Illness perceptions in patients and parents in paediatric oncology during acute treatment and follow‐up care

Recent evidence suggests that illness perceptions in paediatric patients and their parents may differ, with parents holding more negative views compared to their children. Little is known about illness perceptions of very young patients and their parents. This study investigates illness perceptions in paediatric cancer patients aged 4–18 years and their parents in acute treatment or follow‐up care, distinguishing patients by age (4–11, 12–18) and stage of medical treatment.


| BACKGROUND
Every year, approximately 1 in 400 children in Germany is diagnosed with cancer. 1 In the last decades, treatment options have improved significantly, leading to increased survival rates of 82% for any cancer diagnosis, and even 90% for acute lymphoblastic leukaemia. 1 Despite improved prognosis, the cancer diagnosis and treatment remain extremely burdensome, and many paediatric patients and their relatives develop adverse psychological symptoms during and after acute treatment (syn. intensive therapy) such as anxiety and depressive symptoms, 2-4 posttraumatic stress symptoms, 5-7 lower health-related quality of life (HRQoL) [8][9][10] and elevated fear of progression, 11 indicating the importance of psychosocial support for the family during and after treatment.
One influential explanatory model on adverse psychological reactions due to illness is the Common-Sense Model of Illness Representation (CSM). 12 It poses that individuals create cognitive representations of a perceived health threat, for example, a cancer diagnosis. These so-called illness perceptions are based on available information (e.g., conversations with others, information by medical staff, own experience with the illness) and launch a self-regulation process which aims to overcome the health threat. They include perceptions on physical symptoms associated with the illness, expected timeline, consequences of the illness on the individual's life, causes, controllability and curability of the illness, and illness-related emotions 13 (for further information, see Section 2).
In paediatric oncology, the diagnosis of a chronic or lifethreatening illness is not only a personal experience for the affected child but also burdensome for parents. 14 When patients and parents are confronted with the cancer diagnosis, they may be exposed to each other's perceptions and anxieties. Research suggests that illness perceptions in patients and parents may differ. Studies revealed that children (aged 8-18 and 12-24 years, respectively) suffering from chronic illness such as cancer or spinal muscular atrophy held more positive views concerning their illness than their parents during acute treatment. [15][16][17] These findings highlight the importance of examining both the patients and the parents to ensure appropriate medical treatment and psychosocial care. For example, if parent and child indicate different assumptions regarding duration, this could be addressed in a family-centred intervention to talk about realistic expectations based on current professional knowledge.
Illness perceptions are also important for children and parents in follow-up care, for example, in psychosocial interventions to support coping. However, research on illness perceptions in child survivors of paediatric cancer and their parents is sparse. Evidence shows that adult survivors tend to hold more positive views as compared to their partners/carers who express more anxiety, more negative perceptions on the survivor's personal control and illness consequences. 18 Finally, child's and parent's illness perceptions may be associated with the child's HRQoL, as has been shown in different samples during acute treatment (e.g., 8-12-year-old cancer patients, 19 8-18-year-old children suffering from spinal muscular atrophy, 16 parents of young children with head and neck cancer 20 ). Similarly in follow-up care, adult survivor's illness perceptions may be associated with their HRQoL. 21 Interestingly, one study on diabetes indicated that higher agreement (syn. similarity) between child's and parent's illness perceptions may be associated with better emotional adjustment in the child. 22 Taken together, previous studies on illness perceptions refer to children aged 8 years or older, adolescents or young adults and are often assessed with the well-established Illness Perception Questionnaire-Revised (IPQ-R) 23 (or shorter adaptations 16,19,24

| Procedure
Participants in acute treatment were recruited in the acute wards for paediatric oncology at the university hospitals in Dresden and Leipzig, Germany. Participants in follow-up care were recruited in the parents' associations in Dresden and Leipzig (Sonnenstrahl e.V. Dresden, Elternhilfe e.V. Leipzig). One parent was also invited to HERZOG ET AL.  For example, immune therapy, BRAF inhibitor therapy.

| Illness Perception Questionnaire-Revised
The IPQ assesses illness perceptions according to the CSM.
For this project, the short German IPQ-R 28  (except timeline-cyclical). 28  -953 illness'). 4-11-year-old children (referred to as 'children') were examined using an age-adapted hand puppet interview containing the IPQ-R questions (available on request). 25,26 For the puppet interview, the items have been dichotomised and two hand puppets tell the child about their opposing views of their own illness (e.g., 'I am sad because of my illness' vs. 'I am not sad…'). The child then decides which puppet's view corresponds with his/her own perception.

| KINDL-R
The KINDL-R was used to measure the child's HRQoL. 29 It contains Likert-scale items on physical well-being, emotional wellbeing, self-esteem, family, friends, daily functioning (school, pre-school/nursery school) and illness. The HRQoL KIscore is calculated by taking the mean of the items after reverse scoring.
Higher scores indicate higher HRQoL. The KINDL-R shows acceptable to good internal consistency and high convergent and discriminant validity. 30,31  , Table A1). Table A2 shows IPQ-R scores and d-scores for this subsample, distinguishing by gender.

| Symptoms related to illness/treatment (illness identity)
Overall, children/adolescents in acute treatment and follow-up care reported significantly fewer symptoms as part of their illness/treatment than their parents (Table 2) 3.1.2 | Timeline (acute-chronic, cyclical), personal control, illness coherence, consequences, emotional representations (Table 2) Among dyads in acute treatment, there were significant differences on the timeline-acute/chronic dimension. However, while adolescents viewed the illness as being less chronic than their parents, children viewed the illness as being more chronic and having a longer duration. Moreover, children perceived significantly fewer negative consequences than parents, while there was no significant difference among adolescent-parent-dyads. Children and adolescents reported significantly less negative emotional representations than parents.
There were no significant differences between children/adolescents and parents concerning perceptions of cyclicity, personal control, and illness coherence.

Wilcoxon (z, p)
Acute -955 representations than parents. There were no differences between children/adolescents and parents concerning perceptions of personal control.

| Association with HRQoL (research question 2)
In both treatment phases, there were negative correlations (adjusted for multiple testing) between adolescent's consequence-dimension and HRQoL, indicating that the perception of fewer negative consequences was related to higher HRQoL (Appendix , Table A3). In both treatment phases and for all age groups, children/adolescents reported significantly less negative emotional representations than parents. A similar result has been found by Fischer and colleagues. 16 A reason might be that patient's and parents' roles during cancer treatment differ substantially: Patients often try to F I G U R E 2 Bar chart of the most commonly experienced symptoms during cancer illness/treatment, by stage of medical treatment and age 956distract themselves or are actively invited by hospital staff to spend time with music, art, or physical therapists, whereas parents may be more worried about their child's survival, monitor them closely to detect symptoms or changes in well-being, 32 spend a lot of time providing care either in the hospital or at home, and deal with other daily demands (e.g., care for other children, work). 33,34 Interestingly, dyads both in acute treatment and in follow-up care did not differ concerning their perceptions of the child's/ adolescent's personal control over the illness. This result is supported by two studies with paediatric patients during acute treatment, 16,17 whereas a study with adult cancer survivors found that they held significantly more positive perceptions concerning their own control over the illness than carers. 18 Possibly, the shared perception of controllability is due to children assuming their parent's cognitive patterns as communicated by them (e.g., illness as stroke of fate vs. illness as something that needs to be 'fought').
In acute treatment, no differences between children/adolescents and parents were found concerning perceptions of cyclicity and illness coherence. This is supported by findings from a sample of An important finding in our study was that children in acute treatment held more negative views concerning the chronicity of their illness than parents, whereas adolescents both in acute treatment and follow-up care held more positive views. A possible explanation might be that children apply a different time frame than parents (or adolescents) 35 : When asked if they think that their illness will last a long time, they might consider a couple of weeks in hospital already a very long time and respond accordingly.
Children in acute treatment perceived fewer negative consequences than parents, whereas adolescent-parent-dyads did not differ in their view on consequences. This might be because children are informed differently about their illness (e.g., survival rates, potential side/long-term effects) than parents or adolescents.
Given that children's cognitive ability to assess risks is still developing, 15

| Study limitations
To our knowledge this is the first study that investigated illness perceptions in paediatric cancer patients from the age of 4 and their parents. The absence of a comprehensive assessment instrument and of similar studies for comparison limits the interpretation of our findings. To assess illness perceptions of younger children aged 4-11 years, we developed a puppet interview based on the IPQ-R.
This allowed us to investigate illness perceptions in children and adolescents from a broad age rage, as well as their parents, using the same items. It could be argued that the dissimilarity in illness perceptions between children and parents was due to the different assessment methods. However, the puppet interview is a state-ofthe-art approach to conduct assessments in young children, 36 and the items of the IPQ-R were carefully age-adapted, retaining their meaning. We used the German short version of the IPQ-R, 37 which has been shortened through item reduction to 32 items as opposed to the 46 items of the full version. On the upside, this allowed the puppet interview to be much shorter and therefore more suitable for very young children. On the downside, the dimension 'treatment control' is not included. The puppet interview has been piloted with n = 11 children in acute treatment and follow-up care and showed good validity, comprehensibility and feasibility. 25,26 Response rates in acute treatment and follow-up care were high.
Nevertheless, sample sizes for child-parent dyads in acute treatment and follow-up care, and for adolescent-parent dyads in acute treatment were small to medium and therefore only able to detect medium to large effect sizes. Future research should incorporate sufficient samples for all age groups. Moreover, future studies should explore differences between cancer types (e.g., leukaemia vs. brain tumour) as they might pose different situations for the affected family. Sample sizes should be big enough to enable research to investigate this.
As we were the first to investigate differences in illness perceptions in cancer patients aged 4-18 years and their parents both during acute treatment and in follow-up care (research question 1) we conducted an exploratory, hypothesis-generating analysis and therefore did not correct for multiple testing for these analyses.
All families in acute treatment, as well as many families in followup care received psychosocial support at the time of the study. To which extent this influences the participants' illness perceptions and coping cannot be resolved here.

| Clinical implications
In line with other research and in accordance with clinical experience, our study suggests that child's/adolescent's and parent's illness perceptions differ significantly from one another. Therefore, parental reports should not be used as proxy reports on the child's/adolescent's illness perceptions and children/adolescents should be asked to report on their own view. This might also HERZOG ET AL. enhance their need for control and autonomy, and might have positive effects for their psychological well-being. 16 However, we would nevertheless like to stress the importance of the parent's self-report as an additional source of information. Obtaining a comprehensive report of both members of the dyad is vital for the medical and psychosocial staff to support children and parents according to their individual needs and views on the illness. Studies with adult patients have shown that interventions directed at changing maladaptive illness perceptions may improve selfmanagement behaviours. 38 Up to date, no illness perception interventions targeted at children are available. It should be kept in mind that these interventions should not only be patient-focussed, but rather family-focussed or parent-focussed as the illness of the child takes place in the whole family system. 39,40 In a family/ parent-focussed intervention, the differences between child's/adolescent's and parent's illness perceptions could be addressed: Parental (more negative) perceptions may be relevant for the child's perceptions and thereby obstruct the child's coping. 20 On the other hand, the child's more positive perceptions could be used as a resource for the family and may be a relief for the parents.
Therefore, family-and parent-focussed interventions would benefit the whole family system. Future research should focus on creating and evaluating such interventions.

| CONCLUSIONS
In summary, our study showed that children/adolescents in paediatric oncology and their parents hold different illness perceptions, with parents in general holding more negative views. This was true both in acute treatment and in follow-up care. As illness perceptions are strongly related to HRQoL, 16,20 psychosocial staff should assess both the child's as well as the parent's illness perceptions, and, if necessary, change them to more positive perceptions using interventions including the whole family system.