Customize your JAMA Network experience by selecting one or more topics from the list below.
Kroshinsky D, Cotliar J, Hughey LC, Shinkai K, Fox LP. Association of Dermatology Consultation With Accuracy of Cutaneous Disorder Diagnoses in Hospitalized Patients: A Multicenter Analysis. JAMA Dermatol. 2016;152(4):477–480. doi:10.1001/jamadermatol.2015.5098
Limited information exists on the activity and impact of hospitalist dermatology consultative services or the nature of dermatologic issues affecting inpatients across US academic medical centers.1-4 We performed a retrospective, multi-institution, cross-sectional cohort study assessing consultations called to 4 full-time dermatology consult teams across the country over the course of 12 months to evaluate the impact of consultative dermatologists on the diagnosis and treatment of hospitalized patients.
Institutional review board approval was obtained at all sites. The 1661 inpatient dermatology consultations performed by full-time inpatient dermatologists with a resident team within 24 to 48 hours were recorded at Massachusetts General Hospital (MGH) (n = 684), University of California Los Angeles (UCLA) (n = 336), University of California San Francisco (UCSF) (n = 385), and the University of Alabama at Birmingham (UAB) (n = 336) for 12 months from 2008 to 2009. The service requesting consultation (primary team) was recorded as Medicine, Surgery, Emergency Department (ED), Intensive Care Units (ICU), Obstetrics-Gynecology, Pediatrics, Psychiatry, or Neurology, and provided its presumptive diagnoses at the time of request. All final diagnoses rendered by dermatology consultants were made based on clinical history, examination findings, and laboratory testing. If more than 1 diagnosis was identified, the additional diagnoses were also recorded. Dermatologic concern duration was defined as acute if present for 7 days or less, subacute if present for 8 to 30 days, chronic if present for 1 to 11 months, and longstanding if present for 1 year or longer. Agreement between the primary team’s and the dermatology team’s diagnosis was determined using 4 categories: “agreement,” “disagreement,” “agreement but identifying at least a second diagnosis on consultation,” and “disagreement but identifying at least a second diagnosis on consultation.” The frequency of follow-up evaluations, defined as repeat visits by the dermatology attending physician to the patient’s bedside, was also recorded. The difference in agreement rate between groups was assessed using a χ2 test. P values were 2-sided and considered significant at the .05 level.
Table 1 lists the patient characteristics and diagnoses. A broad spectrum of dermatologic diagnoses was rendered by the inpatient dermatology services, including inflammatory and neoplastic processes. The agreement between the primary team’s preliminary diagnosis and actual diagnosis per consulting team and by diagnosis are listed in Table 2. The most common consulting team was Medicine (n = 939, 56%), followed by Surgery (n = 251, 15%), ICU (n = 195, 12%), and ED (n = 92, 6%). Cellulitis, leg ulceration, and viral infections were the 3 most commonly undiagnosed and misdiagnosed conditions by the referring team. The majority of primary team preliminary diagnoses included “rash”/unknown (n = 814), followed by cellulitis/abscess (n = 115), and “drug rash” (n = 111), whereas the majority of primary diagnoses by the dermatologist included “drug rash” (n = 292), psoriasis/eczema (n = 170) and benign neoplasm (n = 168). Additional cutaneous issues were identified in 298 (18%) of consults. Diagnosis was confirmed by biopsy in 667 (40.2%) patients. Four hundred (30.9%) patients were admitted because of their skin condition while in the remaining cases, the dermatologic issue was found incidentally or developed during hospitalization. The cutaneous issue duration was recorded in 1226 cases and was acute in 640 patients (52.2%), subacute in 241 patients (19.7%), chronic in 212 patients (17.3%), and longstanding in 233 patients (10.8%). Importantly, dermatology consultation changed the final diagnosis in 71% of consultation requests. Dermatology-specific evaluation and treatment recommendations could be rendered in a single consultation visit in 40.4% of cases (n = 498) or with 1 follow-up evaluation in 29.3% (n = 361) of cases.
This research demonstrates that dermatology consultation is associated with improved diagnostic accuracy of cutaneous disorders in hospitalized patients and facilitates early appropriate intervention, suggesting a potential approach to improving care. The results highlight the positive impact of hospitalist dermatology care in the hospital and reinforce the validity of this model for dermatology specialty care.5 The data presented identify potential practice gaps and areas for targeted education on dermatology topics relevant to primary inpatient teams.
Corresponding Author: Daniela Kroshinsky, MD, MPH, Department of Dermatology, Massachusetts General Hospital, 50 Staniford St, 200 Boston, MA 02114 (firstname.lastname@example.org).
Accepted for Publication: October 23, 2015.
Published Online: January 13, 2016. doi:10.1001/jamadermatol.2015.5098.
Author Contributions: Drs Kroshinsky and Fox 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: All authors.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Kroshinsky, Cotliar.
Critical revision of the manuscript for important intellectual content: All authors.
Obtained funding: Kroshinsky.
Administrative, technical, or material support: Kroshinsky, Cotliar, Hughey.
Study supervision: Kroshinsky, Hughey, Shinkai.
Conflict of Interest Disclosures: None reported.
Additional Contributions: We are indebted to the Dermatology Foundation for their support, and to Kristen Stevenson, PhD candidate, Dana-Farber, Harvard Cancer Center for her assistance with the statistical analysis. She was financially compensated for her contribution.
Create a personal account or sign in to: