Genital and Extragenital Lichen Sclerosus et Atrophicus: A Case Series Written Using ChatGPT

Background Lichen sclerosus et atrophicus (LSEA) is a chronic inflammatory dermatosis of genital and extragenital sites with a prevalence ranging from 9% in prepubertal patients to 50% in postmenopausal patients. Chat generative pre-trained transformer (ChatGPT) is an artificial intelligence tool designed to assist humans based on supervised and reinforcement techniques. In this study, we aimed to evaluate the characteristics of patients with LSEA using ChatGPT. Methods In this retrospective study, we included all patients who presented to the outpatient dermatology department during 2017-2022 at a tertiary care teaching hospital in South India. Information regarding demographic data, characteristics of LSEA, comorbidities, and associated autoimmune disorders was gathered using a medical chart review. Following data analysis and drafting of the manuscript, the utility of ChatGPT-3 and ChatGPT-4 in finalizing the draft was assessed. Results Of 20 patients diagnosed with LSEA, 16 (80%) and four (20%) patients were females and males, respectively. Of them, 50% of female patients had attained menopause. While 65% of patients had genital LSEA, 30% of patients had extragenital LSEA only, and 5% of patients had both genital and extragenital LSEA. Furthermore, four (20%) patients were prepubertal children. Of four male patients, two (50%) were younger than 18 years of age, and one patient was diagnosed with balanitis xerotica obliterans. The commonest associated features in LSEA included joint involvement (30%), hypertension (25%), and anemia (15%). Rare concomitant disorders included psoriasis, asthma, and basal cell carcinoma over the nose. Conclusions LSEA may be confused with other various dermatoses, such as morphea, vitiligo, and lichen planus. A high index of suspicion is required, especially in children, to diagnose it early and intervene to prevent further complications. Its relationship with autoimmune disorders and comorbidities warrants further large-scale studies. ChatGPT was unreliable in the literature search due to the provision of non-existent citations. ChatGPT-4 was better than ChatGPT-3 since it provided few true publications. ChatGPT was used in this study to summarize the articles identified by the authors during the literature search and to correct grammatical errors in the final draft of the manuscript.