LAUSR

BACKGROUND There is no gold standard test to accurately identify patients with cellulitis and therefore misdiagnosis is common. Using the clinical impression of a dermatology or an infectious disease specialist as a reference standard, we sought to determine the prevalence of misdiagnosis of cellulitis among nonspecialist physicians. METHODS A systemic search was performed using MEDLINE, Cochrane Library, and EMBASE databases for studies reporting diagnostic accuracy of cellulitis. Inclusion criteria required dermatology or infectious disease consultation for all patients diagnosed with cellulitis by generalist physicians. We used random effects modeling to estimate the prevalence of misdiagnosis using consultant diagnosis as a reference standard. RESULTS Eight studies contributed to the analysis. For the seven studies involving inpatients, the results were sufficiently homogeneous to justify pooling data. Of 858 inpatients initially diagnosed with cellulitis, 335 (39%, 95% confidence interval: 31-47) received an alternative diagnosis from the specialist. Heterogeneity was large (I2  = 74%) and the greatest contributor to between-study variance was the year of publication. Alternative diagnoses were mostly noninfectious (68%, 221/327), with stasis dermatitis (18%, 60/327) being the most common. An abscess was the most common alternative infectious diagnosis (10%, 32/327). DISCUSSION Cellulitis is commonly misdiagnosed among inpatients, leading to unnecessary hospital admissions and antibiotic overuse. Most alternative diagnoses are noninfectious. Continuing medical education among general practitioners and urgent care providers will likely reduce cellulitis misdiagnoses.