Massive EHR Analysis Reveals Why Atopic Dermatitis Patients Flood Emergency Rooms Instead of Dermatology Clinics

Atopic dermatitis, the most common form of eczema, affects millions of Americans and imposes a relentless burden of itching, sleep disruption, and skin infection. For many patients, the question is not whether to seek care but where to seek it, and a new analysis of more than 630,000 clinical encounters suggests that the answer is increasingly the emergency department rather than the dermatology clinic. The study, conducted by researchers at Stanford University School of Medicine and the George Washington School of Medicine and Health Sciences and published in Archives of Dermatological Research, mined the Epic Cosmos electronic health record database to compare how patients with atopic dermatitis are treated across emergency and outpatient settings, and the findings paint a striking picture of fragmented, unequal access to specialty skin care in the United States.

The research team, led by Mihir M. Shah and corresponding author Gordon H. Bae, performed a cross-sectional analysis of 632,916 encounters tied to atopic dermatitis within Epic Cosmos, one of the largest aggregated electronic health record networks in the world. Epic Cosmos pools de-identified data from hundreds of health systems that use Epic’s software, giving investigators a longitudinal, nationwide window into real-world care patterns that no single institution could provide. Because the dataset is de-identified and aggregated, the study constituted a secondary analysis of existing records rather than human subjects research, and institutional review board approval was not required. The sheer scale of the cohort matters: with over half a million encounters, the analysis can detect utilization differences and demographic patterns that smaller, single-center studies routinely miss.

The technical logic of a cross-sectional EHR design is straightforward but powerful. Rather than following patients forward in time, the investigators snapshot a defined period and classify each encounter by care setting, comparing emergency department visits with outpatient dermatology visits for the same diagnosis. By leveraging standardized diagnostic coding across participating institutions, the team could quantify the proportion of atopic dermatitis care delivered in emergency settings, characterize the patients who end up there, and contrast the clinical resources consumed in each environment. Emergency department care for a chronic inflammatory skin disease is, on its face, a mismatch: emergency physicians are trained to stabilize acute conditions, not to manage the long-term topology of a relapsing dermatologic disease that requires emollient regimens, topical anti-inflammatory therapy, phototherapy, and increasingly sophisticated systemic agents.

Why do patients with a chronic rash end up in the emergency department at all? The new study sits within a well-documented context of dermatologic access failure in the United States. Prior work has shown that dermatologist density varies dramatically between urban and rural counties, that appointment wait times and insurance acceptance differ across practice types, and that patients from racial and ethnic minority and underserved populations face disproportionate barriers to allergy and immunology care. When a patient’s eczema flares at two in the morning and the next dermatology appointment is months away, the emergency room becomes the only open door. Earlier research has also quantified the financial toll: emergency department visits for atopic dermatitis carry substantial costs to patients and the health system, costs that are arguably avoidable when disease is controlled in the outpatient setting.

The Cosmos findings give that national picture an unprecedented empirical foundation. With 632,916 encounters analyzed, the study provides one of the largest published snapshots of where atopic dermatitis care actually happens in American medicine. The comparison between settings is not merely descriptive bookkeeping. Emergency department encounters for a chronic skin condition signal a system-level failure of continuity: the patient arrives in crisis, receives episodic treatment, and is discharged without the longitudinal disease management that guidelines from the American Academy of Allergy, Asthma and Immunology and the American College of Allergy, Asthma and Immunology recommend. Those 2023 guidelines emphasize stepped-care management, trigger avoidance, skin barrier maintenance, and early escalation to systemic therapies for moderate to severe disease, all of which depend on an established outpatient relationship that emergency visits cannot supply.

The demographic dimension of the analysis is where the public health stakes become sharpest. Studies of allergic and immunologic disease have repeatedly documented that underserved racial and ethnic populations experience higher disease burden and lower access to specialty care, and atopic dermatitis is no exception. If the Cosmos data show that emergency departments absorb a disproportionate share of atopic dermatitis encounters among particular insurance categories, age groups, or communities, the implication is that the emergency room is functioning as a safety-net dermatology clinic of last resort. That substitution is clinically suboptimal and economically inefficient, because the emergency setting delivers the most expensive per-encounter care in the health system while offering the least continuity for a disease whose successful management is fundamentally longitudinal.

For dermatologists and health policy analysts, the study offers a measurable benchmark against which interventions can be judged. Teledermatology, for example, has been proposed as a way to extend specialty expertise into underserved regions and after-hours windows, potentially diverting flare-related visits away from emergency departments. Same-week urgent dermatology slots, integrated itch clinics, and pharmacist-guided topical therapy refills represent other structural responses. The value of a dataset like Epic Cosmos is that it establishes the baseline: how many encounters, in which settings, among which populations. Without that baseline, it is impossible to know whether an intervention actually moved the needle. With it, health systems can set concrete targets, such as reducing the fraction of atopic dermatitis care delivered in emergency settings year over year.

The study also showcases a methodological shift reshaping clinical research. Epic Cosmos aggregates billing, diagnostic, and demographic data across a vast consortium of health systems, allowing cross-sectional analyses at a scale that approaches population-level surveillance. Such datasets democratize research: a research letter from a dermatology department can now characterize national care patterns that once required multi-year, multi-center prospective registries costing millions of dollars. The trade-offs are real, however. EHR data reflect care that was documented, not care that was needed but never sought, and coding practices vary across institutions. Cross-sectional designs capture association, not causation, and cannot fully disentangle whether patients choose emergency care, are forced into it, or are triaged there by constrained outpatient capacity. The authors acknowledge these constraints by framing the work as a cross-sectional analysis rather than a causal study.

What emerges from the 632,916 encounters is ultimately a systems diagnosis rather than a laboratory finding. Atopic dermatitis is a controllable chronic disease with an expanding therapeutic arsenal, from topical calcineurin inhibitors to biologic and oral systemic agents. Yet when a substantial share of its care migrates into emergency departments, the message is that the routine, relationship-based infrastructure needed to deploy that arsenal is not reaching everyone. The Stanford-led team, which reported no conflicts of interest and no dedicated funding for the work, offers the analysis as a call to quantify and close the gap between where eczema care is delivered and where it should be. As electronic health record networks grow and therapeutic options multiply, studies of this scale will increasingly define the frontier of dermatologic health services research, turning the humble rash into a lens through which the inequities of American health care come into sharp focus.

Subject of Research: Emergency department versus outpatient dermatology care utilization for atopic dermatitis analyzed across 632,916 encounters in the Epic Cosmos electronic health record database.

Article Title: Emergency department vs. outpatient dermatology care for atopic dermatitis: a cross-sectional analysis of 632,916 encounters in epic cosmos

Article References: Shah, M. M., Pour Mohammad, A., Bhatt, M., Ko, J. M., & Bae, G. H. (2026). Emergency department vs. outpatient dermatology care for atopic dermatitis: a cross-sectional analysis of 632,916 encounters in epic cosmos. Archives of Dermatological Research, 318(1), Article 454. https://doi.org/10.1007/s00403-026-04957-1

Image Credits: AI Generated

DOI: 10.1007/s00403-026-04957-1

Keywords: atopic dermatitis, eczema, emergency department, outpatient dermatology, Epic Cosmos, electronic health records, health care utilization, dermatologist access, health disparities, cross-sectional analysis, Emergency, department