Acute Bacterial Cellulitis: Emergency Department Management, Aetiological Epidemiology and Clinical Risk Factors for Antibiotic Treatment Failure

This thesis consists of 6 work-packages (WPs) that describe the Emergency Department (ED) management of cellulitis; it’s aetiological epidemiology, and risk factors for antibiotic treatment failure.

In order to provide a context for this thesis, I performed a scientometric analysis of research output related to skin and soft tissue infections (SSTIs) between 1945 and 2014. Overall, research output was low but increased as a result of the community associated methicillin-resistant Staphylococcus aureus (CA-MRSA) epidemic in North American settings.

My observational research work began with two studies describing the ED incidence of cellulitis, antibiotic prescribing practices and physician adherence to cellulitis treatment guidelines. The incidence of cellulitis ranged from 6.8-12 per 1000 ED patient attendances. Lower limb cellulitis among males accounted for the majority of cases. There was poor adherence to published guideline recommendations with between 32.9-43.5% of patients treated with intravenous (IV) antibiotics despite guideline recommendation for oral treatment. This finding suggested that a clinical prediction rule (CPR) may help clinicians to risk stratify ED patients with “uncomplicated” cellulitis. I therefore focused on investigating the risk factors for oral antibiotic treatment failure among adult ED patients with “uncomplicated” cellulitis in the remaining WPs of my thesis.

In order to define potential risk factors for cellulitis treatment failure, a systematic review and meta-analysis (SRMA) of published case-control studies examining risk factors for the development of non-purulent leg cellulitis was performed. The SRMA revealed that local risk factors may be more important than systemic risk factors, including diabetes, in the pathogenesis of non-purulent leg cellulitis.

I then performed a retrospective cohort study aimed at identifying risk factors for adjustment in type, duration and setting of prescribed IV antibiotic treatment for adult patients with cellulitis, treated in the largest outpatient parenteral antimicrobial therapy (OPAT) service in Ireland. I found that OPAT was an effective and safe treatment option for cellulitis, with hospital admission occuring in 4.8% of patients and treatment adjustment in 11.7%. Increased patient age was a consistent risk factor for adjustment of OPAT and prolongation of treatment beyond 7 days, indicating the need for heightened vigilance among patients aged over 65 years.

My final study (WP-6) was a multicentre prospective cohort study describing the prevalence and predictors of oral antibiotic treatment failure among adult ED patients with cellulitis. Inter-observer reliability for patient risk factor assessment, and assessment of the eligibility and loss to follow-up rate was also performed in order to pilot study methods for any future, larger CPR derivation study. This study revealed that, depending on how treatment failure was defined, the rate of oral antibiotic treatment failure among adults discharged from 3 urban EDs with cellulitis was between 8.9% and 24.8%. Increased surface area of infection was associated with both definitions of treatment failure used in the study, and further risk factors for treatment failure are described.

This thesis concludes that, in current clinical practice, the risk stratification and antibiotic treatment of adult ED patients with cellulitis is not evidence-based. It also concludes that there is a need for a consensus definition of treatment failure for cellulitis in order to provide reliable comparison of treatment failure rates between studies, and to accurately associate risk factors with treatment failure. A number of implications for future research are discussed.