Comment
We found that testing positive for 4 of 7 clinical criteria for assessing cellulitis was highly specific (95%) and sensitive (100%) for a diagnosis of cellulitis among its range of mimics (Figure 3). These cellulitis criteria can be remembered, with some modification, using NEW HAvUN as a mnemonic device (New onset, Erythema, Warmth, History of associated trauma, Ache, Unilaterality, and Number of white blood cells). This aid to memory could prove a valuable tool in the efficient evaluation of a patient in an emergency, inpatient, or outpatient medical setting.
Figure 3. Clinical criteria score (1 point each for 7 clinical criteria) stratified by final diagnosis of cellulitis or noncellulitis. A score of 4 was a distinct inflection point for either clinical outcome.
Consistent with the literature, pyrexia, history of associated trauma, and unilaterality also were predictors of cellulitis diagnosis. Unilaterality often is used as a diagnostic tool by dermatologist consultants when a patient lacks other criteria for cellulitis, so these findings are intuitive and consistent with our institutional experience. Interestingly, leukocytosis was seen in only 65% of cellulitis cases and 8% of noncellulitis cases and therefore might not serve as a sensitive independent predictor of a diagnosis of cellulitis, emphasizing the importance of the multifactorial scoring system we have put forward. Additionally, acuity of onset, erythema, and tenderness are not independently associated with cellulitis when assessing a patient because several of those findings are present in other dermatologic conditions of the lower extremity; when combined with the other criteria, however, these 3 findings can play a role in diagnosis.
Effective cellulitis diagnosis provides well-recognized challenges in the acute medical setting because many clinical mimics exist. The estimated rate of misdiagnosed cellulitis is certainly well-established: 30% to 75% in independent and multi-institutional studies. These studies also revealed that patients admitted for bilateral “cellulitis” overwhelmingly tended to be stasis clinical pictures.13,19
Cost implications from inappropriate diagnosis largely regard inappropriate antibiotic use and the potential for microbial resistance, with associated costs estimated to be more than $50 billion (2004 dollars).20,21 The true cost burden is extremely difficult to model or predict due to remarkable variations in the institutional misdiagnosis rate, prescribing pattern, and antibiotic cost and could represent avenues of further study. Misappropriation of antibiotics includes not only a monetary cost that encompasses all aspects of acute treatment and hospitalization but also an unquantifiable cost: human lives associated with the consequences of antibiotic resistance.
Conclusion
There is a lack of consensus or criteria for differentiating cellulitis from its most common clinical counterparts. Here, we propose a convenient clinical correlation system that we hope will lead to more efficient allocation of clinical resources, including antibiotics and hospital admissions, while lowering the incidence of adverse events and leading to better patient outcomes. We recognize that the small sample size of our study may limit broad application of these criteria, though we anticipate that further prospective studies can improve the diagnostic relevance and risk-assessment power of the NEW HAvUN criteria put forth here for assessing cellulitis in the acute medical setting.
Acknowledgement—Author H.H.E. recognizes the loving memory of Nadia Ezaldein for her profound influence on and motivation behind this research.