AI-Driven DERMACLEAR Insights Into the Real-World Burden of Chronic Inflammatory Skin Diseases

Findings from the recent DERMACLEAR study highlight both the clinical complexity and system-wide burden of 4 chronic inflammatory skin conditions: hidradenitis suppurativa (HS), chronic urticaria (CU), psoriasis (PsO), and atopic dermatitis (AD).¹ The large-scale, real-world analysis used artificial intelligence (AI) to extract data from electronic health records (EHRs) across 7 hospitals in Spain between 2016 and 2021.

Background

According to the researchers, EHRs are considered to be one of the most complex data objects in the information processing industry.² To combat this, AI systems have been increasingly used in medicine to retrieve and analyze large amounts of data, ultimately saving time.

“Integrating AI in health care provides promising advancements in processing and analyzing big data more effectively by learning from data and identifying patterns. In particular, AI is revolutionizing dermatology research and has shown the ability to learn skin lesion features, allowing it to detect, analyze, and diagnose skin diseases,” the authors wrote.

Study Design and Demographics

Among nearly 50,000 patients with a single diagnosis, PsO was most common (58.8%), followed by AD (19.6%), CU (13.4%), and HS (8.3%). The mean patient age was 52.9 years, and more than half of the included participants were women. Across all diseases, comorbidity burden was substantial: Infections (68.7%) and respiratory disorders (59.1%) were most frequent, with notable rates of metabolic, cardiovascular, psychiatric, and neoplastic conditions, reinforcing the need for comprehensive screening and multidisciplinary management in routine practice.

Disease-specific patterns were also observed. Patients with HS were younger, more frequent smokers (58.1%), and had high rates of metabolic and inflammatory comorbidities. Patients with PsO were skewed older and demonstrated strong associations with cardiovascular and metabolic disease, whereas AD and CU populations showed high rates of respiratory and allergic comorbidities. Importantly, 11.2% of patients were recorded as deceased during the study period.

Health Care and Treatment Patterns

Health care utilization was significant. PsO accounted for the highest number of dermatology visits annually, followed by AD, CU, and HS. Nearly half of all patients required hospitalization, often likely driven by comorbidities rather than skin disease alone. Patients frequently interacted with multiple specialties, including psychiatry, internal medicine, and emergency care. The study also demonstrated a rising prevalence of all 4 diseases in dermatology settings over time, with a notable increase in 2021, possibly reflecting delayed care during the COVID-19 pandemic.

Treatment patterns revealed important gaps. Although most patients with HS (82.8%) and CU (70.7%) received some form of therapy, only about half of patients with PsO (56.2%) and AD (58.4%) had documented disease-specific treatments. Biologic use was notably low (8.4% in PsO), despite established efficacy in moderate to severe disease. Instead, reliance on nonspecific therapies such as nonsteroidal anti-inflammatory drugs and corticosteroids was common, suggesting potential undertreatment or barriers to advanced care.

Clinical Implications

A key strength of DERMACLEAR is its use of AI and natural language processing to analyze unstructured EHR data at scale, achieving high precision and enabling the detection of real-world trends not easily captured in traditional studies. The system achieved high precision (> 95%), enabling analysis at a scale not feasible with manual review. However, limitations include missing or inconsistent documentation and the inability to fully attribute outcomes like hospitalization to specific diseases.

Overall, this study highlights the high prevalence, systemic comorbidity burden, and substantial health care utilization associated with HS, CU, PsO, and AD. It also reveals underuse of advanced therapies, particularly biologics, pointing to opportunities for improving care.

For clinicians, these findings emphasize the importance of holistic, patient-centered care. Chronic inflammatory skin diseases extend beyond the skin, requiring vigilance for comorbidities, proactive treatment optimization, and coordination across specialties. The study also highlights an opportunity to improve documentation and leverage AI tools to enhance clinical insight, population health management, and, ultimately, patient outcomes.

References

1. Giménez-Arnau AM, Ortiz de Frutos FJ, Rivera-Díaz R, et al. DERMACLEAR: AI-powered insights into four chronic inflammatory skin diseases in Spain. JEADV Clinical Practice. Published online April 20, 2026. doi:10.1002/jvc2.70339

2. Sauer CM, Chen LC, Hyland SL, Girbes A, Elbers P, Celi LA. Leveraging electronic health records for data science: common pitfalls and how to avoid them. Lancet Digit Health. 2022;4(12):e893-e898. doi:10.1016/S2589-7500(22)00154-6