The intervention was developed using the Person-Based Approach (PBA) for planning, developing, and evaluating the feasibility of the intervention [22]. The aim of this method is to ground the intervention in the views and experiences of the people who will use it to ensure that it is persuasive, accessible, and engaging for the target population [22]. The PBA involves in-depth qualitative research to identify key objectives and barriers and facilitators to target behaviors [22]. We carried out a systematic review and synthesis of qualitative research to explore the qualitative literature on acne among patients, carers, and health care professionals [23]. The review protocol was registered on PROSPERO (International Prospective Register of Systematic Reviews; registration number CRD42016050525). A secondary analysis of qualitative interviews with people with acne was also carried out to understand young people’s views and experiences with acne and its treatments [24].

Alongside intervention planning, guiding principles were drafted and iteratively developed throughout, identifying distinctive intervention features to address these. This method involved highlighting key objectives from qualitative research (1). to support young people in gaining autonomy and competence around acne management, (2) to support and promote autonomy in making treatment choices, and (3) to provide support and acknowledge the psychological impact of acne (see Table 1 for guiding principles developed for this intervention).

The hypothesized outcome of the intervention was to improve QoL for young people with acne through the target behavior appropriate use of topical treatments. This target behavior was chosen as it has been shown that effective use of topical treatments can improve acne [5] and benefit QoL [6,7]. For addressing this target behavior, barriers and facilitators identified from the qualitative research were described along with the proposed intervention element.

Evidence from the literature and qualitative research (including the systematic review and synthesis of qualitative data [23] and the secondary analysis of interview data with young people [24]) highlighted several barriers to the appropriate use of topical treatments that needed to be addressed in the intervention. These included concerns about side effects, confusion about the different types of topical treatments, beliefs around the ineffectiveness of topical treatments, belief that acne is a short-term condition that will resolve on its own, confusion about how to use treatment, the time-consuming nature of topical treatments, and the belief that oral treatments were more effective than topical treatments.

Alongside the PBA, a behavioral analysis was carried out to map the intervention components to the behavior change taxonomy, which is a list of consensually agreed techniques for specifying interventions [25]. The behavioral analysis showed that the intervention targeted nine behavior change techniques from the 93 behavior change taxonomies [25]. A central behavior change technique was instructions on how to perform the behavior in terms of advice about choosing the right topical treatment and instructions and demonstrations on how to use topical treatments appropriately. The intervention components were also mapped onto the COM-B model, part of the behavior change wheel [26], to map the target constructs and functions for the intervention [26]. This included six target constructs (physical capability, psychological capability, physical opportunity, social opportunity, automatic motivation, and reflective motivation) and five intervention functions (persuasion, education, training, enablement, and modeling).

Qualitative research showed the Extended Common Sense Model of Illness [27] to be a useful model for understanding how people with acne conceptualize illness and treatment and was therefore used in the behavioral analysis to check that all important components of the model were covered in the intervention (Table 2).

The web-based intervention, Spotless, was developed using the LifeGuide software [31]. The intervention was delivered on the web via the internet and included a compulsory core module on topical treatment. This included information about the different types of topical treatments available, how they work, how to use them appropriately, common side effects, and how to manage them. Information was adapted from accurate web-based sources, including National Health Service [32], National Institute for Health and Care Excellence [33], and the British National Formulary [34]. This was initially carried out by artificial intelligence, and the team (MS, AG, PL, and IM) provided suggestions throughout. The purpose of adapting the information was to ensure that it was easily understood by young persons. An example of this was using information about types of treatments, including how they are used and the side effects, but rewriting this in lay language. Six optional modules were highlighted as important for the self-management of acne in earlier qualitative studies (Textbox 1).

As part of the development stage, think-aloud interviews [35] were carried out with 19 participants with acne using the draft intervention to gather feedback and further modify the intervention. Participants were recruited through mail-out from primary care practices, opportunistic sampling using posters, and advertising via social media. The inclusion criteria for the study were young people aged 14-25 years with acne or those who had consulted about their acne or obtained a prescription for their acne in the past year. Potential participants were excluded if they were outside the age range or did not have acne. General practitioners (GPs) were also asked to screen lists to ensure that the invitation pack was not sent to patients where they felt this would be inappropriate. Face-to-face think-aloud interviews were conducted by following a semistructured interview guide to ensure that all topics were covered while also allowing participants to discuss any concerns they had about the intervention. This process involved asking participants to use the intervention while speaking out their thoughts aloud. Interviews were transcribed and analyzed using a deductive approach to code the data using the objectives of the study (engagement, persuasiveness, and usability) and identify positive and negative comments to aid intervention development.

Overall, participants found the intervention engaging, persuasive, and usable, with some suggestions for changes. Main changes made as a result of the interviews were adding pseudonyms and ages to quotes (these quotes were adapted from the qualitative interview study [24] and included to provide other peoples’ experiences in managing acne); changing the context of certain quotes to make them more relatable to the intended user; providing further clarification on how people can manage sun sensitivity as a potential side effect of topical treatments; further clarification on steps for applying topical treatments (time of day and quantity) and what sensitive areas of the face referred to; changing the 6-week challenge to 4 weeks as some participants felt that 6 weeks would be too long to commit and based on evidence that topical treatments could take effect sooner [10]; changing the core module name from universal core treatments to core treatments so that participants would not misinterpret the website as advertising something; and changing the layout of the intervention including the banner, images, and color scheme.

This was a randomized, unblinded feasibility trial comparing two parallel groups: usual care and usual care plus web-based intervention.

The intervention was aimed at young people with acne managed through primary care in the United Kingdom. Participants were recruited through mail-outs from 20 GP practices in the South of England to people aged 14-25 years whose electronic record included a diagnosis of acne and who had received one or more prescriptions for acne in the past 6 months.

People who had previously taken part in the think-aloud study were excluded, as were people who said their acne had cleared and those taking oral isotretinoin, as it is not recommended to use topical acne treatments at the same time as isotretinoin because of the side effects of dry skin.

Patients aged ≥16 years, who met the criteria, were sent an adult study pack from their GP, and patients <16 years received a child study pack (addressed to the parent or carer). Initially, the pack included an information sheet, a freepost envelope, and a covering letter. Those interested returned a reply slip, and a member of the study team contacted the participant, providing them with a unique participant identification number and the link to the web-based intervention. Amendments were made to the process, and these were approved by both the university and National Health Service ethics committees. Changes included an additional A5 flyer about the study to appeal to the target population and a sign-up sheet providing participants with their unique identification number and a link to the intervention. These changes were essential for assessing the feasibility of the study with a challenging population to recruit. Implied parental consent was approved for participants aged <16 years as invitation letters were sent to the parents; therefore, passing log-in details to their child implied consent. This is because young people from 14 years usually self-manage their acne and are responsible for using topical treatments themselves. The link directed all participants to further information and a web-based consent procedure. After consenting, participants were asked to complete a set of baseline questionnaires before being randomized into 1 of 2 groups. Follow-up questionnaires for the trial were conducted at 4 and 6 weeks as a recent study suggested that topical treatments could take effect within 1 to 4 weeks and that continuation after the 4 weeks would lead to further improvements [10]. Participants received an automated email followed by a reminder email a week after (5 and 7 weeks) if they had not completed these. Further text and subsequent phone follow-ups were conducted for nonresponders to complete the outcome measures, particularly the primary outcome Skindex-16.

The usual care group received treatment as usual from their GP, including appointments, prescriptions, and referrals to the dermatologist, if necessary. Participants in this group were given access to the intervention after they had completed the 6-week follow-up questionnaires.

Participants in the usual care plus web-based intervention group received care as usual with immediate access to the website as described to help them self-manage their acne.

We sought to assess a range of feasibility outcomes including the following:

Outcome measures included the following: Skindex-16 [36] was included as a skin-specific QoL measure. Skindex-16 is a validated measure that includes 16 items on a 6-point Likert scale ranging from 0 (never bothered) to 6 (always bothered), which are transformed into a 100-point scale, with higher scores indicating a lower level of QoL [36].

EQ-5D-5L [37] was included as a health-related QoL measure collected at all intervals. It comprised five domains (mobility, self-care, usual activities, pain or discomfort, and anxiety or depression) with five response levels (no problems, slight problems, moderate problems, severe problems, and extreme problems) that describe the current health state. The visual analog scale was also included alongside this [37].

The Problematic Experiences Therapy Scale (PETS) [38] was included and data collected at each interval to explore barriers to treatment adherence. This measure includes 12 items with four subscales: problems due to symptoms, problems due to uncertainty about therapy, problems due to doubts about treatment efficacy, and practical problems. Participant responses were scored on a scale ranging from 1 (disagree strongly) to 5 (agree strongly), with higher scores indicating fewer barriers to adherence [38].

Participants also completed the Credibility/Expectancy Questionnaire at baseline as a process predictor. It measures how a person thinks and feels about their therapy and its likely success [39]. These are measured using two types of rating scales, one from 1 (not at all) to 9 (very much) and another from 0% (not at all) to 100% (very much), and it provides an overall score ranging from 3-27 for each factor [40].

The Patient Health Questionnaire-4 (PHQ-4) was used to measure anxiety and depression [41] collected at all intervals. This brief screening tool has been shown to be a reliable and valid measure in young people [42] and includes 4 items measured on a 4-point Likert scale ranging from 0 (not at all) to 4 (nearly every day) [41].

Treatment monitoring questions were included in order to collect data on what topical treatments participants were using, whether they experienced side effects, how they dealt with these, how often they were using treatment, and any other treatments they were using for their acne.

Sociodemographic questions included age, gender, education, age of onset of acne, and whether living with parents or independently.

The Skindex-16 overall mean score at baseline was 55.4 (SD 21.8) across both groups. There was a substantial improvement in both groups, and the mean differences between groups, when controlling for baseline scores and covariates (gender, age, age onset, and education), suggested a trend toward benefit at both 4 and 6 weeks: at 4 weeks, the intervention group had a score 5.2 points lower (95% CI −14.58 to 4.09) than the usual care group and at 6 weeks 2.9 points lower (95% CI −13.27 to 7.47; Table 5).

There was no evidence of a trend toward benefit in the symptoms subscale (intervention group 5.4 points higher at 4 weeks: 95% CI −8.41 to 19.22; 0.9 points lower at 6 weeks: 95% CI −11.76 to −10.03); however, some evidence of a trend toward benefit in the emotional subscale (intervention 12.4 points lower at 4 weeks: 95% CI −24.23 to −0.67; 3.9 points lower at 6 weeks: 95% CI −16.65 to 8.75) and functioning subscale (intervention group 6.4 points lower at 4 weeks: 95% CI −20.52 to 7.79; 3.4 points lower at 6 weeks: 95% CI −16.75 to 9.9; Table 5).

More people in the usual care group reported using topicals at baseline compared with those in the intervention group. In the intervention group, the percentage of people using topicals increased from baseline to 4 weeks by 13.5% and decreased by 0.8% in the usual care group (Table 6).

At 4 and 6 weeks, the usual care groups reported similar rates of side effects compared with the intervention group (Table 7). There was an increase of 13.5% from baseline to 4 weeks in the number of people reporting continuing treatment (including altering application as advised by the website) when experiencing minor side effects compared with the usual, which decreased by 2.4% at 4 weeks (Table 8). In both groups, the most common frequency of application at all intervals was once or more than once a day or most days. The intervention group and the usual care decreased similarly in the number of people reporting application of once or more than once a day or most days at 4 weeks (Table 9).