Online behavioural interventions for children and young people with eczema: a quantitative evaluation

Background Two online behavioural interventions (one website for parents/carers of children with eczema; and one for young people with eczema) have been shown in randomised controlled trials to facilitate a sustained improvement in eczema severity. Aim To describe intervention use and examine potential mediators of intervention outcomes and contextual factors that may influence intervention delivery and outcomes. Design and setting Quantitative process evaluation in UK primary care. Method Parents/carers and young people were recruited through primary care. Intervention use was recorded and summarised descriptively. Logistic regression explored sociodemographic and other factors associated with intervention engagement. Mediation analysis investigated whether patient enablement (ability to understand and cope with health issues), treatment use, and barriers to adherence were mediators of intervention effect. Subgroup analysis compared intervention effects among pre-specified participant subsets. Results A total of 340 parents/carers and 337 young people were recruited. Most parents/carers (87%, n = 148/171) and young people (91%, n = 153/168) in the intervention group viewed the core introduction by 24 weeks. At 24 weeks, users had spent approximately 20 minutes on average on the interventions. Among parents/carers, greater intervention engagement was associated with higher education levels, uncertainty about carrying out treatments, and doubts about treatment efficacy at baseline. Among young people, higher intervention use was associated with higher baseline eczema severity. Patient enablement (the ability to understand and cope with health issues) accounted for approximately 30% of the intervention effect among parents/carers and 50% among young people. Conclusion This study demonstrated that positive intervention outcomes depended on a modest time commitment from users. This provides further support that the wider implementation of Eczema Care Online is justified.


Aim
To describe intervention use and examine potential mediators of intervention outcomes and contextual factors that may influence intervention delivery and outcomes.

Design and setting
Quantitative process evaluation in UK primary care.

Method
Parents/carers and young people were recruited through primary care.Intervention use was recorded and summarised descriptively.Logistic regression explored sociodemographic and other factors associated with intervention engagement.Mediation analysis investigated whether patient enablement (ability to understand and cope with health issues), treatment use, and barriers to adherence were mediators of intervention effect.Subgroup analysis compared intervention effects among pre-specified participant subsets.

Results
A total of 340 parents/carers and 337 young people were recruited.Most parents/carers (87%, n = 148/171) and young people (91%, n = 153/168) in the intervention group viewed the core introduction by 24 weeks.At 24 weeks, users had spent approximately 20 minutes on average on the interventions.Among parents/ carers, greater intervention engagement was associated with higher education levels, uncertainty about carrying out treatments, and doubts about treatment efficacy at baseline.Among young people, higher intervention use was associated with higher baseline eczema severity.Patient enablement (the ability to understand and cope with health issues) accounted for approximately 30% of the intervention effect among parents/carers and 50% among young people.

Introduction
Eczema (also known as atopic eczema/ dermatitis) is a common long-term skin condition characterised by itchy, dry, and inflamed skin.It affects around 20% of children in the UK and often persists into adulthood. 1Eczema can have substantial impact on quality of life for those affected and their families. 2,3First-line eczema treatment includes emollients and topical corticosteroids alongside the avoidance of irritants/triggers (such as soap). 4Topical treatments are often underused because of uncertainty and hesitancy, and irritants and triggers are often not well understood. 5,6While there is evidence to suggest that intensive group education delivered by a multidisciplinary team can improve outcomes in eczema, 7 less is known about the effectiveness of online behavioural interventions. 8,9e authors developed two online behavioural interventions (called Eczema Care Online): one for parents/ carers of children with eczema; and one for young people with eczema. 10,11hese interventions were evaluated in two randomised controlled trials (RCTs), which demonstrated that the interventions provided a useful, sustained improvement in the eczema severity symptoms for up to 52 weeks in both children and young people, when offered in addition to usual care.The mean difference in Patient-Oriented Eczema Measure (POEM) score was −1.5 (95% confidence interval [CI] = −2.5 to −0.6; P = 0.002) in the parents/carers trial, and −1.9 (95% CI = −3.0 to −0.8; P<0.001) in the young people trial. 12hese RCTs also explored the impact of the interventions on two hypothesised behavioural mechanisms: emollient use and topical corticosteroid/topical calcineurin inhibitor use; as well as two psychological mechanisms: patient enablement (the self-perceived ability to understand and cope with health issues)

RESEARCH | e380
and perceived barriers to treatment.There were no significant differences between groups found for either RCT in self-reported treatment use at 24 weeks or perceived barriers to treatments at 24 weeks. 12However, improvements in patient enablement were found in the intervention groups in both trials: adjusted mean difference at 24 weeks was −0.7 (95% CI = −1.0 to −0.4) for parents/carers and −0.9 (95% CI = −1.3 to −0.6) for young people. 12qualitative process evaluation provided further support for the role of patient enablement. 13Specifically, parents/carers and young people reported that the interventions supported them to feel confident in managing eczema and discussing treatments with healthcare professionals, to normalise and accept eczema, and (for parents) to involve their child in eczema management.
In line with the Medical Research Council guidelines for developing and evaluating complex interventions, 14,15 the current study involved a quantitative process evaluation to further explore how the interventions worked and for whom.These guidelines recommend exploring: 1. implementation: the extent to which the intervention is used as intended; 2. mechanisms: the processes by which an intervention leads to changes in the intended outcome; and

Design
This study involved a quantitative process evaluation embedded within two RCTs (details of which are reported elsewhere). 12,16Trial participants were randomised into either an intervention group who were given access to the relevant online intervention in addition to usual care, or into a usual care group who received usual care and were recommended a standard informational website (https://eczema.org),and given access to the intervention on study completion.

Intervention
The two digital interventions were developed using evidence-, theory-, and person-based approaches. 17,18The aim of both interventions was to reduce eczema severity via several behavioural mechanisms applying to children and young people: increased use of emollients; improved use of topical corticosteroids or topical calcineurin inhibitors; improved management of irritants and triggers; reduced scratching; and improved emotional management.The first two mechanisms were identified as core behaviours likely to have the greatest effect on eczema severity. 10,11Intervention development and the content and design features are summarised in Box 1 and described in detail elsewhere. 10,11See Supplementary Figures S1 and S2 for details of logic models of hypothesised intervention mechanisms.that the focus should be on 'effective engagement', defined as the minimal level of engagement necessary for achieving the intended outcomes of the intervention, rather than broad engagement measures, such as number of logins or time spent on the intervention. 21The effective engagement threshold may involve viewing certain intervention content that is judged to be most likely to lead to behaviour change. 22Following the AMUsED (Analyzing and Measuring Usage and Engagement Data) framework 22 for analysing and measuring engagement data in digital interventions, the authors chose two patterns of intervention use to explore as potential effective engagement thresholds.The first was viewing the core introductory content that contains the key content the authors deemed necessary for behaviour change (minimum engagement threshold).The second was viewing the core introductory content and at least one optional module (higher engagement threshold).

How this fits in
Mediators and outcomes were measured at 24 weeks, as this was the primary outcome timepoint in the RCT.The modified Patient Enablement Instrument (PEI) 23,24 was tailored to be eczema-and parent/young person-specific, and the question was amended so that responders

Box 1. Summary of key intervention content and design features Intervention content
Intervention content encouraged users to engage in five target behaviours: 1. use of emollient; 2. use of topical corticosteroids or topical calcineurin inhibitors during a flare-up; 3. management of irritants and triggers; 4. reduced scratching (children and young people); and 5. emotional management (children and young people).

Key design features
• Websites accessible via a mobile device • At the beginning of the intervention, users first progressed through a brief (9 pages) introductory section containing the key content necessary for facilitating behaviour change • Two core modules about topical treatments: emollients and topical corticosteroids or topical calcineurin inhibitors • Optional modules (14 in young people intervention and 16 in parent/carer intervention) • Videos (four in young people intervention and five in parent/child intervention; lasting approximately 2 minutes) briefly summarising key behavioural messages • A '2-week challenge' that supported people to get into a routine of applying emollients consistently • A brief eczema assessment that provided tailored advice on which treatment modules (emollients or flare control creams) would be most helpful • Quotes from other parents/carers and young people with eczema that share their experiences of eczema and management advice • Monthly automated email or SMS with additional information and advice for 6 months The two tested websites have since been combined into one (https://www.eczemacareonline.org.uk).

RESEARCH | e382
Research indicated how they felt 'as a result of the eczema care and support you have received in the past 6 months'.

Data analysis
Intervention use at 24 weeks (primary outcome point) and 52 weeks (entire study period) was summarised descriptively.Eczema severity (POEM) was analysed at 24 weeks, to enable mediation and adherence analyses to be carried out, as these cannot be implemented using repeated measures models as were used in the main RCT. 12 Mediation analysis determined whether patient enablement, treatment use, or barriers to adherence mediated the effect of the intervention on eczema severity.A structural equation model was fitted with the PEI (patient enablement) at 24 weeks as a mediator and POEM score at 24 weeks as the outcome (Figure 1).Because of potential confounding between the mediator and outcome, baseline POEM score and baseline PEI score were also adjusted for in the mediation model (Figure 1).This analysis was repeated using total treatment use and total Problematic Experiences of Therapy Scale (PETS) 25 score as mediators, adjusting for baseline POEM and baseline treatment use/baseline PETS scores, respectively.Total treatment use was calculated by combining weekly emollient use (total number of times per week), weekly topical corticosteroid use, and weekly topical calcineurin inhibitor use.
Subgroup analysis explored whether the intervention effect was different among people with pre-specified categories of baseline variables: sex, age group, deprivation, baseline eczema severity, emollient use (days per week), and barriers to adherence.The mean difference in POEM score at 24 weeks in each subgroup was reported, along with the interaction term (difference in treatment effect between subgroups) adjusted for baseline POEM score, recruitment region, ethnicity, previous belief in intervention, use of other eczema websites, and highest parental qualification (parent/carer group only).Associations between high intervention use and various demographic and baseline characteristics, including age, sex, education, deprivation, baseline severity, baseline treatment use, and baseline barriers to adherence, were explored using logistic regression.
As the intervention effect at 24 weeks was similar in both the parent/carer and young people trials, sensitivity analyses pooling these participants were also carried out.All analyses were undertaken on complete cases using Stata (version 17), with statistical significance taken as P<0.05.

Results
Participants A total of 314/340 (92%) parents/ carers and 304/337 (90%) young people completed 24-week POEM (primary timepoint).Overall, 171 parents/ carers and 168 young people were allocated to the intervention group (see Supplementary Table S1 for details of baseline participant characteristics).

Implementation
Table 2 presents summaries of general intervention use for each intervention group across 24 and 52 weeks (see Supplementary Table S2 for details of key intervention component use).Most intervention participants viewed the core introductory content (containing the key content deemed necessary for behaviour change) by 24 weeks (87%, n = 148/171 in the parent/carer trial and 91%, n = 153/168 in the young people trial) and by 52 weeks (88%, n = 151/171 in the parent/carer trial and 93%, n = 156/168 in the young people trial).About half of participants viewed the core introductory content and at least one optional module (high users) by 24 weeks (58%, n = 99/171 in the parent/carer trial and 57%, n = 96/168 in the young people trial) and by 52 weeks (61%, n = 104/171 in the parent/carer trial and 59%, n = 99/168 in the young people trial).

Mechanisms
For parents/carers, patient enablement had a statistically significant mediating effect of the intervention on the child's POEM score (-0.6, 95% CI = -1.0 to -0.2).As a proportion of the total effect, this corresponds to about 30% of the intervention effect on the POEM score at 24 weeks (mediating effect/ total effect = -0.6/-1.9)(Table 3).For young people, patient enablement also had a statistically significant mediating  effect (-1.2, 95% CI = -1.9 to -0.5).This corresponds to about 50% of the intervention effect on the POEM score at 24 weeks (mediating effect/ total effect = -1.2/-2.4).There was no evidence of a mediating effect of total treatment use or perceived barriers to treatment in either trial.

Context
Most subgroup effects were not statistically significant, except for parents/carers with children with severe eczema at baseline, who had a significantly larger treatment effect than those with children with mild eczema at baseline (-4.0 versus 0.8, adjusted interaction term -4.0; 95% CI = -7.7 to -0.2) (see Supplementary Tables S3-S5 for details).This could be because of a floor effect among those with mild eczema.
In the parent/carer trial, having a degree, having uncertainty about how to carry out treatment, and having doubts about treatment efficacy were significantly associated with meeting the higher intervention engagement threshold (Table 4).In the young people trial, higher baseline POEM score (reflecting worse eczema severity) was significantly associated with meeting the higher intervention engagement threshold.

Summary
Most participants in the intervention group met the minimum effective engagement threshold of viewing the core introductory content, suggesting a high level of user engagement.Users spent approximately 20 minutes on average on the interventions, demonstrating that positive outcomes on eczema severity depended on minimal time commitment from users.The study findings suggested that a substantial amount (30%-50%) of intervention effect on eczema severity at 24 weeks was mediated by increasing patient enablement.Among parents/carers, greater intervention engagement across 24 weeks was associated with higher levels of education, uncertainty about how to carry out treatment, and doubts about treatment efficacy at baseline.Among young people, higher intervention use was associated with higher baseline eczema severity.Indirect (mediating) effect -0.2 (-0.9 to 0.5) -0.01 (-0.6 to 0.6) Direct effect -2.0 (-3.5 to -0.5) -1.5 (-2.9 to -0.1)Total effect -2.2 (-3.8 to -0.6) -1.5 (-3.0 to 0.003) PETS score at 24 weeks Indirect (mediating) effect 0.1 (-0.3 to 0.5) 0.1 (-0.1 to 0.4) Direct effect -1.8 (-3.2 to 0.4) -1.9 (-3.1 to -0.7)Total effect -1.7 (-3.1 to 0.2) -1.8 (-3.0 to -0.5)

Research
Most of the associations between user characteristics and eczema severity were not statistically significant, indicating that there is little evidence to support a differential effect between user characteristics.However, this analysis was exploratory in nature, as the trial was not powered to detect differences between subgroups.

Strengths and limitations
By including two different populations, this study was able to explore how implementation, mechanisms, and contextual factors may differ between groups.Trial participants received follow-up questionnaire email and/ or SMS reminders prompting users to revisit the intervention.Time spent on the intervention was based on how long users spent on webpages, which may not be the same as actively engaging with the content.Therefore, the reported use is likely to be inflated.
Use of multiple timepoints allowed the exploration of changes in mediators and outcomes.However, in the RCTs, the primary outcome and potential mediators were measured at the same timepoint.Ideally, the mediators would be measured at an intermediate timepoint when the change is occurring, after the use of the intervention and before the measurement of the outcome at 24 weeks.However, the earliest measurement of the mediators after baseline was at 24 weeks when both mediator and outcome had changed significantly.
In the mediation analysis, total treatment use was included as a potential mediator.The effect for emollient use and topical corticosteroid/topical calcineurin inhibitor use were not explored separately, but there is unlikely to be a mediating effect as they were not statistically significant, and the effect sizes were small.
The study also had a high proportion of females in both groups and excluded those with very mild eczema, which may have limited the generalisability of the findings.

Comparison with existing literature
][28] However, in both RCTs, there were no significant differences between groups in self-reported treatment use (the trials' key hypothesised mechanisms) or perceived barriers to treatments.This finding be a result of the challenges around measuring the complexities of treatment adherence in eczema.Only frequency of treatment use was measured, but participants in a qualitative process evaluation reported additional positive treatment outcomes, such as increasing the quantity (rather than frequency) of emollients used, starting topical corticosteroids more promptly after a flare-up, or reducing their use of topical corticosteroids to prevent overuse. 13articipants in the qualitative study also cited eczema-specific treatment barriers not captured by the PETS, including uncertainty about why treatments are used and the difference between the two

Ethical approval
Ethical approval was granted by South Central -Oxford A Research Ethics Committee (reference number: 19/ SC/0351).

Data
The data that support the findings of this study are available from the corresponding author on reasonable request.

Provenance
Freely submitted; externally peer reviewed.
British Journal of General Practice, June 2024 treatments, and concerns about the longterm safety of treatments.
At baseline, higher eczema severity, higher level of education (among parents/carers), and having doubts and uncertainties about treatment use were significantly associated with higher intervention engagement.One explanation for these relationships is that the beliefs, concerns, and knowledge gaps may have motivated these participants to use the intervention.This is in line with the qualitative findings that suggested that participants who believed they had high levels of eczema knowledge, good eczema control, and a good treatment regimen tended to be less engaged with the intervention. 13

Implications for research and practice
This quantitative process evaluation suggests that the positive outcomes from the associated RCTs depended on only a minimal time commitment from users, providing further support that the wider implementation of Eczema Care Online (https://eczemacareonline.org.uk) is justified.Furthermore, the findings demonstrate that patient enablement is likely to mediate a substantial proportion of the effect of the intervention on eczema severity.However, other mechanisms, such as adherence to treatment regimens, management of irritants/triggers, and treatment concerns, are likely to also play a combined role.Future research should explore how these interventions are used and experienced by a more diverse cohort of people with eczema and their families in a real-world setting, outside a trial context.It would be useful to explore the extent to which users' eczema-specific treatment concerns explain changes in intervention outcomes and associations between time since diagnosis and intervention outcome.

Table 1 . Quantitative measures and timepoints
aAs recruitment start times were staggered, this will be a different 52-week period for each participant.

Table 2 . General website use across 24 and 52 weeks for both intervention groups
Excluding visits to complete research questionnaires and users who visited only once.b Total modules for parents/carers is 19; total modules for young people is 17.IQR = interquartile range.N/A = not applicable. a

Table 4 . Predictors of high intervention use
a Denominator is different.b Total treatment use is weekly combined emollient, topical corticosteroid, and topical calcineurin inhibitor use.PETS = Problematic Experiences of Therapy Scale.POEM = Patient-Oriented Eczema Measure.SD = standard deviation.Bold numbers indicate statistical significance.