Feasibility of a randomized trial comparing different emollients for childhood eczema: findings from Choice of Moisturiser in Eczema Treatment (COMET) study

Talk Code: 

The problem

Around 20% of children have eczema and the basis of treatment is the regular application of emollients. Primary care clinicians can prescribe over 30 types of emollient, but evidence to guide choice is poor. Conducting a clinical trial to compare the effectiveness and acceptability of different emollients is potentially difficult due to uncertainty around: the best way to recruit children into the study; parent/carer willingness to use a randomly allocated emollient and be followed-up for a clinically meaningful duration; and whether researchers assessing outcomes can be kept masked to allocation.

The approach

The COMET study is a randomised trial that aims to establish the feasibility of a definitive study to compare the clinical and cost effectiveness of four commonly used emollients (Aveeno Lotion, Diprobase cream, Doublebase gel and Hydromol ointment) in primary care. Children with eczema are recruited via GP surgeries (mail-out, self-referral, or in-consultation) and are followed-up at 0, 28, 56 and 84 days. At these visits, a masked researcher undertakes objective assessments of eczema severity (including Eczema Area Severity Index). Patient-reported outcomes are collected via a daily diary, which includes a weekly POEM (Patient Orientated Eczema Measure) score.


Recruitment commenced in May 2014 and is on-going. Searches of medical records at 16 GP surgeries identified 1080 potentially eligible children. GPs excluded 158 (14.6%) children on clinical or social grounds. From 922 invitation letters, 167 responses were received (18.1%) and 2 people self-referred. By January 2015, 100 children had been randomized: 56 from the mail-out/self-referral route and 44 recruited in-consultation. Baseline characteristics of participants are: mean age 18.5 months (SD 8.9), POEM 8.5 (SD 6.4) and EASI 2.9 (SD 4.1). 11 participants have withdrawn from the study and researchers have become unmasked to 4 participants. Recruitment via the mail-out/self-referral route has been the easiest and most efficient route.Consequence: Recruitment will finish in Spring 2015 and updated figures will be presented at the conference. Data so far suggest a definitive trial is feasible: recruitment is to target and retention/unmasking rates are acceptable. While practices continue to recruit children in-consultation, changes have been made to maintain recruitment trajectory to the target sample size of 160: the upper age eligibility limit has been increased from <3 years to <5 years; and additional practices have been brought on board. The findings from this study will provide valuable new information and inform the design and delivery of the definitive trial, which will ultimately lead to better evidenced-based choices for clinicians and parents/carers.


  • Matthew Ridd, University of Bath, Bath, UK
  • Victoria Wilson, University of Bath, Bath, UK
  • Kingsley Powell, University of Bath, Bath, UK
  • Elizabeth Folkes, University of Bath, Bath, UK
  • Niamh Redmond, University of Bath, Bath, UK
  • Sandra Hollinghurst, University of Bath, Bath, UK
  • Kirsty Garfield, University of Bath, Bath, UK
  • Nicola Ball, University of Bath, Bath, UK
  • Lindsay Shaw
  • Athene Lane, University of Bath, Bath, UK
  • Chris Metcalfe, University of Bath, Bath, UK
  • Daisy Gaunt, University of Bath, Bath, UK
  • Richard Guy, University Hospitals Bristol NHS Trust, Bristol, UK
  • Sarah Purdy, University of Bath, Bath, UK