This trial found that when treating childhood eczema that clinically looked suspicious of moderate infection, adding antibiotic tablets or creams to the usual treatment of oils, lotions, creams and corticosteroids was not clearly beneficial.
When eczema becomes infected, NICE recommends using antibiotics that are applied to the skin for small-scale infections and oral antibiotics for the treatment of widespread infection.
Reducing the use of ineffective treatments reduce harm to patients and could also reduce the development of antimicrobial resistance so it is important to know how strong the evidence is that supports the use of antibiotics when eczema looks as if it might be infected.
There are some issues to consider when interpreting this study, including the fact that the trial was slow to recruit children and stopped early with only 40% of the planned sample size. This might be a problem if the results are applied beyond those who agreed to take part in the trial. For example, it could be that parents and clinicians were not keen on treating eczema without antibiotics, and therefore those who did enrol could be unusual in some way and perhaps there was more doubt about whether infection was present.
However, given the lack of an important difference between groups, the researchers were clear that, for children with apparently infected eczema, the routine prescribing of antibiotics for all is probably unnecessary. Towards the milder end of the spectrum there may be infected eczema that improves without antibiotics. But given the uncertainties, it may be too soon to be confident in specifying which children might be managed without antibiotics.
Why was this study needed?
It was important to do this study not only because of the need to identify effective treatments for children with eczema, but also because it may be possible to reduce the wider use antibiotics, if they were shown to be ineffective. Widespread use of antimicrobials contributes to increasing antimicrobial resistance and exposes children to possible harms from adverse effects. So, it is now recognised that antibiotics are justifiable only where there is clear evidence of benefit.
Eczema affects around one in five children in the UK. Eczema is managed primarily using emollient cream to maintain moisture in the skin, relieve itchiness and prevent severe outbreaks, or “flare-ups”. When “flare-ups” occur, people are also given a short course of corticosteroids that are applied to the skin to reduce soreness and control the flare up. Skin is vulnerable to breaking and infection during a flare-up. These infections are sometimes treated using antibiotics taken by mouth (orally) or applied directly to the skin (topically).
A Cochrane systematic review, published in 2010, found a lack of clear evidence that antibiotics helped when eczema flares up and questioned their use, recommending that better and longer-term studies were needed to make a judgment.
What did this study do?
Children were recruited from 32 UK GP practices and one dermatology clinic with clinically suspected infection. Children with severe infection were excluded from the study. This analysis included 113 children, aged three months up to eight years old (average age 3.1 years).
All children received standard treatment for a flare-up of eczema, using emollient cream and topical corticosteroids. The participants were then randomly allocated to additionally receive one of the following for one week:
- oral antibiotics and placebo topical cream (36 children),
- topical antibiotics and oral placebo (37 children), or
- both oral and topical placebos to act as a control (40 children).
This study had intended to recruit 137 participants into each treatment arm, but the trial was discontinued due to slow recruitment. The final sample of 113 participants was 40% of the original target, resulting in the study being underpowered; this means it may have missed a clinically important effect, if one existed.
What did it find?
Despite the randomisation, the three groups were not matched in terms of level of severity of infection at baseline. Children in the topical antibiotics group had the most severely infected eczema to start with, and those in the control group the least severe. This potential bias would have tended to reduce the relative effectiveness of topical antibiotics.
- After two weeks, there was no significant difference between the three groups according to a parent-completed scoring, the 28-point patient-orientated eczema measure (POEM) scale. Compared to the placebo group, the difference for the oral antibiotic group was 1.52 (95% confidence interval [CI] -1.35 to 4.40) and for the topical antibiotic group was 1.49 (95% CI -1.55 to 4.53). Similar results were found at four weeks and three months. The clinically important difference for this score is said to be 3.4.
- There was no clinically significant difference in scores between the three groups on the nurse assessed 72-point eczema area and severity index (EASI) scale after two weeks. The difference in EASI score compared to the placebo group was 0.20 for the oral antibiotic group (95% CI ‑0.12 to 0.52) and 0.42 for the topical antibiotic group (95% CI 0.09 to 0.75).
- There was no difference, a limited difference or a trend towards worse outcomes in the intervention groups for all other measures, including the impact on the family, quality of life, daily symptom score and longer-term outcomes.
- Participants reported taking 61.3% of oral antibiotics or matched placebo tablets, and 81.8% of topical antibiotics or matched placebo applications, which is relatively low adherence and might also reduce the apparent effectiveness of antibiotics.
What does current guidance say on this issue?
NICE’s 2007 guideline on managing atopic eczema in under 12s recommends topical antibiotics are administered in combination with topical corticosteroids for children with localised infection. For widespread infection a one to two week course of oral antibiotics is recommended. The first choice of antibiotic is flucloxacillin, with erythromycin used for children who are allergic to penicillins or whose infection is resistant to flucloxacillin. Clarithromycin is another option if the child experiences side effects with erythromycin.
What are the implications?
Oral or topical antibiotic treatment did not improve scores on either subjective parental or objective nurse assessments of eczema severity in this trial. However, both groups of children given antibiotics had worse eczema to start with than the group given control treatments, suggesting some selection bias.
Also as the trial did not recruit as many people as planned, fewer infections led to wide confidence intervals around the results, reflecting some underlying uncertainty. It is important to note that children with severe eczema were excluded from the trial and so advice from NICE for children with flare ups of severe eczema still stands. Children with apparent infected flare ups of moderate severity eczema, as in this trial (POEM scores 13 to 17), where parents and clinicians agree to a ‘wait and see’ approach, could well be spared antibiotics, with arrangements to follow up, if necessary.
Emerging resistance to antibiotics means that antibiotics should only be prescribed when they are known to improve patient outcomes.
Citation and Funding
Francis NA, Ridd MJ, Thomas-Jones E, et al. A randomised placebo-controlled trial of oral and topical antibiotics for children with clinically infected eczema in the community: the ChildRen with Eczema, Antibiotic Management (CREAM) study. Health Technol Assess. 2016; 20(19):1-84.
This project was funded by the National Institute for Health Research HTA programme (project number 09/117/03).
DermNet NZ. EASI score. Hamilton (NZ): DermNet NZ; 2015.
NICE. Atopic eczema in under 12s: diagnosis and management. CG57. London: National Institute for Health and Care Excellence; 2007.
The University of Nottingham. How to use EASI. Nottingham: The University of Nottingham; 2014.
The University of Nottingham. POEM: patient-oriented eczema measure. Nottingham: The University of Nottingham; 2015.
Produced by the University of Southampton and Bazian on behalf of NIHR through the NIHR Dissemination Centre