Article Text
PDF
Abstract
Objective To assess the effectiveness of a potent topical corticosteroid (TCS) as an initial treatment in primary care for children with moderate flare-ups of atopic dermatitis (AD), compared to starting on a mild TCS.
Design An observational prospective cohort study with an embedded pragmatic multicentre open-label randomised controlled trial.
Setting A total of 53 general practices in the southwest of the Netherlands took part in the study.
Participants 209 children aged 3 months to 17 years diagnosed with AD (International Classification of Primary Care codes S87 or S88) who visited their general practitioner (GP) for AD or received repeat prescriptions for AD in the previous 12 months were included in the cohort study through the general practices. Finally, 32 patients (15%) were randomised and assigned to the trial (13 girls; 19 boys; median age 4.0 years).
Interventions If cohort participants experienced a moderate flare-up (ie, need to intensify topical treatment from the child’s and/or parents’ point of view of AD and a three-item severity score from three to<6 scored by their GP) during cohort follow-up, they were randomised to either the intervention group, a strong TCS (class III, fluticasone propionate 0.05%), or the control group, a mild TCS (class I, hydrocortisone acetate 1%).
Primary and secondary outcome measures We measured outcomes at baseline and at 1, 4 and 24 weeks. The primary outcome was AD-related symptoms (Patient-Oriented Eczema Measure (POEM) score) measured over 24 weeks of follow-up. Secondary outcomes included the Eczema Area and Severity Index, the Investigators Global Assessment, quality of life (QoL), Patient Global Assessment, Numeric Itch Intensity Score and TCS use.
Results The primary outcome showed a significant difference in the POEM scores over 24 weeks of follow-up between the intervention group (n=17) and the control group (n=15) (3.3 vs 9.4, p=0.023). The potent TCS also significantly improved the POEM at 1 week (5.5 vs 12.0, p=0.042) and 4 weeks (4.3 vs 12.7, p=0.030). Improvement in the QoL was significant at 4 weeks (1.0 vs 4.5, p=0.014) and 24 weeks (0.0 vs 2.0, p=<0.000).
Conclusion Despite the small sample size, the data suggests a clinical benefit from starting with a potent TCS rather than a mild TCS when a flare-up of AD is moderate.
Trial registration The Netherlands National Trial Register: NTR6679.
- Randomised Controlled Trial
- Eczema
- Primary Care
Data availability statement
Data are available upon reasonable request.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
STRENGTHS AND LIMITATIONS OF THIS STUDY
A randomised trial in primary care.
Long follow-up evaluating the effectiveness of a potent TCS versus a mild TCS for moderate AD in children.
Small sample size.
Open label.
Introduction
Atopic dermatitis (AD) is a chronic inflammatory skin disease characterised by itching and repeated skin lesions.1 AD is the skin condition with the largest overall disease burden, with a prevalence of up to 20% in children and 10% in adults.2 AD is in the top 10 most prevalent disorders in general practice in children aged up to 18.3 Fewer than 6% of all AD patients are referred to secondary care, so the majority of patients are managed by their own GPs with a combination of emollients and topical corticosteroids (TCSs).4 5
The guidelines for managing AD that are now available internationally range greatly in quality and offer a wide variety of therapeutic suggestions.6 Because of the lack of evidence, recommendations are frequently based solely on expert opinions. Various initial therapeutic strategies are advocated for disease flare-ups, such as starting with a short burst of high-potency TCS, starting with the least potent TCS possible and starting with a TCS tailored to the severity. The Dutch guideline on AD for GP advocates a stepwise approach, starting with emollients if the AD is mild and starting with a mild TCS if the AD is moderate.7 In the Netherlands, it has been determined that most of the affected children in the general population have moderate AD, as determined by the patient-reported outcome measure (Patient-Oriented Eczema Measure, POEM), and use a moderate TCS followed by a potent TCS.8 Particularly in moderate or severe AD, a potent TCS is probably more effective than a mild TCS, and we hypothesise that a potent TCS will result in a larger number of patients whose treatment is successful.9 Furthermore, treatment with a potent TCS compared with a mild TCS might lead to faster and better results, ultimately with less overall use of TCS, improved quality of life (QoL) and better compliance.10 However, trials comparing different strengths of TCS are often more than 20 years old, inferior in quality and only with short follow-ups (14–42 days).11 12 In addition, the majority of these trials were conducted in outpatient or other hospital settings, whereas most patients with AD are treated in primary care in many countries.3 13 Commencing treatment with a potent TCS may have the additional benefit of preventing a flare-up from becoming chronic skin damage. This in turn avoids the skin being exposed to allergens.
POEM has been chosen as the recommended patient-reported outcome measure for symptoms of AD internationally.14 15 As far as we are aware, no studies have evaluated the effectiveness of treatment using the recommended POEM measurement. Additionally, no studies have assessed the effectiveness of treatment with mild TCS versus a potent TCS in children with moderate AD in primary care. We therefore conducted the Rotterdam Eczema Study to assess the effectiveness of a potent TCS as an initial treatment in primary care for children with moderate AD flare-ups, compared with starting with a mild TCS. The primary objective was to investigate whether starting with a potent TCS is superior to starting with a mild TCS for long-term disease control (follow-up for 6 months) when measured by the POEM.16
Methods
Study design
The Rotterdam Eczema Study was an observational prospective cohort study with an embedded pragmatic multicentre open-label randomised controlled trial. We chose this design as it can be difficult to randomise children who are experiencing a flare-up when they visit their GP. Our approach allowed us to monitor the AD-affected children who were already enrolled in the cohort and incorporate them into the study immediately in the event of a flare-up. Furthermore, the overall aim of the cohort was to determine the frequency, burden and determinants of flare-ups of AD during follow-up. A detailed study protocol has been published.16 The Medical Ethics Committee (MEC) of the Erasmus Medical Centre Rotterdam approved the protocol (MEC-2017–328).
Patients
We included patients through GP practices. The GP confirmed their willingness to take part in the research and was responsible for the patient’s treatment and management. The inclusion criteria for patients in the cohort were age between 12 weeks and 18 years, diagnosis of eczema (International Classification of Primary Care codes S87 and S88), additional confirmation of the diagnosis by the GP, a consultation or repeat prescription in the previous 12 months and informed consent. We excluded patients from the cohort if their own GP was aware of potential problems that would limit participation in a trial, such as family problems, current treatment for AD in secondary care, contraindications for the study medication, language barriers or no internet access (which was needed to fill-in the weekly online questionnaires). If the child was willing to participate, the informed consent was signed by the child (if aged ≥12 years) and/or the parents (when the child was aged ≤16). Inclusion criteria for the trial were participation in the cohort through their GP, a flare-up (ie, the need to intensify topical treatment) from the child’s and/or parent’s point of view and a Three-Item Severity (TIS) score between 3 and 5. The TIS score is used in the Dutch GP guideline on AD to determine disease severity (0–2 mild, 3–5 moderate, 6–9 severe). Patients were excluded from the trial if they had used a TCS in the 2 weeks before inclusion in the trial, had AD on their eyelids, had >50% of the body affected by AD, had other skin disorders hampering proper assessment of AD, were pregnant or breastfeeding or had clinical signs and symptoms of untreated skin infections (bacterial, viral, fungal or parasitic).
Interventions
The intervention group received a potent TCS (class III, fluticasone propionate 0.05%) once daily to start with at each flare-up during the trial follow-up period. If children were aged <2 years, they were reassessed by the GP after 1–2 weeks. For children over the age of 2, the follow-up was left to the GP’s discretion, and they may have chosen to schedule a reassessment or not. The control group was instructed to receive care as stated in the Dutch GP guideline for all flare-ups during the trial period.7 This consisted of starting with a mild TCS (class I, hydrocortisone acetate 1%) once daily. If the condition did not improve within 1–2 weeks, a moderate-potency class II TCS was prescribed once daily. If the moderate (class II) TCS did not improve the symptoms within 1–2 weeks, a potent TCS (class III) was prescribed once daily. The TCS formula was selected by the GP based on the Dutch GP guideline for AD. The guideline suggests using ointment or cream depending on whether the eczema is dry or wet. In addition, the instructions for using the TCS are in line with the fingertip unit as specified in the Dutch GP guideline. When symptoms improved, the children followed a predefined tapering plan.7 In addition to using a TCS, all children were advised to use an emollient (namely cetomacrogol). The advice was to use the emollient daily.
Randomisation
We used our electronic data capture system to prepare a computerised generalised randomisation list using 1:1 allocation with randomly permuted blocks. Children were stratified by TIS score (ie, TIS score between 3 and 5) to ensure equal distribution of AD severities between the two groups. After the inclusion and exclusion criteria were checked and baseline assessments were finished, the physician assistant performed the randomisation on the encrypted website and then informed the patient and GP of the randomisation result.
Primary outcome
Patients received a weekly online questionnaire for 24 weeks. The primary outcome was the average subjective disease severity over 24 weeks of follow-up in the trial, measured by the POEM.
Secondary outcomes
We also measured outcomes at baseline and 1, 4 and 24 weeks after randomisation. Secondary outcomes included the objective disease severity at 1, 4 and 24 weeks, as measured with (1) the Eczema Area and Severity Index (EASI), (2) the Investigators’ Global Assessment (IGA) and (3) QoL using the Infants’ Dermatitis Quality of Life Index or Children’s Dermatology Life Quality Index (depending on age). The secondary outcomes measured weekly in the online questionnaire were (4) the Patient Global Assessment (‘How is your AD today?’ on a 6-point scale: clear, almost clear, mild, moderate, severe, very severe), (5) the Numeric Itch Intensity Score (‘Indicate how bad the itching was that you experienced during the past day’; ranging from 0=no itching to 10=worst itching imaginable), (6) frequency and type of TCS application (for description of mild, moderate and potent TCS see online supplemental file 1 and (7) side effects. For the weekly questionnaire, see online supplemental file 2.
Supplemental material
Supplemental material
Sample size
This study was powered to test whether a potent TCS (class III) was superior to a mild-potency TCS (class I) as measured by the mean POEM score over 24 weeks of follow-up. We used data from a feasibility study by Ridd et al with a mean score of 8.8 (SD 5.9) in a similar population (ie, primary care patients) in terms of the baseline POEM characteristics.17 The minimum clinically important difference (MCID) for the POEM score was estimated to be 3.18 A sample size of 122 patients (61 per group) was needed to achieve 80% power at a significance level of α=0.05 to confirm or disprove superiority. We expected a loss during follow-up of 15% and therefore aimed to recruit 144 patients.
Statistical analysis
We described the baseline characteristics using the medians or the numbers, as appropriate. Because the data was not normally distributed—probably due to the small sample size—we used a different analysis to the one stated in the protocol. We used the Quade non parametric ANCOVA (analysis of covariance) for all outcomes, with age, sex, baseline POEM and baseline EASI as the covariates. The analysis for the primary outcome was done according to the intention-to-treat principle, that is, irrespective of compliance. Secondly, a per-protocol analysis was carried out, excluding children for whom there had been major deviations from the protocol. Major protocol deviations were defined as withdrawn from the study or lost to follow-up, or a medication compliance rate of <75% (non-compliance defined as a POEM >8 and no use of TCS during a specific week). Secondary outcomes of the trial, including statistical comparisons between the treatment groups for changes in disease severity and QoL after 1 week and over 4 weeks, were also carried out using the Quade non parametric ANCOVA. We tested for the possible confounders of age, sex and baseline POEM score. Because the data was not normally distributed, we imputed POEM scores for missing weeks (14.4%) using the mean per patient of the non-missing weeks and did a sensitivity analysis with missing values=0 and last observation carried forward for the primary outcome. All analyses were performed using Statistical Package for the Social Sciences (SPSS), V.28.0 (IBM Corp).
Patient and public involvement
The outcome measures are based on the recommendations stated by the HOME initiative (Harmonising Outcome Measures for Eczema).14 In this initiative, patients were intensively involved in the development of the core outcome set and its measurement tools.
Results
Patients
Between January 2018 and September 2020, 209 patients were included in the cohort through 53 general practices. Ultimately, 32 patients (15%) were randomly assigned to a trial group (13 girls; 19 boys; median age 4.0 years); 17 were allocated to the potent TCS group and 15 to the mild TCS group (figure 1). Unfortunately, the calculated minimum sample size of 144 was not reached. One patient in the control group had a medication compliance rate of <75%; 31 patients were therefore included in the per-protocol analysis, whereas all 32 were included in the intention-to-treat analysis. The baseline characteristics are presented in table 1. The POEM scores differed between the two groups.
Baseline characteristics of participants
Flowchart showing the recruitment and randomisation of patients.
Primary outcome
In both groups, the POEM score improved over the 24 weeks of follow-up. The primary outcome, the mean POEM score over 24 weeks, was almost three times lower in the intervention group (3.3 vs 9.4, p=0.023). The mean improvements in the intervention group versus the control group were greater than the minimal clinically significant difference of 3.0 at 1 week (5.5 vs 12.0, p=0.042), 4 weeks (4.3 vs 12.7, p=0.030) and 24 weeks (3.3 vs 9.4, p=0.023). The differences between the groups were significant at all time points (table 2). These results were comparable to the sensitivity analysis in the intention-to-treat population for follow-up at 1 week and over 4 weeks (see online supplemental file 3) (table 2).
Supplemental material
POEM scores for primary outcome (mean over 24 weeks) and secondary outcomes (1 and 4 weeks), intention-to-treat analysis
Secondary outcomes
Improvement in the QoL was significant at 4 weeks (1.0 vs 4.5, p=0.014) and 24 weeks (0.0 vs 2.0, p=<0.001). The median TCS use at 1-week follow-up was 7 days. For the 4-week and 24-week follow-up, this decreased for both groups without there being a significant difference between the two groups. The IGA scores did not improve in the intervention group, in which two patients reported a side effect during the study, namely stinging skin when applying the TCS and telangiectasias (small, dilated blood vessels on the skin) (table 3).
Secondary outcomes, median
Per-protocol analysis
In the per-protocol analysis, 31 patients were included. The POEM score improved in both groups at all time points. The primary outcome, the mean POEM score over 24 weeks, was almost three times lower in the intervention group and differed significantly (3.3 vs 9.2, p=0.025). The inter-group difference was significant over 4 weeks (p=0.032) of follow-up (table 4).
POEM scores for primary outcome (mean over 24 weeks) and secondary outcomes (1 and 4 weeks), per-protocol analysis (n=31)
Discussion
Summary
Although the number recruited was less than intended, this trial still suggests evidence for the effectiveness in primary care of using a potent TCS rather than a mild TCS for children presenting with a moderate flare-up of AD. During both long-term (24 weeks) and short-term (1 and 4 weeks) follow-ups, there was a statistically significant improvement in the patient-reported outcome measure POEM, with fewer symptoms for children on a potent TCS compared with those using a mild-potency TCS for flare-ups of moderate AD in primary care. For both the short-term (4 weeks) and long-term (24 weeks) QoL, a statistically significant difference in favour of a potent TCS was found.
Strengths and limitations
As far as we are aware, this is the first trial in primary care with a long follow-up evaluating the effectiveness of a potent TCS versus a mild TCS for moderate AD in children. The original intention was to include 144 children for the trial rather than the 32 who were actually enrolled. The failure to enrol enough patients has been described and analysed in our article.19 The main reason why cohort participants did not take part in the trial was that they did not contact their GP when experiencing a flare-up of AD. Furthermore, the majority of patients who contacted their GP did not match the trial’s inclusion criteria. Although the number of 32 children included was fewer than intended, we did find clinically relevant and significant differences for the primary outcome in this randomised trial. On the other hand, the small sample size could have led to sampling bias, and in particular self-selection bias, because patients had to contact the GP themselves before they could participate in the trial. This may affect the generalisability. A limitation of an open-label trial is that a patient’s or parent’s knowledge of the treatment received could influence their views and the reporting of their symptoms, creating a potential bias that favours the intervention. This performance bias could also apply to our study. Because the GPs had to be able to adjust the treatment strategy or refer the child if required, it was not possible to have them blinded. Furthermore, the high costs associated with blinding meant it was unfortunately not possible to blind the participants to their medication in this pragmatic study. Another limitation is that multiple testing was used for the secondary outcomes and, therefore, the results (the beneficial effect of a potent TCS on QoL) can at best be used for generating hypotheses. A statistical limitation is that we summarised repeated measures over 24 weeks as an average.
Comparison with existing literature
The MCID we used for the sample size was calculated by Gaunt et al in 2016. In 2018, Howells et al presented interpretations of changes in the POEM, where a change of ≥4 was very likely to be a clinically important change.20 For our primary outcome, the mean POEM over 24 weeks, we found a difference between the two groups of 6.1 POEM points, which indicates a clinically relevant difference. However, we calculated a mean change over all subjects and this will therefore not apply to all individuals. We did choose POEM as the primary outcome. As proposed by HOME in 2011, long-term control was identified as a key outcome domain.21 In 2021, an instrument was selected by HOME to assess long-term control.22 As our study started in 2018, we opted for a follow-up period of 24 weeks to evaluate long-term control. Although the EASI is a more objective measure, it was not feasible to capture the EASI score on a weekly basis over 24 weeks. Furthermore, in mild eczema, EASI is not responsive enough. In our study, the median EASI at baseline was 2.8. The MCID for EASI is 6.6.23 This indicates that the EASI cannot be very responsive for our study population. The evidence we found for the effectiveness of a potent TCS matches the results of a Cochrane review by Lax et al about the effectiveness and safety of different ways of using TCS for treating AD.9 This Cochrane review stated that trials typically lasted 1–5 weeks, and included adults and children with moderate or severe AD, and that most trials were conducted in outpatient or other hospital settings. One study by Thomas et al is comparable to ours: children with mild or moderate AD (clinical severity of AD) were randomly assigned either to treatment with a potent TCS (0.1% betamethasone valerate) applied for 3 days followed by the base ointment for 4 days or to treatment with a mild TCS (1% hydrocortisone) applied for 7 days. No differences were found between the two groups for all outcomes, and they concluded that a short burst of a potent TCS is just as effective as prolonged use of a mild TCS in children with mild or moderate AD. We included only children with moderate AD, which may explain the differences in outcome compared to Thomas et al as most of the children in their trial had mild AD. The short burst (only 3 days) of a potent TCS could also have led to there being no differences between the two groups. In our trial, the median TCS use was 7 days in both groups after 1 week of follow-up. For faster and better disease control, longer application of a potent TCS may be needed during a flare-up. The Cochrane review by Lax et al also studied the safety of different ways of using the TCS. They found only 26 cases of abnormal skin thinning in 2266 participants (1% across 22 trials). Most cases were in participants using higher-potency TCSs (16 with very potent and six with potent TCSs). The evidence was assessed as having very low certainty for very potent versus potent TCS. In our trial, one patient reported telangiectasia as a side effect. However, the telangiectasia was not confirmed by a researcher or physician, and the side effects need to be studied in a larger sample to provide better evidence. Although the differences in objective outcomes (IGA, EASI) were not statistically significant, the EASI score was lower in the intervention group at all time points, which is possibly an indication of a smaller AD-affected area. Although a potent TCS leads to higher exposure, a possibly smaller affected area plus a lower frequency of TCS use in the intervention group may reduce the overall exposure to TCS. The IGA did not differ; this is likely to be due to the relatively insensitive 5-point scale; this is less responsive to changes than the EASI scale of 72 points.
Implications for research and/or practice
Despite the small sample size and the limitations, the findings of this trial suggest benefits (in symptoms/POEM and QoL) in general practice from starting children with moderate AD on a potent TCS. The clinically important improvements in disease severity and QoL compared with the baseline were significant. The difference was twice the MCID of the POEM for the primary outcome, which suggests clinical relevance. Our findings may help in providing clear recommendations in guidelines for the management of AD in children in primary care. Future studies with larger samples and blinding of the participants and GPs may confirm these results and potentially show significant findings for our secondary outcomes.
Data availability statement
Data are available upon reasonable request.
Ethics statements
Patient consent for publication
Ethics approval
This study involves human participants and was approved by The Medical Ethics Committee (MEC) of the Erasmus Medical Center Rotterdam (MEC-2017-328). Participants gave informed consent to participate in the study before taking part.
Acknowledgments
We would like to thank all participants and their parents for their support.
Footnotes
Contributors KFvH, GE and AMB conceived the study and initial study design in collaboration with PJEB and SGMAP. KFvH and AMB conducted the analyses. All the authors contributed to drafting this paper, led by KFvH, and approved the final manuscript. The corresponding author GE attests that all the listed authors meet authorship criteria and that no others meeting the criteria have been omitted. KFvH is the guarantor.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Patient and public involvement Patients and/or the public were involved in the design, conduct, reporting or dissemination plans of this research. Refer to the Methods section for further details.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.