Cellulitis – Observe, Admit, or Discharge?

Cellulitis is commonly diagnosed and treated in the Emergency Department, with considerable variation in treatment and disposition patterns. These patients are either:

  1. Given a dose of intravenous (IV) antibiotics before discharge home
  2. Placed on observation units for IV antibiotics before transitioning to oral medications prior to discharge
  3. Discharged with oral medications
  4. Admitted for continuous IV antibiotics

Is there evidence to help guide our decision-making for these patients?



Should I Stay or Should I Go?

The Infectious Disease Society of America (IDSA) released guidelines in 2014 for management of skin and soft tissue infections. It is important to remember there are only a few Emergency Department based studies used to create these guidelines, which may limit their generalizability to an ED population. Let’s focus our discussion on nonpurulent cellulitis, since the main treatment for purulent infections is incision and drainage.

IDSA algorithm for SSTI (1)

Per their algorithm, PO antibiotics are suitable for nonpurulent cellulitis if the patient lacks systemic signs of infection and is immunocompetent, hemodynamically stable, and with normal mental status. The IDSA defined systemic signs of infection as essentially the SIRS criteria (fever >38°C, tachycardia >90, tachypnea >24, WBC >12,000 or <4000). Note that the commonly prescribed trimethoprim-sulfamethoxazole is not recommended as a first-line agent for mild SSTI and that only one of the oral antibiotics (clindamycin) has MRSA activity. Those with moderate or severe nonpurulent cellulitis that have systemic signs of infection should receive IV antibiotics (1). Diabetics, while at increased risk for infections, are not considered immunocompromised in these guidelines and there are separate guidelines for diabetic foot infections.


If I Go There Will Be Trouble

Abetz et al recently performed a systematic review looking at SSTI management failures in ED observation units, with a primary outcome of admission, greater than 24 hours stay in ED observation, or death. Nine studies fit their criteria. Their conclusion from these studies was that the presence of fever, elevated inflammatory markers (such as leukocytosis, CRP, ESR, or lactate), or a known MRSA exposure increased the risk for the primary outcome. Two of these studies found that hand cellulitis may be more prone to treatment failure (2).


A well-done, prospective cohort study by Peterson et. al looked at predictors of failure for empiric outpatient antibiotic treatment from the Emergency Department. The primary outcome was admission or failed initial antibiotic regimen. This study had a treatment failure rate of 20.5% and reported 18 different antibiotic regimens using different combinations of IV and oral medications (3).

Risk Factor Odds Ratio 95% confidence Interval
Fever at triage 4.3 1.6 to 11.7
Chronic Leg Ulcers 2.5 1.1 to 5.2
Chronic edema/lymphedema 2.5 1.5 to 4.2
Prior cellulitis in same area 2.1 1.3 to 3.5
Wound cellulitis 1.9 1.2 to 3.0


And If I Stay It Will Be Double

Patients placed on observation, short stay units, or admitted to the hospital for cellulitis will likely receive IV antibiotics. This comes at an increased cost to the healthcare system and, often, an increased risk to the patient. Patients are subjected to additional testing, exposure to other pathogens in the hospital, and further iatrogenic harm. The IV route of antibiotic administration is associated with higher rates of phlebitis, bacteremia, and fluid overload. Most antibiotics utilized for cellulitis have excellent bioavailability with favorable pharmacokinetics that would not benefit from the faster peak levels attainable by the intravenous route (4,5). Haran et. al performed a prospective multicenter study that evaluated rates of antibiotic-associated diarrhea (AAD) and found that 25.7% of patients who received IV antibiotics developed AAD compared to 12.3% of patients treated only with oral antibiotics. The therapeutic failure rates, ESI level, proportion with tachycardia or fever, and demographics were not statistically significant between groups. Interestingly, in this study, Hispanic heritage had an odds ratio of 3.04 for the development of AAD compared to other ethnicities (6).

Pharmacokinetics of Antibiotics (4)

In fact, there is no RCT that clearly demonstrates the superiority of IV over PO antibiotics. Most of the studies comparing IV to PO antibiotics are in pediatrics, with few examining the difference in cellulitis. Li et al reviewed multiple studies that demonstrated no reduction in treatment failures of IV antibiotics for pediatric and adult pneumonia, pediatric pyelonephritis, low-risk febrile neutropenia, infective endocarditis, chronic osteomyelitis, and acute pediatric osteomyelitis (5). For cellulitis, Bernard performed two open-label noninferiority studies, one decade apart, comparing IV penicillin vs. oral roxithromycin (a macrolide) and oral pristinamycin (a streptogramin). Both drugs are similar to erythromycin and are unfortunately not available in the United States. It should be noted that pristinamycin has activity against MRSA, which is a significant confounder. In these studies, there were no significant differences between groups for patients withdrawing due to adverse events (7,8).


So Come On And Let Me Know

Based on the existing literature, which is limited, clinicians should reserve IV antibiotics for patients with severe cellulitis. Those with moderate cellulitis should be considered for empiric outpatient treatment if they do not have any of  the risk factors for failed outpatient therapy listed below:



Peer Reviewers: Wendy Chan, MD and Raul Hernandez, MD

Faculty Advisor: Ian deSouza, MD



  1. Dennis L. Stevens, Alan L. Bisno, Henry F. Chambers, E. Patchen Dellinger, Ellie J. C. Goldstein, Sherwood L. Gorbach, Jan V. Hirschmann, Sheldon L. Kaplan, Jose G. Montoya, and James C. Wade. Practice Guidelines for the Diagnosis and Management of Skin and Soft Tissue Infections: 2014 Update by the Infectious Diseases Society of America. Clin Infect Dis. (2014) 59 (2): e10-e52
  2. Abetz J, Adams N, Mitra B. Skin and soft tissue infection management failure in the emergency department observation unit: a systematic review. Emergency Medicine Journal: EMJ [serial online]. October 25, 2016
  3. Peterson D, McLeod S, Woolfrey K, McRae A. Predictors of failure of empiric outpatient antibiotic therapy in emergency department patients with uncomplicated cellulitis. Academic Emergency Medicine: Official Journal Of The Society For Academic Emergency Medicine [serial online]. May 2014;21(5):526-531.
  4. MacGregor R, Graziani A. Oral administration of antibiotics: a rational alternative to the parenteral route. Clinical Infectious Diseases: An Official Publication Of The Infectious Diseases Society Of America [serial online]. March 1997;24(3):457-467.
  5. Li H, Agweyu A, English M, Bejon P. An Unsupported Preference for Intravenous Antibiotics. Plos Medicine [serial online]. May 2015;12(5):1-7. Available from: Academic Search Complete, Ipswich, MA. Accessed November 26, 2016.
  6. Haran J, Hayward G, Boyer E, et al. Original Contribution: Factors influencing the development of antibiotic associated diarrhea in ED patients discharged home: risk of administering IV antibiotics. American Journal Of Emergency Medicine [serial online]. October 1, 2014;32:1195-1199.
  7. Bernard P, Plantin P, Scheimberg A, et al. Roxithromycin versus penicillin in the treatment of erysipelas in adults: a comparative study. The British Journal Of Dermatology [serial online]. August 1992;127(2):155-159
  8. Bernard P, Chosidow O, Vaillant L. Oral pristinamycin versus standard penicillin regimen to treat erysipelas in adults: randomised, non-inferiority, open trial. BMJ : British Medical Journal. 2002;325(7369):864.
The following two tabs change content below.


Eden Kim, DO, MPH PGY-3 Emergency Medicine Resident

Latest posts by edenkim (see all)

2 comments for “Cellulitis – Observe, Admit, or Discharge?

  1. Anonymous
    November 28, 2016 at 10:57 pm

    Literally had a case of non-purulent cellulitis today (and everyday…)

    Strong work Dr. Kim

  2. November 29, 2016 at 1:37 pm

    Awesome post Eden!

    Doxycycline is another PO med for MRSA skin infection, probably underutilized because it has little anti-strep activity.


Leave a Reply

Your email address will not be published.