Despite several expert panel recommendations and cellulitis treatment guidelines, there are currently no clinical decision rules to assist clinicians in decideding which Emergency Department (ED) patients should be treated with oral antibiotics and which patients require IV therapy at first presentation of cellulitis amenable to outpatient treatment.

Cellulitis is an acute or subacute infection of the dermis and subcutaneous tissue of presumed bacterial etiology charactrerized by warmth, erythema, swelling, and tenderness. Physician visits for cellulitis increased from 32 to 48 visits per 1,000 patients from 1997 to 2005 with the largest relative increase occurring in EDs (Hersh 2008). The majority of publications reporting risk factors for cellulitis have been case-control studies of patients hospitalized with cellulitis. These studies showed that disruption of the cutaneous barrier, venous insufficiency, leg edema, obesity, fungal infections, history of cellulitis, prior saphenetomy, white race, and the presence of Staph aureus and/or B-hemolytic Streptococci in the toe webs to be independently associated with cellulitis (Dupuy 1999). A retrospective study of admitted patients with cellulitis reported that lower extremity edema, BMI, smoking and homelessness were independently associated with recurrent cellulitis (Lewis 2006).

Although these studies have reported potential risk factors associated with the development of cellulitis, there is very little data on risk factors for treatment failure of cellulitis in the ED patient population.

Clinical Question

What risk factors are independently associated with failure of empiric oral antibiotic therapy in ED patients initially treated as outpatients.


Adult patients over 18 years old presenting to one of two EDs at an academic tertiary care center in London, Ontario, Canada whose chief complaint were consistent with skin or soft tissue infections who were invited to participate in this study once an Emergency Physician confirmed a cellulitis infection.


Patients filled out a health and lifestyle questionnaire also recorded were any physical exam findings and the triage vitals. Pts were treated either with PO antibiotics, IV antibiotics (the hospital had an outpatient IV antibiotic infusion clinic) or a combination of both PO and IV. Patients were then reached two weeks after the initial visit to evaluate response to antibiotics.


Treatment failure was defined as worsening cellulitis requiring either hospitalization or changing antibiotics.


Prospective cohort, multicenter convenience sample


Patients who required hospitalization for cellulitis, those who had taken antibiotics prior to initial evaluation, presence of an abscess, cognitive impairment and those unable to speak English.

Primary Results

  • 497 patients met inclusion criteria and were discharged with antibiotics
  • Total of 102 patients had treatment failures
  • Of the 185 patients discharged on oral antibiotics 39 (21%) were deemed treatment failure
  • Of the 81 patients discharged on both oral and IV antibiotics 22 (27%) were deemed treatment failure
  • Of the 231 patients discharged on IV antibiotics 41 (17%) were deemed treatment failure

Critical Findings

  • Univariate analyses of all potential patient risk factors were completed to allow for multi-variable logistic regression models to determine predictor variables independently associated with treatment failure
  • Independent predictors of treatment failure include: Fever >38C at triage (OR:4.3), chronic leg ulcers (OR:2.5), chronic edema or lymphedema (OR:2.5), prior cellulitis in the same area (OR:2.1), cellulitis at wound site (OR:1.9).


  • Large prospective cohort study with <10% of patients lost to follow up
  • Well defined criteria for treatment failure


  • No data regarding patients who were initially admitted for cellulitis. Unclear what prompted the inpatient admission vs decision to treat outpatient.
  • Certain patient characteristics are not well represented such as HIV, cancer, kidney disease, liver disease and hypoxia, so we cannot extrapolate these findings to those patients.
  • The wide range of antibiotics prescribed at the physician’s discretion makes it unclear what role the choice of antibiotics had in terms of treatment failure.
  • MRSA rate is 13.2% in their patient population which is less pervasive compared to other communities
  • Convenience sample of patients only taken from 7am to 10 pm

Author's Conclusions

“This is the largest prospective study to date evaluating potential risk factors for adult patients presenting to the ED with cellulitis who are initially treated as outpatients who fail initial antibiotic therapy and require a change of antibiotics or admission to the hospital. Fever at triage, chronic leg ulcers, chronic edema or lymphedema, prior cellulitis in the same area and cellulitis at wound site were independently associated with failure of empiric antibiotic therapy in ED patients with cellulitis. These risk factors should be considered when initiating empiric outpatient antibiotic therapy for patients with cellulitis.

Our Conclusions

The authors of this paper were able to identify certain independent predictors of outpatient cellulitis treatment failure based on multivariable analysis using data obtained from history and physical exam. However one possible predictor of treatment failure was the choice of antibiotic regimen that was left to individual physician discretion. Therefore, it is unclear what percentage of treatment failures were secondary to improper antibiotics. That being said, the five independent predictors of treatment failure, especially fever, should be strongly considered as reasons for potential hospital admission for treatment of cellulitis.

Potential Impact To Current Practice

This study provides some guidance on features that may predict treatment failure and can be used to aid in disposition decisions.

Bottom Line

Patients with cellulitis presenting with fever, chronic leg ulcers, edema or lymphedema, prior cellulitis to same area, or cellulitis at wound site have a statistically increased risk of outpatient treatment failure.

Read More


Hersh AL. National trends in ambulatory visits and antibiotic prescribing for skin and soft-tissue infections. Arch Intern Med 2008;168;1585-91. PMID: 18663172

Dupuy A et al, Risk factors for erysipelas of the leg; case-control study. BMJ 1999;318:1591-4. PMID: 10364117

Lewis SD et al. Risk factors for recurrent lower extremity cellulitis in a US veterans medical center population. Am J Med Sci 2006;332:304-7. PMID: 17170620