Article Text

Download PDFPDF
Randomised controlled trial
Antibiotics of no benefit to children with eczema and features of cutaneous infection but controversy remains unresolved
  1. Ting Fan Leung
  1. Department of Paediatrics, The Chinese University of Hong Kong, Shatin, Hong Kong
  1. Correspondence to: Professor Ting Fan Leung, Department of Paediatrics, The Chinese University of Hong Kong, Room 84043, Lui Che Woo Clinical Sciences Building, Prince of Wales Hospital, Shatin, Hong Kong; tfleung{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Commentary on: OpenUrl.


Eczema poses substantial adverse impact on the quality of life of patients and their families. Although most patients have cutaneous Staphylococcus aureus colonisation,1 the relationship between S. aureus and eczema flares remains unclear.2 Oral and topical antibiotics are widely prescribed to treat clinically infected eczema despite the lack of high-quality evidence to justify this practice. This multicentre randomised, double-blind, placebo-controlled trial (RDBPCT) examined whether antibiotics confer benefits to these patients.


The ChildRen with Eczema, Antibiotic Management (CREAM) study randomly assigned children with clinically infected eczema to receive 1-week treatment with: (1) oral flucloxacillin and topical placebo; (2) topical fusidic acid and oral placebo; or (3) oral and topical placebos. The children continued to use emollients and topical corticosteroids. The main outcome was Patient-Orientated Eczema Measure (POEM) 1-week after treatment completion. Secondary outcomes included Eczema Area and Severity Index (EASI) and S. aureus on skin swabs. The primary analyses compared POEM scores at 2 weeks between placebo and antibiotic groups.


One hundred and thirteen children aged 3 months to 7 years with a diagnosis of eczema and clinical suspicion of infection were recruited. Patients who received oral and topical antibiotics had negligible changes in POEM as compared to placebo (1.52 (95% CI −1.35 to 4.40) and 1.49 (95% CI −1.55 to 4.53) increase, respectively). Recipients of topical antibiotics had higher (worse) EASI 0.42 (95% CI 0.09 to 0.75) than placebo-treated patients, which was not observed among patients who received oral antibiotics (0.20 and 95% CI −0.12 to 0.52). Interventions tended to confer worse quality of life, daily symptom scores and longer term outcomes. Similar findings were obtained on sensitivity analyses following adjustment for compliance and imputation for missing data.


This RDBPCT concluded that oral and topical antibiotics had no effect or a harmful effect, on eczema severity in children with clinically infected eczema. However, patient recruitment was prematurely terminated due to slow progress. The authors rightly pointed out that this study was underpowered to detect a clinically important difference of 3.4 for POEM score.3 Accordingly, they chose to present effect sizes and CIs rather than tests of significance. It would be risky to draw definitive conclusions from this RDBPCT, although the ‘small’ effect sizes do not suggest topical and oral antibiotics to be effective in improving eczema severity in children.

This study recruited children with mild-to-moderate eczema primarily from community clinics, and excluded eczematous children with clinically severe infections or significant comorbid diseases. The majority would not be expected to develop serious complications of cutaneous staphylococcal infections, and should not be the target population for aggressive antimicrobial treatments in clinical practice. Future studies must improve the recruitment strategy for patients in order to better define a population that may benefit from antibiotic treatment.

The emergence of antimicrobial resistance is a public health concern. Fusidic acid resistance was found to be related to high levels of prescription.1 ,4 This RDBPCT observed decrease in cutaneous S. aureus carriage at 2 weeks and 3 months among antibiotic-treated patients, but it was unable to relate such changes to clinical outcomes of eczema severity. Similar proportions of patients from the three groups carried resistant S. aureus.

Implications for practice

The findings of this RDBPCT are not definitive due to a lack of study power and suboptimal selection criteria for patients. It highlights the need for further research on optimal selection of candidates for antibiotics (potentially based on biomarkers such as vitamin D5), optimal regimens (type, route, duration) and potential effectiveness of combined antibiotic/corticosteroid treatment.1


View Abstract


  • Competing interests None declared.

  • Provenance and peer review Commissioned; internally peer reviewed.