97%). CONCLUSIONS: GPT-4 appeared to outperform ChatGPT-3.5, DocsGPT, and DermGPT at the fifth-grade FKRL for both common and rare conditions, although both ChatGPT-3.5 and DocsGPT performed better than GPT-4 at the seventh-grade FKRL for rare conditions. LLM-produced PEMs may reliably meet seventh-grade FKRLs for select common and rare dermatologic conditions and are easy to read, understandable for patients, and mostly accurate. LLMs may play a role in enhancing health literacy and disseminating accessible, understandable PEMs in dermatology.

Updates in atopic dermatitis for the primary care physician: A review of advances in the understanding and treatment of atopic dermatitis.

Atopic dermatitis (AD) is a common inflammatory skin condition occurring in both pediatric and adult patients. Pruritus is a clinical hallmark of the disease, and patients with AD often experience disruptions to their quality of life. The pathogenesis of AD is a complex and multifactorial interplay between genetic factors, epidermal barrier disruption, and immune dysregulation. Clinically, AD is characterized by pruritus, eczematous skin changes, and age-specific lesion distribution patterns. Infants and young children tend to have AD lesions on their face and extensor surfaces of their extremities while older children and adults tend to have AD lesions on flexural surfaces of their extremities. Many patients also experience a chronic and relapsing disease course. Due to the chronicity and severe pruritus, lesions often undergo secondary changes like lichenification. Patients with AD can experience a number of comorbidities including other atopic disease (i.e. allergic rhinitis, asthma), skin infections, cardiovascular, and neuropsychiatric illnesses. Management of AD depends on the severity of the disease as well as the distribution of the disease. Traditionally, treatment of AD included the use of moisturizers / emollients, topical corticosteroids or topical calcineurin inhibitors, or systemic therapy with non-selective immunosuppressants such as corticosteroids, cyclosporine, azathioprine, or similar. However, in the past decade, new biologic and small molecule drugs, both topical and systemic, have become important therapeutic options for AD patients, especially for those with moderate-to-severe disease. The development of these medications, following decades of research to better understand AD, are designed to specifically target various components of immune dysregulation and inflammation implicated in the pathogenesis of AD. Their successful development and deployment now allow for an exciting new era of treatment for individuals suffering from atopic dermatitis.

Effects of Telemedicine on Pediatric Clinical Care During COVID-19.

Coronavirus disease 2019 (COVID-19) has provided unprecedented opportunity for the growth of telemedicine in pediatrics. Clinics rapidly expanded their telemedicine capabilities and converted in-person visits to virtual visits to offer patients continued care while limiting infection risks. Although the transition to video visits has several benefits-patient convenience, versatility of care, ease of follow up, and management of chronic conditions-these gains are only available to those who have the requisite technology access and literacy needed to conduct a video visit. The need for technology-dependent care stands to further widen preexisting disparities in access to care for those who cannot use or afford the necessary technology. Our article explores some of these benefits and challenges and reminds us that although telemedicine has proven to have high utility for many patients, there are improvements that must still be made to ensure equitable accessibility and durability of virtual visits. [Pediatr Ann. 2021;50(12):e503-e508.].

Digitally Disconnected: Qualitative Study of Patient Perspectives on the Digital Divide and Potential Solutions.

BACKGROUND: As telemedicine utilization increased during the COVID-19 pandemic, divergent usage patterns for video and audio-only telephone visits emerged. Older, low-income, minority, and non-English speaking Medicaid patients are at highest risk of experiencing technology access and digital literacy barriers. This raises concern for disparities in health care access and widening of the "digital divide," the separation of those with technological access and knowledge and those without. While studies demonstrate correlation between racial and socioeconomic demographics and technological access and ability, individual patients' perspectives of the divide and its impacts remain unclear. OBJECTIVE: We aimed to interview patients to understand their perspectives on (1) the definition, causes, and impact of the digital divide; (2) whose responsibility it is to address this divide, and (3) potential solutions to mitigate the digital divide. METHODS: Between December 2020 and March 2021, we conducted 54 semistructured telephone interviews with adult patients and parents of pediatric patients who had virtual visits (phone, video, or both) between March and September 2020 at the University of Chicago Medical Center (UCMC) primary care clinics. A grounded theory approach was used to analyze interview data. RESULTS: Patients were keenly aware of the digital divide and described impacts beyond health care, including employment, education, community and social contexts, and personal economic stability. Patients described that individuals, government, libraries, schools, health care organizations, and even private businesses all shared the responsibility to address the divide. Proposed solutions to address the divide included conducting community technology needs assessments and improving technology access, literacy training, and resource awareness. Recognizing that some individuals will never cross the divide, patients also emphasized continued support of low-tech communication methods and health care delivery to prevent widening of the digital divide. Furthermore, patients viewed technology access and literacy as drivers of the social determinants of health (SDOH), profoundly influencing how SDOH function to worsen or improve health disparities. CONCLUSIONS: Patient perspectives provide valuable insight into the digital divide and can inform solutions to mitigate health and resulting societal inequities. Future work is needed to understand the digital needs of disconnected individuals and communities. As clinical care and delivery continue to integrate telehealth, studies are needed to explore whether having a video or audio-only phone visit results in different patient outcomes and utilization. Advocacy efforts to disseminate public and private resources can also expand device and broadband internet access, improve technology literacy, and increase funding to support both high- and low-tech forms of health care delivery for the disconnected.

Clinician Perspectives on Telemedicine: Observational Cross-sectional Study.

BACKGROUND: Since the COVID-19 pandemic onset, telemedicine has increased exponentially across numerous outpatient departments and specialties. Qualitative studies examining clinician telemedicine perspectives during the pandemic identified challenges with physical examination, workflow concerns, burnout, and reduced personal connection with patients. However, these studies only included a relatively small number of physicians or were limited to a single specialty, and few assessed perspectives on integrating trainees into workflows, an important area to address to support the clinical learning environment. As telemedicine use continues, it is necessary to understand a range of clinician perspectives. OBJECTIVE: This study aims to survey pediatric and adult medicine clinicians at the University of Chicago Medical Center to understand their telemedicine benefits and barriers, workflow impacts, and training and support needs. METHODS: In July 2020, we conducted an observational cross-sectional study of University of Chicago Medical Center faculty and advanced practice providers in the Department of Medicine (DOM) and Department of Pediatrics (DOP). RESULTS: The overall response rate was 39% (200/517; DOM: 135/325, 42%; DOP: 65/192, 34%); most respondents were physicians (DOM: 100/135, 74%; DOP: 51/65, 79%). One-third took longer to prepare for (65/200, 33%) and conduct (62/200, 32%) video visits compared to in-person visits. Male clinicians reported conducting a higher percentage of telemedicine visits by video than their female counterparts (P=.02), with no differences in the number of half-days per week providing direct outpatient care or supervising trainees. Further, clinicians who conducted a higher percentage of their telemedicine by video were less likely to feel overwhelmed (P=.02), with no difference in reported burnout. Female clinicians were "more overwhelmed" with video visits compared to males (41/130, 32% vs 12/64, 19%; P=.05). Clinicians 50 years or older were "less overwhelmed" than those younger than 50 years (30/85, 35% vs 23/113, 20%; P=.02). Those who received more video visit training modalities (eg, a document and webinar on technical issues) were less likely to feel overwhelmed by the conversion to video visits (P=.007) or burnt out (P=.009). In addition, those reporting a higher ability to technically navigate a video visit were also less likely to feel overwhelmed by video visits (P=.02) or burnt out (P=.001). The top telemedicine barriers were patient-related: lack of technology access, lack of skill, and reluctance. Training needs to be focused on integrating learners into workflows. Open-ended responses highlighted a need for increased support staff. Overall, more than half "enjoyed conducting video visits" (119/200, 60%) and wanted to continue using video visits in the future (150/200, 75%). CONCLUSIONS: Despite positive telemedicine experiences, more support to facilitate video visits for patients and clinicians is needed. Further, clinicians need additional training on trainee education and integration into workflows. Further work is needed to better understand why gender and age differences exist. In conclusion, interventions to address clinician and patient barriers, and enhance clinician training are needed to support telemedicine's durability.

Teaching Telemedicine: The Next Frontier for Medical Educators.

The COVID-19 pandemic has pushed telemedicine to the forefront of health care delivery, and for many clinicians, virtual visits are the new normal. Although telemedicine has allowed clinicians to safely care for patients from a distance during the current pandemic, its rapid adoption has outpaced clinician training and development of best practices. Additionally, telemedicine has pulled trainees into a new virtual education environment that finds them oftentimes physically separated from their preceptors. Medical educators are challenged with figuring out how to integrate learners into virtual workflows while teaching and providing patient-centered virtual care. In this viewpoint, we review principles of patient-centered care in the in-person setting, explore the concept of patient-centered virtual care, and advocate for the development and implementation of patient-centered telemedicine competencies. We also recommend strategies for teaching patient-centered virtual care, integrating trainees into virtual workflows, and developing telemedicine curricula for graduate medical education trainees by using our TELEMEDS framework as a model.