RESUMEN
BACKGROUND: ChatGPT is a free artificial intelligence (AI)-based natural language processing tool that generates complex responses to inputs from users. OBJECTIVES: To determine whether ChatGPT is able to generate high-quality responses to patient-submitted questions in the patient portal. METHODS: Patient-submitted questions and the corresponding responses from their dermatology physician were extracted from the electronic medical record for analysis. The questions were input into ChatGPT (version 3.5) and the outputs extracted for analysis, with manual removal of verbiage pertaining to ChatGPT's inability to provide medical advice. Ten blinded reviewers (seven physicians and three nonphysicians) rated and selected their preference in terms of 'overall quality', 'readability', 'accuracy', 'thoroughness' and 'level of empathy' of the physician- and ChatGPT-generated responses. RESULTS: Thirty-one messages and responses were analysed. Physician-generated responses were vastly preferred over the ChatGPT -responses by the physician and nonphysician reviewers and received significantly higher ratings for 'readability' and 'level of empathy'. CONCLUSIONS: The results of this study suggest that physician-generated responses to patients' portal messages are still preferred over ChatGPT, but generative AI tools may be helpful in generating the first drafts of responses and providing information on education resources for patients.
Asunto(s)
Dermatología , Registros Electrónicos de Salud , Procesamiento de Lenguaje Natural , Humanos , Inteligencia Artificial , Portales del Paciente , Relaciones Médico-Paciente , Médicos/psicologíaRESUMEN
Background: During the COVID-19 pandemic, teledermatology became a popular mode of health care delivery. Thus, deciphering which diagnoses are best suited for synchronous video visits is important to guide providers on appropriate patient care. Methods: We conducted a retrospective study of 1,647 submitted synchronous video visits from September 1, 2020 to March 31, 2021 at a single, large academic institution. Results: Video visits' follow-up rate was significantly associated with diagnosis subtype (p < 0.001). Compared with patients with skin lesions and nonskin dermatologic conditions, patients with a rash had higher odds of being recommended to have their follow-up visit as a video visit (odds ratio [OR] = 0.222, p < 0.001; OR = 0.296, p < 0.001). Patients with a rash had lower odds of being recommended to have their follow-up visit as an in-person office visit when compared with skin lesions (OR = 9.679, p < 0.001), nonskin dermatologic conditions (OR = 4.055, p < 0.001), and other skin dermatologic conditions (OR = 2.23, p < 0.01). Demographically, employed, middle-aged patients with private insurance made up the majority of video visit usage. African American patients were less likely to utilize a video visit compared with Asian patients (OR = 2.06, p < 0.038). Conclusions: Certain dermatologic diagnoses, most notably rashes, are more conducive to video visit management. Rashes made up 86% of new patient video visits, were more likely to have video visit follow-up if needed and were more likely to not require further follow-up indicating that the management of rashes from initial diagnosis to completion in care is suitable for video visit management.
Asunto(s)
COVID-19 , Dermatología , Enfermedades de la Piel , Telemedicina , Humanos , COVID-19/epidemiología , Estudios Retrospectivos , Femenino , Masculino , Persona de Mediana Edad , Dermatología/métodos , Enfermedades de la Piel/diagnóstico , Adulto , Anciano , Adolescente , SARS-CoV-2 , Adulto Joven , Niño , Pandemias , Preescolar , LactanteRESUMEN
ChatGPT (OpenAI) is an artificial intelligence-based free natural language processing model that generates complex responses to user-generated prompts. The advent of this tool comes at a time when physician burnout is at an all-time high, which is attributed at least in part to time spent outside of the patient encounter within the electronic medical record (documenting the encounter, responding to patient messages, etc). Although ChatGPT is not specifically designed to provide medical information, it can generate preliminary responses to patients' questions about their medical conditions and can precipitately create educational patient resources, which do inevitably require rigorous editing and fact-checking on the part of the health care provider to ensure accuracy. In this way, this assistive technology has the potential to not only enhance a physician's efficiency and work-life balance but also enrich the patient-physician relationship and ultimately improve patient outcomes.
RESUMEN
Objective: To determine if dermoscopy, a technique widely utilized in dermatology for improved diagnosis of skin lesions, can be used comfortably for evaluating periorbital, eyelid, and conjunctival lesions. Design: Proof-of-concept study in which a technique for performing dermoscopy near the eye was developed, related educational material was prepared, and a protocol for dermoscopic image capture was created. Methods: Technicians used the developed materials to learn to take high-quality pictures with a 10x dermoscope attached to a standard cell phone camera. The images were assessed for diagnostic utility by an oculoplastic surgeon and two dermatologists. Participants: 115 patients recruited from ophthalmology clinics from July 2021 to April 2023 were photographed, yielding 129 lesions with high-quality dermoscopic images as assessed by an oculoplastic surgeon and two dermatologists. Results: Technicians reported a significant increase in confidence (measured on a 1-10 scale) with dermoscopy after training (pre-instruction mean = 1.72, median = 1, mode = 1, IQR = 1.25 vs mean = 7.69, median = 7.75, mode = 7 and 8, IQR = 1.5 post-instruction. Wilcoxon rank sum test with continuity correction, W = 0, p < 0.001, paired t = 13.95, p < 0.0001). Incorporating a contact plate with a 4 × 4mm reticule on the dermoscope aided in photographing ocular and periocular lesions. Conclusion: Medical support staff in eye-care offices can be taught to use dermoscopes to capture high-quality images of periorbital, eyelid, and conjunctival lesions. Dermoscopy illuminates diagnostic features of lesions and thus offers a new avenue to improve decision-making in ophthalmology. Dermoscopy can be incorporated into telemedicine evaluations by ophthalmologists, oculoplastic surgeons, or affiliated dermatologists for triage of or rendering advice to patients and for planning of surgery if needed.