Figure. Breakdown of initial diagnoses over the study period.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Lee HY, Walsh S, Creamer D. Initial Presentation of DRESS: Often Misdiagnosed as Infections. Arch Dermatol. 2012;148(9):1085–1087. doi:10.1001/archdermatol.2012.1079
Author Affiliations: Department of Dermatology, King's College Hospital, Denmark Hill, London, England.
Drug reaction eosinophilia and systemic symptoms (DRESS) is a severe cutaneous event characterized by a skin eruption, fever, hematologic abnormalities (eosinophilia or atypical lymphocytes), and internal organ involvement.1 DRESS may be complicated by multiple-organ failure requiring treatment in the intensive care unit.2 The overall mortality rate has been estimated at 10%,1,3 although more recent studies have suggested a rate nearer 2%.4 Early recognition of this syndrome, discontinuation of treatment with the suspected drug, and the institution of supportive and/or specific measures are vital.
However, the heterogeneity of initial presentation often results in the patient being initially managed by nondermatologists. In DRESS, the latency period is long (delay between initiation of treatment with the culprit drug and onset of reaction), between 3 weeks and 2 months, which further obscures drug culpability. The failure to recognize and manage DRESS as an adverse drug reaction was recently proposed in the Archives as a “Practice Gap.”5 The aim of our study was to review the presentation, initial diagnosis, and early management of DRESS in our institution and to determine if such a practice gap exists.
A retrospective review was performed on cases of DRESS managed in our institution from 2005 through 2011. All patients were recruited onto the SCAR-UK registry (Severe Cutaneous Adverse Reactions, United Kingdom). All cases were defined as probable or definite DRESS based on the RegiSCAR diagnostic criteria.6 Patients were either admitted following presentation or were already inpatients at the time of diagnosis. Case records were systematically reviewed, and data on initial clinical presentation, laboratory parameters, initial diagnosis, and treatment were collected. Diagnostic lag was defined as the interval from initial diagnosis at presentation to time of DRESS diagnosis. This study was approved by our institutional review board at King's College Hospital, London, England.
There were a total of 26 patients with DRESS included in the analysis (11 male; 15 female) with a mean age of 42 years (age range, 7-73 years). The validation scores for DRESS cases according to RegiSCAR criteria are listed shown in Table 1. The clinical features and laboratory findings at presentation are summarized in Table 1. Fever and malaise affected more than 75% of patients (Table 2). Seventeen patients initially presented to medical departments (internal medicine, 7; dermatology, 3; rheumatology, 2; hepatology, 2; pediatrics, 1; hematology, 1; and chest medicine, 1). Four patients presented to neurosurgery; 2 patients were on the intensive care ward; and 3 patients were initially seen in the emergency department.
The initial diagnosis was presumed to be solely infection in 13 of 26 patients (50%), which led to treatment with antibiotics (Figure). Lymphoma and drug hypersensitivity were the other initial diagnoses suspected (Figure). The mean lag between the initial diagnosis and the diagnosis of DRESS was 1.7 days. The diagnostic accuracy varied over the study period (2005-2011), as shown in the Figure. The proportion of cases presumed to be infection-related decreased from 0.6 in the 2005-2007 period to 0.5 in the 2007-2009 period to 0.38 in the 2010-2011 period.
The clinical presentation of DRESS is heterogeneous. The constellation of fever, constitutional symptoms, eruption, and multiple-organ involvement led the initial consulting doctors in our study to consider an infectious illness as the primary diagnosis in 13 of 26 patients. Infection was also implicated in an additional 6 patients, although other differential diagnoses were considered. Drug hypersensitivity was deemed as the most likely diagnosis in only 7 patients (27%), and an initial diagnosis of DRESS was never made.
The initial misdiagnosis of DRESS led to an average diagnostic delay of 1.7 days. This short delay is likely attributed to the fact that in our institution, liaison inpatient dermatology services are available 24 hours a day. On analyzing the diagnostic accuracy over the study period, we found that an increasing proportion of cases were considered to have an underlying drug cause. This improvement most likely reflects our efforts to educate our medical colleagues about the drug-induced dermatoses and dissemination of information on DRESS in particular.
Our study further illustrates how difficult the diagnosis of DRESS can be. A close cooperation between dermatologists and other hospital physicians may decrease the delay in diagnosis as well as bring about a greater awareness of this syndrome.
Correspondence: Dr Creamer, Department of Dermatology, King's College Hospital, Denmark Hill, London, SE5 9RS United Kingdom (email@example.com).
Author Contributions: Drs Lee and Creamer had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Creamer. Acquisition of data: Lee, Walsh, and Creamer. Analysis and interpretation of data: Lee, Walsh, and Creamer. Drafting of the manuscript: Lee. Critical revision of the manuscript for important intellectual content: Walsh and Creamer. Administrative, technical, and material support: Walsh. Study supervision: Walsh and Creamer.
Financial Disclosure: None reported.