Health Technology Assessment (Oct 2023)

Comparison of lotions, creams, gels and ointments for the treatment of childhood eczema: the BEE RCT

  • Matthew J Ridd,
  • Sian Wells,
  • Stephanie J MacNeill,
  • Emily Sanderson,
  • Douglas Webb,
  • Jonathan Banks,
  • Eileen Sutton,
  • Alison RG Shaw,
  • Zoe Wilkins,
  • Julie Clayton,
  • Amanda Roberts,
  • Kirsty Garfield,
  • Lyn Liddiard,
  • Tiffany J Barrett,
  • J Athene Lane,
  • Helen Baxter,
  • Laura Howells,
  • Jodi Taylor,
  • Alastair D Hay,
  • Hywel C Williams,
  • Kim S Thomas,
  • Miriam Santer

DOI
https://doi.org/10.3310/GZQW6681
Journal volume & issue
Vol. 27, no. 19

Abstract

Read online

Background Emollients are recommended for children with eczema (atopic eczema/dermatitis). A lack of head-to-head comparisons of the effectiveness and acceptability of the different types of emollients has resulted in a ‘trial and error’ approach to prescribing. Objective To compare the effectiveness and acceptability of four commonly used types of emollients for the treatment of childhood eczema. Design Four group, parallel, individually randomised, superiority randomised clinical trials with a nested qualitative study, completed in 2021. A purposeful sample of parents/children was interviewed at ≈ 4 and ≈ 16 weeks. Setting Primary care (78 general practitioner surgeries) in England. Participants Children aged between 6 months and 12 years with eczema, of at least mild severity, and with no known sensitivity to the study emollients or their constituents. Interventions Study emollients sharing the same characteristics in the four types of lotion, cream, gel or ointment, alongside usual care, and allocated using a web-based randomisation system. Participants were unmasked and the researcher assessing the Eczema Area Severity Index scores was masked. Main outcome measures The primary outcome was Patient-Oriented Eczema Measure scores over 16 weeks. The secondary outcomes were Patient-Oriented Eczema Measure scores over 52 weeks, Eczema Area Severity Index score at 16 weeks, quality of life (Atopic Dermatitis Quality of Life, Child Health Utility-9 Dimensions and EuroQol-5 Dimensions, five-level version, scores), Dermatitis Family Impact and satisfaction levels at 16 weeks. Results A total of 550 children were randomised to receive lotion (analysed for primary outcome 131/allocated 137), cream (137/140), gel (130/135) or ointment (126/138). At baseline, 86.0% of participants were white and 46.4% were female. The median (interquartile range) age was 4 (2–8) years and the median Patient-Oriented Eczema Measure score was 9.3 (SD 5.5). There was no evidence of a difference in mean Patient-Oriented Eczema Measure scores over the first 16 weeks between emollient types (global p = 0.765): adjusted Patient-Oriented Eczema Measure pairwise differences – cream–lotion 0.42 (95% confidence interval –0.48 to 1.32), gel–lotion 0.17 (95% confidence interval –0.75 to 1.09), ointment–lotion –0.01 (95% confidence interval –0.93 to 0.91), gel–cream –0.25 (95% confidence interval –1.15 to 0.65), ointment–cream –0.43 (95% confidence interval –1.34 to 0.48) and ointment–gel –0.18 (95% confidence interval –1.11 to 0.75). There was no effect modification by parent expectation, age, disease severity or the application of UK diagnostic criteria, and no differences between groups in any of the secondary outcomes. Median weekly use of allocated emollient, non-allocated emollient and topical corticosteroids was similar across groups. Overall satisfaction was highest for lotions and gels. There was no difference in the number of adverse reactions and there were no significant adverse events. In the nested qualitative study (n = 44 parents, n = 25 children), opinions about the acceptability of creams and ointments varied most, yet problems with all types were reported. Effectiveness may be favoured over acceptability. Parents preferred pumps and bottles over tubs and reported improved knowledge about, and use of, emollients as a result of taking part in the trial. Limitations Parents and clinicians were unmasked to allocation. The findings may not apply to non-study emollients of the same type or to children from more ethnically diverse backgrounds. Conclusions The four emollient types were equally effective. Satisfaction with the same emollient types varies, with different parents/children favouring different ones. Users need to be able to choose from a range of emollient types to find one that suits them. Future work Future work could focus on how best to support shared decision-making of different emollient types and evaluations of other paraffin-based, non-paraffin and ‘novel’ emollients. Trial registration This trial is registered as ISRCTN84540529 and EudraCT 2017-000688-34. Funding This project was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (HTA 15/130/07) and will be published in full in Health Technology Assessment; Vol. 27, No. 19. See the NIHR Journals Library website for further project information. Plain language summary One in five children in the UK have eczema, a long-term, itchy, dry skin condition. It can significantly affect both the child and their family. Most children are diagnosed and looked after by their family doctor (general practitioner) and are prescribed moisturisers (also called emollients) to relieve skin dryness and other creams (topical corticosteroids) to control flare-ups. However, there are many different types of emollients and, to our knowledge, limited research to show which is better. In the Best Emollients for Eczema clinical trial, we compared the four main types of moisturisers – lotions, creams, gels and ointments. These types vary in their consistency, from thin to thick. We recruited 550 children (most of whom were white and had moderate eczema) and randomly assigned them to use one of the four different types as their main moisturiser for 16 weeks. We found no difference in effectiveness. Parent-reported eczema symptoms, eczema severity and quality of life were the same for all the four types of moisturisers. However, overall satisfaction was highest for lotions and gels. Ointments may need to be used less and cause less stinging. We interviewed 44 parents and 25 children who took part. Opinions of all four types of moisturisers varied. What one family liked about a moisturiser was not necessarily the same for another and preferences were individual to each user. Sometimes there was a tension between how well a moisturiser worked (effectiveness) and how easy it was to use (acceptability). In these cases, effectiveness tended to decide whether or not parents kept using it. People found moisturisers in pumps and bottles easier to use than those in tubs. A number of participants valued the information they were given about how to use moisturisers. Our results suggest that the type of moisturiser matters less than finding one that suits the child and family. Scientific summary Background Eczema (also called atopic eczema/dermatitis) is a common condition that usually first appears in early childhood. It is characterised by dry, itchy skin. Emollients are recommended for all patients, used as a ‘leave-on’ treatment to add and help retain moisture in the skin. For all but the mildest disease, they are used in combination with topical anti-inflammatory (topical corticosteroids or calcineurin inhibitors) to treat and prevent eczema flare-ups. Despite the accepted importance of emollients in treating the dry skin of eczema, there is limited evidence to guide prescribers and users on what types to use. In previous research, parents/carers (hereafter, parents) of children with eczema have spoken of a ‘trial and error’ approach to finding an emollient that suits them and their child, with the attendant frustration, waste and costs to both users and the health service. Objectives We sought to: compare the effectiveness and acceptability of four commonly used types of emollients (lotion, cream, gel and ointment) in the treatment of childhood eczema explore carers’ and children’s experiences of study emollient use and their views about perceived effectiveness and/or acceptability of study emollients. Methods We recruited children with eczema via general practitioners’ (GPs’) surgeries based in three centres (West of England, Wessex and East Midlands). To be eligible, children had to be aged between 6 months and 12 years and to have at least a mild form of disease, and parents had to be willing to be randomly allocated to any of the types as their main emollient. Children with a known sensitivity to study emollients or their constituents were excluded. Participants were randomised to lotion, cream, gel or ointment groups in a 1 : 1 : 1 : 1 manner, stratified by centre and minimised by baseline eczema severity as determined by the Patient Orientated Eczema Measure (POEM) score [mild (3–7 points) vs. moderate/severe (≥ 8 points)] and participant age (< 2 years vs. ≥ 2 years). Participants received their allocated type of emollient via their GP, who prescribed products on their local formulary that were study approved. All study emollients were paraffin based and none contained antimicrobials or urea. Study lotions contained glycerol [Cetraben (Thornton & Ross Ltd, Huddersfield, UK), Diprobase (Bayer UK Ltd, Reading, UK), QV (QV Skincare, Melbourne, VIC, Australia)], study creams had no humectant or lanolin [AproDerm (Fontus Health, Walsall, UK), Aquamax (Intrapharm Laboratories, Maidenhead, UK), Diprobase, Epimax (Aspire Pharma, Petersfield, UK), Zerobase (Thornton & Ross Ltd, Huddersfield, UK)], gels did not contain povidine [AproDerm (Fontus Health, Walsall, UK), Doublebase (Diomed Developments, Hitchin, UK), Isomol (Aspire Pharma, Petersfield, UK), MyriBase (Penlan Healthcare,Weybridge, UK) and Zerodouble (Thornton & Ross Ltd, Huddersfield, UK)] and study ointments had no additives (Diprobase, Emulsifying, Paraffin White soft, Paraffin Yellow soft ointment, White soft/Liquid paraffin 50/50). Participants were asked to use their study emollient as their only leave-on treatment for the first 16 weeks; thereafter, they were free to change. However, if they had problems with or disliked their study emollient, they could stop it and seek an alternative from their GP. Participants’ skin was assessed using the Eczema Area Severity Index (EASI) at baseline and at 16 weeks by a researcher masked to treatment allocation. Other data were collected by self-completed questionnaires weekly (first 16 weeks) and then 4-weekly (until 52 weeks). The primary outcome was eczema symptoms measured using the POEM over 16 weeks. We sought to recruit 520 children to detect a difference of 3.0 points in POEM scores between any two groups with 90% power and a significance level of 0.05, allowing for 20% loss to follow-up and multiple comparison testing. We conducted semistructured interviews with parents and older children at around 4 weeks and 16 weeks after randomisation, sampling on the characteristics of the children. The interviews were conducted face to face and by telephone, informed by a topic guide that was revised during the course of the study. Interviews were audio-recorded and transcribed verbatim. Analysis was thematic and carried out alongside data collection, which stopped once saturation was reached. Ethics approval was granted by the NHS Research Ethics Committee (South West – Central Bristol Research Ethics Committee 17/SW/0089). Results Between January 2018 and October 2019, 78 GP surgeries sent 9437 invitations to potentially eligible children. Expressions of interest were received from 910 parents, and 550 children attended a baseline visit, were eligible and enrolled. At baseline, the characteristics of participants were balanced except for sex, as there were more girls in the cream group (55%) than in the gel group (40%). Most children were white (86.0%), with a median age of 4 years (interquartile range 2–8 years) and moderate severity eczema (mean POEM score 9.32 points; standard deviation 5.46 points). Creams (94.5%) were most likely to have been used before, followed by ointments (66%), lotions (63.0%) and gels (25.0%). Participants were randomised to receive lotion (n = 137), cream (n = 140), gel (n = 135) or ointment (n = 138), with prescriptions issued by their GP. The median number of days between randomisation and reported first use of emollient was 4 days (interquartile range 3–7 days), with 80% reporting first use within 7 days of randomisation. There were 29 withdrawals (24 in the first 16 weeks). A total of 95% of participants provided at least one post-baseline POEM score and, therefore, were included in the primary outcome analysis. The researcher undertaking the EASI assessment identified the correct allocation in seven participants. There was no difference in the primary outcome (repeated-measures analysis of weekly POEM scores over the first 16 weeks) between the different groups (global p = 0.765). The adjusted differences in mean POEM scores [95% confidence interval (CI)] for pairwise comparisons were as follows: lotion versus cream 0.42 (95% CI –0.48 to 1.32; p = 0.360); lotion versus gel 0.17 (95% CI –0.75 to 1.09; p = 0.718); lotion versus ointment –0.01 (95% CI –0.93 to 0.91; p = 0.983); cream versus gel –0.25 (95% CI –1.15 to 0.65; p = 0.586); cream versus ointment –0.43 (95% CI –1.34 to 0.48; p = 0.354); and gel versus ointment –0.18 (95% CI –1.11 to 0.75; p = 0.704). Adjusting for sex imbalance at baseline and imputing missing data using multiple imputation did not meaningfully alter these results. There was no evidence of a difference in mean POEM scores between treatment groups or in any of the pairwise comparisons in first 16 weeks in ‘per-protocol’ analysis (p = 0.238) or over the 52 weeks (p = 0.909). There was no evidence of effect modification in prespecified subgroup analyses by parent prior emollient expectation (p = 0.935), participant age (p = 0.343), participant eczema severity (p = 0.042), or UK diagnostic criteria for atopic dermatitis (p = 0.291). No differences between groups were seen in the following secondary outcomes at 16 weeks (or 52 weeks, if also collected): EASI scores, Atopic Dermatitis Quality of Life scores, Child Health Utility 9-Dimension scores, Dermatitis Family Impact scores and well-controlled weeks. During the first 16 weeks, median reported weekly use of the allocated emollient appeared to be higher in the lotion and cream groups (6 days per week) and lower in the gel (5 days) and ointment (3 days) groups, but this difference in usage was not statistically significant (p = 0.481). Similarly, there was no difference between groups in median reported daily use of non-allocated emollient or topical corticosteroids. Overall satisfaction was highest with lotions and gels (67.3% and 64.5% very or mostly satisfied, respectively) and dissatisfaction was highest with creams and ointments (34.2% and 40.4% dissatisfied or very dissatisfied, respectively) (p = 0.003). Overall, 37% of participants reported at least one adverse reaction, the most common being ‘application site reactions’ (e.g. worsening of eczema). There was no evidence that the proportion of children reporting adverse reactions in the first 16 weeks of follow-up differed by treatment group (p = 0.794). There were no significant adverse events. In the nested qualitative study, 44 parents were interviewed, 20 at both weeks 4 and 24 (including five repeat interviews at week 16). Children took part in 25 interviews. Participants judged the effectiveness of study emollients by comparing them with others they had used in the past (i.e. the perceived hydrating action of the emollient, skin feel after application, skin symptoms and appearance, the number of flare-ups, and the need for topical corticosteroid use). Other factors identified as affecting effectiveness were weather and the frequency and quantity of emollient application. Acceptability was usually considered alongside or as part of the effectiveness of an emollient. Characteristics of the emollients that participants considered were how it felt on the skin, ease of application and absorbency, with smell being less important. In terms of containers, participants favoured pumps and squeezer bottles for practical reasons, including older child being able to self-apply. Many participants in the lotion and gel groups reported ease of application but felt that these types of emollients had to be used more often, and there was a perception that they ‘maintained’ rather than improved the skin. Although participants using study creams and study ointment were more likely to report improvements, opinions about their acceptability were more divergent. Problems were reported with all types of emollients. At 16 weeks, there was no clear pattern or differentiation between the emollient types in terms of continued use over and above the factors listed above. Parents of children with very dry and/or rough skin tended to prefer an emollient with a thicker consistency, such as a cream or an ointment. Age was also reported to influence emollient use: application may be easier/more frequent in younger children, accompanying nappy changes, and more difficult in older children as they become more independent and attend school. Some participants thought that their child’s eczema and the effectiveness of their emollient was related to their ethnicity. Changes in behaviour and knowledge were reported as a result of taking part in the trial. For some, the regular study questionnaires had reminded them to apply their emollient regularly. Others reported persisting more with one treatment or that the emollient information sheet had improved their knowledge of eczema management. Conclusions No one type of emollient was found to be superior, although these findings may not apply to children from more ethnically diverse backgrounds. Parents and children should be made aware that application site reactions are common; persistence may be required to find an emollient that works for them; and preferences may change with time, season and body site. We have also demonstrated the need for choice and education around the use of emollients for the treatment of eczema in children. Without this, emollients may be applied incorrectly or not as frequently as prescribed. Most children with eczema will need, in addition to their emollient, a topical anti-inflammatory treatment (usually corticosteroids) appropriate to the site and severity of their eczema to get, and keep, control of their condition. Guidelines should advocate for, and formularies support, a range of emollients, with lotions, creams, gels and ointments all available. Prescribers and pharmacists have an important role in ensuring that families are aware of the different emollients available, to support them in selecting a type most likely to suit them and to advise on optimal use. Verbal advice could be accompanied by written or online information (including videos) on the role of emollients and how to use them, perhaps accompanied by a planned review at the end of an agreed trial period. Future research could evaluate a decision aid to support parents and clinicians in deciding which type of emollient to try first and the clinical effectiveness and cost-effectiveness of providing ‘tester pots’ of each type of emollient to try before selecting their preferred emollient. Further trials may be appropriate to compare emollients in more ethnically diverse populations and of different types not evaluated in this study, for example ointments with emulsifiers and humectant-containing, plant-based and ‘novel’ emollients, including those designed to alter the skin microbiome. Research in this field would benefit from an internationally agreed system of classifying different emollients and a common approach to measuring and reporting treatment use. Finally, further research is needed to determine how emollients best fit into an overall package of eczema care, which includes frequency of use, bathing, use of other topical treatments and avoidance of triggers. Trial registration This trial is registered as ISRCTN84540529 and EudraCT 2017-000688-34. Funding This project was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (HTA 15/130/07) and will be published in full in Health Technology Assessment; Vol. 27, No. 19. See the NIHR Journals Library website for further project information.