The wounded self—lonely in a crowd: A qualitative study of the voices of children living with atopic dermatitis in Hong Kong

2.1 Participants and sampling

We recruited participants from a group of children (n = 55) who were waiting to participate in a psychosocial program specifically for children living with AD in Hong Kong. Only primary schoolers clinically diagnosed as having AD were eligible to participate in this program. By using the SCORing Atopic Dermatitis (SCORAD) index (European Task Force on Atopic Dermatitis, 1993), the assessments of children's severity of AD were administered by the first author, who was well-trained by a dermatologist in conducting the assessments. The assessments were conducted between June and September 2017. The sociodemographic characteristics of the children were reported by their parents in a questionnaire.

To achieve the maximum variation (in age, gender and severity of AD), we purposely selected a heterogeneous group of 17 children living with AD (eight girls and nine boys aged between 8 and 12 years) to participate in our study (Table 1). Eleven had severe AD, with a SCORAD score above 50 (maximum total score = 103). The remaining six had moderate AD, with a SCORAD score between 25 and 50. Ten participants had shown AD symptoms in the first year of life. Seven had endured AD symptoms for more than 10 years. The six most common types of treatments received by the participants were ointment (n = 17), partial steroid treatments (n = 17), traditional Chinese medicine (n = 16), partial non-steroid treatments (n = 14), wet wrap therapy (n = 8) and oral antibiotics (n = 8). Five children had other health conditions, such as allergic rhinitis (F6 and M4), asthma (M7), ADHD (M3) and ASD (M5). Two participants (M4 and M6) were from low-income families with a monthly household income of less than HK$20,000. Fourteen families had more than one child. Nine families had more than one affected child.

2.2 Data collection

Semi-structured interviews were conducted with 17 children between September 2017 and February 2018. The draw-and-explain technique was used in these interviews to increase the amount of information reported by the children and to enhance their interview performance (Günindi, 2015; Wennström, Hallberg, & Bergh, 2008). Before each interview, the child was provided with one A4 sheet of white paper, one A4 sheet of paper with five concentric circles and a set of 12 coloured pencils. In each interview, the child was asked to draw two pictures. For the first picture, the child was asked to draw whatever he or she would like on a blank piece of paper when thinking of his or her AD. Then, the child was asked to explain his or her drawing. For the second picture, the child was given a sheet of paper with five concentric circles in the centre. (S)he was asked to draw himself or herself in the inner circle and to draw people related to him/her in the surrounding circles; (s)he was asked to place those most important or those he/she liked most closest to him or her and to place those less important or those he/she did not like in the outer circles (Mason & Tipper, 2008). Then, the child was asked to explain his/her relationships with these people and the reasons for their specific locations. A set of open-ended questions was asked when children explained their drawings; such questions included “Could you please tell me more about this drawing?” and “How do you feel about this?” Two pilot interviews were conducted to fine-tune the interview procedures and the interview protocol.

The interviews were conducted by five interviewers with a professional background in social work or psychology within a room selected in each social service centre. Each interviewer received 3 hr training: a 1-hr lecture, 1 hr of role play and 1 hr of interview modelling; professional supervision was provided by the first author. All interviews were audio-recorded. The tape-recorded interviews lasted from 22 to 75 min (mean duration: 53 min). Interviews were conducted in Chinese, and the transcripts were reviewed for accuracy by the first author.

2.3 Data analysis

All Chinese transcripts of the children's verbal expressions during the interviews were organised and analysed using NVivo software (version 12). Moustakas’s (1994) method was used to analyse the phenomenological data. All transcripts were read in their entirety a minimum of three times before analysis to obtain a sense of the whole database. Significant statements—segments of text relevant to the research questions—were identified and listed. Then, the significant statements were grouped into meaning units (codes), which were then classified into themes. In addition, memos were written to extract concepts from particular phrases, individual interviews and reviews in multiple interviews to track the development of ideas or themes (Creswell & Poth, 2018).

“Epoché” is an important concept in phenomenological research, which was attempted during the data analysis. This concept requires researchers to suspend as much as possible their prejudgments regarding the phenomenon under investigation before attempting to explore the experiences of the participants (Tufford & Newman, 2012). The main investigator (the first author) is a licensed social worker and a Ph.D. with several years of experience conducting research with children living with health issues. She had previously worked closely with more than one hundred families with a child diagnosed as having AD. This experience helped her to understand the difficulties faced by the participants. However, this experience might have biased her understanding of the whole picture of the experiences of children living with AD by inhibiting her ability to see the resilience or strength of the participants.

To reduce bias, a research assistant with a master's degree in social work and who had no experience working with children living with AD also performed the data analysis. The research assistant analysed all the transcripts independently by applying a coding framework developed by the first author. The initial interrater reliability was high (k = 0.83). Disagreements were resolved through discussions, which led to the creation of a revised codebook (Table 2). Then, the first author reviewed all the transcripts according to the revised codebook. The transcripts cited in this study were translated into English by the first author and reviewed by the other two authors.

2.4 Ethical considerations

We received ethical approval for the study from The University of Hong Kong Research Ethics Committee (Ref. EA1707024). Participation was completely voluntary. The purpose and procedures of the study and the rights of the participants were explained to all the participants and their parents. Written informed consent was obtained before the interviews.

3.1 The child: A vicious circle of skin and mind

3.2 The family: A combination of conflicts and support

3.2.3 High educational stress from attempting to meet parental expectations

Nine participants explicitly referred to their educational stress and connected their academic challenges to their AD. The participants believed that their skin became worse when taking tests, as they were under “enormous pressure.” F8 said, “Every time when there is an exam, all AD will come out.” Figure 1 presents her drawing.

More importantly, the participants indicated that their educational stress was mainly due to their parents’ high expectations. As M1 said:

My mother makes high demands on my academic performance. She always wants me to study. It’s stressful.

F3 described her father’ reaction when she wanted to sleep and didn't want to study:

If I cried, then I could sleep. If I said that I wanted to sleep, he still asked me to continue to study… I think my skin would be better if I could sleep more.

Some parents even insisted that good school performance might help to prevent the discrimination children faced as the result of AD. The participants often felt very worried, tense or even guilty when their academic performance did not meet parental expectations.

3.3 The school: Full of stress, little support

3.3.2 Bullying and victimisation

All participants, except for F6 and M8, described their problematic relationships with peers in school. Moreover, 12 participants indicated that they were bullied verbally, socially and even physically because of their impaired skin. Classmates often laughed at the skin of the participants and even used insulting and demeaning words, such as “virus,” “dirtiness,” “infectious diseases,” “garbage,” “waste” and “ugliness.” M2 depicted being bullied in school in his drawing (Figure 2).

3.4 The neighbourhood: Stigmatisation and discrimination

3.5 The wounded self: An outcome of skin, mind and environment

3.5.1 Lonely in a crowd

The participants indicated that they felt very lonely because they were not being understood at home and had no friends in school. F2 said that AD “stole the key to my interpersonal relationships.” Her drawing is presented in Figure 3. Deteriorated interpersonal relationships caused their negative self-image and low self-esteem, thus aggravating their psychological disturbances.

3.6 A child-centred biopsychosocial framework

The composite description of the experiences of children living with AD contributed to developing a child-centred biopsychosocial framework (Figure 4). Living with AD means contending not only with the biological symptoms for school-aged children but also with an accumulation of challenges and crises in their family, school and neighbourhood. The experiences of the affected children were impacted by the synergy of individual and environment factors that subsequently resulted in an incapable and wounded “self.”

4.1 Implications

Healthcare professionals should target different challenges faced by children living with AD at the individual and environment levels. At the individual level, three important aspects are identified: First, topical therapies for treating physical symptoms should be primary in the management of paediatric AD (Stein & Cifu, 2016). Second, screening procedures for identifying psychological disturbances in children living with AD should be emphasised in conjunction with medical treatment (Bronkhorst et al., 2016). Third, improving the coping skills of children for dealing with teasing and bullying and empowering them to separate their core identity from AD may be useful in helping children living with AD to develop a positive self-concept and adapt to their environment.

At the family level, resolving conflicts and highlighting resources are equally important. We suggest providing intervention services to children living with AD and their parents together in a parallel format. Contents of intervention programs should integrate educational components (such as those that provide AD-related information and teach care skills) and psychosocial components (such as those that reduce the psychological stress of both children and parents, improve parenting strategies and improve parent–children relationships).

Working with teachers may be the key to the success of school-based programs. How teachers treat affected children may not only directly impact the psychological well-being of these children but also indirectly influence other students’ responses to them. Training teachers about AD, facilitating their awareness of the needs of this group of children, encouraging them to share information on AD with students and improving their skills in recognising and dealing with school bullying may be effective in developing a supportive school environment for children living with AD.

In neighbourhoods, a prevention strategy for attracting social attention to children living with AD and changing public attitudes about AD through education should be the first step.

4.2 Limitations and future studies

First, sampling bias might exist in this study. The sample was recruited from a group of children living with AD who registered together with their parents to participate in a psychosocial program in Hong Kong. The participants may be more likely to have a good relationship with their parents and to receive support from them than children whose parents refused to participate in this psychosocial program. A good parent–child relationship may be one reason that AD did not fundamentally impair the parent–child relationships in this sample. More attention should be paid to affected children living in more vulnerable families. In addition, future empirical efforts should also explore the experiences of affected children in different age groups and in other sociocultural contexts. Second, the interviews may be inconsistent because the data were collected by multiple interviewers. Some techniques were used to minimise this limitation and improve the quality of the interviews; these techniques included providing training to all interviewers, using a detailed interview protocol, and involving a supervisor in each interview. Third, the data analysis used in this study was based mainly on children's verbal expressions rather than their drawings. Future research may need to incorporate an analysis of visual data.