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.

BACKGROUND: A rapid expansion of systemic immunological treatment options for atopic dermatitis (AD) has created a need for clinically relevant and understandable comparative efficacy and safety information for patients and clinicians. Given the scarcity of head-to-head trials, network meta-analysis (NMA) is an alternative way to enable robust comparisons among treatment options; however, NMA results are often complex and difficult to directly implement in shared decision-making.
OBJECTIVE: The aim of this study is to develop a website that effectively presents the results of a living systematic review and NMA on AD treatments to patient and clinician users.
METHODS: We conducted a multimethod study using iterative feedback from adults with AD, adult caregivers of children with AD, dermatologists, and allergists within a user-centered design framework. We used questionnaires followed by workshops among patients and clinicians to develop and improve the website interface. Usability testing was done with a caregiver of a patient with eczema.
RESULTS: Questionnaires were completed by 31 adults with AD or caregivers and 94 clinicians. Patients and caregivers felt it was very important to know about new treatments (20/31, 65%). Clinicians felt the lack of evidence-based comparisons between treatments was a barrier to care (55/93, 59%). \"Avoiding dangerous side effects\" was ranked as the most important priority for patients (weighted ranking 5.2/7, with higher ranking being more important), and \"improving patients\' overall symptoms\" was the most important priority for clinicians (weighted ranking 5.0/6). A total of 4 patients and 7 clinicians participated in workshops; they appreciated visualizations of the NMA results and found the website valuable for comparing different treatments. The patients suggested changes to simplify the interface and clarify terminology related to comparative efficacy. The user in the usability testing found the website intuitive to navigate.
CONCLUSIONS: We developed a website, \"eczematherapies.com,\" with a user-centered design approach. Visualizations of NMA results enable users to compare treatments as part of their shared decision-making process.

BACKGROUND: Self-help interventions have the potential to increase access to evidence-based mental health care. Self-help can be delivered via different formats, including print media or digital mental health interventions (DMHIs). However, we do not know which delivery format is more likely to result in higher engagement.
OBJECTIVE: The aims of this study were to identify if there is a preference for engaging in print media versus DMHIs and whether there are individual differences in relative preferences.
METHODS: Participants were 423 adults between the ages of 18 and 82 years (201/423, 47.5% female) recruited on Prolific as a nationally representative sample of the US population, including non-Hispanic White (293/423, 69.2%), non-Hispanic Black (52/423, 12%), Asian (31/423, 7%), Hispanic (25/423, 6%), and other individuals (22/423, 5%). We provided individuals with psychoeducation in different self-help formats and measured their willingness to use print media versus DMHIs. We also assessed participants\' demographics, personality, and perception of each format\'s availability and helpfulness and used these to predict individual differences in the relative preferences.
RESULTS: Participants reported being more willing to engage with print media than with DMHIs (B=0.41, SE 0.08; t422=4.91; P<.001; d=0.24, 95% CI 0.05-0.43). This preference appeared to be influenced by education level (B=0.22, SE 0.09; t413=2.41; P=.02; d=0.13, 95% CI -0.06 to 0.32), perceived helpfulness (B=0.78, SE 0.06; t411=13.66; P<.001; d=0.46, 95% CI 0.27-0.66), and perceived availability (B=0.20, SE 0.58; t411=3.25; P=.001; d=0.12, 95% CI 0.07-0.30) of the self-help format.
CONCLUSIONS: This study suggests an overall preference for print media over DMHIs. Future work should investigate whether receiving mental health treatment via participants\' preferred delivery format can lead to higher engagement.

BACKGROUND: Telerehabilitation for musculoskeletal (MSK) conditions may produce similar or better outcomes than usual care, but most telerehabilitation studies address only chronic or postsurgical pain.
OBJECTIVE: We aimed to examine pain and function at 3, 6, and 12 weeks for individuals with acute and subacute MSK pain who took part in a digital MSK program versus a nonparticipant comparison group.
METHODS: We conducted an observational, longitudinal study with a nonparticipant comparison group. The intervention group had video visits with physical therapists who recommended exercise therapies and educational articles delivered via an app. Nonparticipants were those who were registered but unable to participate because their benefit coverage had not yet begun. We collected pain and function outcomes through surveys delivered at 3-, 6-, and 12-week follow-ups. We conducted descriptive analyses, unadjusted regression, and mixed effects regression adjusting for baseline characteristics, time as fixed effects, and a time*group interaction term.
RESULTS: The analysis included data from 675 nonparticipants and 262 intervention group participants. Compared to baseline, the intervention group showed significantly more pain improvement at 3, 6, and 12 weeks versus nonparticipants after adjusting for baseline factors. Specifically, the intervention group\'s pain scores decreased by 55.8% at 3 weeks versus baseline, 69.1% at 6 weeks, and 73% at 12 weeks. The intervention group\'s adjusted pain scores decreased from 43.7 (95% CI 41.1-46.2) at baseline to 19.3 (95% CI 16.8-21.8) at 3 weeks to 13.5 (95% CI 10.8-16.2) at 6 weeks to 11.8 (95% CI 9-14.6) at 12 weeks. In contrast, nonparticipants\' pain scores decreased by 30.8% at 3 weeks versus baseline, 45.8% at 6 weeks, and 46.7% at 12 weeks. Nonparticipants\' adjusted pain scores decreased from 43.8 (95% CI 42-45.5) at baseline to 30.3 (95% CI 27.1-33.5) at 3 weeks to 23.7 (95% CI 20-27.5) at 6 weeks to 23.3 (95% CI 19.6-27) at 12 weeks. After adjustments, the percentage of participants reporting that pain was better or much better at follow-up was significantly higher by 40.6% at 3 weeks, 31.4% at 6 weeks, and 31.2% at 12 weeks for intervention group participants versus nonparticipants. After adjustments, the percentage of participants with meaningful functional improvement at follow-up was significantly higher by 15.2% at 3 weeks and 24.6% at 12 weeks for intervention group participants versus nonparticipants.
CONCLUSIONS: A digital MSK program may help to improve pain and function in the short term among those with acute and subacute MSK pain.