Misdiagnosis of Uncomplicated Cellulitis: a Systematic Review and Meta-analysis

Rachael Nightingale*, Krishan Yadav, Laura Hamill, Paul Glasziou, Anna Mae Scott, Justin Clark, Gerben Keijzers

*Corresponding author for this work

Research output: Contribution to journalReview articleResearchpeer-review

1 Citation (Scopus)

Abstract

BACKGROUND: Cellulitis is a clinical diagnosis with several mimics and no gold standard diagnostic criteria. Misdiagnosis is common. This review aims to quantify the proportion of cellulitis misdiagnosis in primary or unscheduled care settings based on a second clinical assessment and describe the proportion and types of alternative diagnoses.

METHODS: Electronic searches of Medline, Embase and Cochrane library (including CENTRAL) using MeSH and other subject terms identified 887 randomised and non-randomised clinical trials, and cohort studies. Included articles assessed the proportion of cellulitis misdiagnosis in primary or unscheduled care settings through a second clinical assessment up to 14 days post initial diagnosis of uncomplicated cellulitis. Studies on infants and patients with (peri-)orbital, purulent and severe or complex cellulitis were excluded. Screening and data extraction was conducted independently in pairs. Risk of bias was assessed using a modified risk of bias tool from Hoy et al. Meta-analyses were undertaken where ≥ 3 studies reported the same outcome.

RESULTS: Nine studies conducted in the USA, UK and Canada, including a total of 1600 participants, were eligible for inclusion. Six studies were conducted in the inpatient setting; three were in outpatient clinics. All nine included studies provided estimates of the proportion cellulitis misdiagnosis, with a range from 19 to 83%. The mean proportion misdiagnosed was 41% (95% CI 28 to 56% for random effects model). Heterogeneity between studies was very high both statistically (I 2 96%, p-value for heterogeneity < 0.001) and clinically. Of the misdiagnoses, 54% were attributed to three conditions (stasis dermatitis, eczematous dermatitis and edema/lymphedema).

DISCUSSION: The proportion of cellulitis misdiagnosis when reviewed within 14 days was substantial though highly variable, with the majority attributable to three diagnoses. This highlights the need for timely clinical reassessment and system initiatives to improve diagnostic accuracy of cellulitis and its most common mimics.

TRIAL REGISTRATION: Open Science Framework ( https://osf.io/9zt72 ).

Original languageEnglish
Pages (from-to)1-9
Number of pages9
JournalJournal of General Internal Medicine
Early online date25 May 2023
DOIs
Publication statusPublished - 25 May 2023

Fingerprint

Dive into the research topics of 'Misdiagnosis of Uncomplicated Cellulitis: a Systematic Review and Meta-analysis'. Together they form a unique fingerprint.

Cite this