Self- and proxy reports of quality of life among adolescents living in residential youth care compared to adolescents in the general population and mental health services

Adolescents are placed in Residential Youth Care (RYC) mainly because of neglect, abuse or severe behavioural problems, coming from families exposed to psycho-social strains and parental addictive- and psychiatric problems. Adolescents in RYC show a high prevalence of mental disorders [1]. These factors increase the risks of adverse outcomes in adulthood with regard to physical health, mental disorders, and psychological well-being [2, 3]. Children in out-of-home care represent a high-risk group for low educational outcome [4‐6] which is associated with marginalisation and social exclusion [7, 8]. In Norway approximately 1200 youths aged 13–24 were placed in RYC in 2013 [9].

In children, Quality of Life (QoL) is defined as subjectively perceived well-being and satisfaction that can be best evaluated by the child itself, according to his/hers own experiences across several life domains [10]. The QoL concept is partly comprised of positive and negative affect towards health and life circumstances, as well as an emotional state that is determined by inter-personal aspects, temperament, etc. [11].

The perspective of positive psychology, i.e., to identify human strengths, fostering well-being, resilience, prosocial behaviour, and QoL [12], correspond well to the objectives of child welfare services. QoL should therefore be an important outcome measure. However, there is a lacuna of research on QoL in this area. The subjective component to an individual’s well-being [13, 14] has not received the same attention as mental health and psychosocial problems [15, 16]. We have only found seven publications addressing QoL among adolescents in child welfare [17‐23], but two publications [21, 22] were based on the same QoL data. Two studies [17, 19] had a small sample size, limiting the generalizability of the results. Australian children and adolescents, 6–17 year olds living in home-based foster care, showed significantly poorer QoL on a wide range of domains compared to the general community [18]. Polish children from children’s homes in one county, showed poorer QoL compared to children living in normal families [20]. It is unclear, whether the Polish version of the KINDL was reliable. Only three studies included QoL reports from youths living in RYC [17, 21, 23]. Of these, only two were representative, allowing for comparisons. Damnjanovic et al. [21, 22] compared 216 children and adolescents from residential- and foster care in Serbia to the general population, using the Pediatric Quality of Life Inventory v.4 (PedsQL^TM) [24]. Responders living in RYC reported significantly poorer QoL than those living in foster care or in biological families. In contrast, a Scottish study by Carrol et al. [23] found no difference in QoL as measured with the PedsQL^TM between adolescents living in residential care and a control group [23]. In these two studies adult proxy reports were not available to provide a supplementary perspective in addition to self-reports.

It is well-known from QoL research that there is a considerable disagreement between adolescent self-report and by proxy report, dependent on the study sample, i.e., general or clinical population, type of clinical population, etc. [13]. By definition of the WHO [25], QoL focuses on subjective well-being, thus self-report should always be the first choice and proxy report used when necessary. However, self-report can be biased by age, mental illness or by too positive self-evaluation [26]. Our recent publication on the same national sample showed a very high prevalence of psychiatric disorders. We found high comorbidity between emotional and disruptive behavioral disorders (conduct disorders), as well as a high prevalence of ADHD (32.3 %) and Asperger syndrome (23.2 %) [1]. Therefore a proxy information was relevant to give supplement information in addition to self-reports by the adolescents in the current study. Also, self-perceptions of children with ADHD have been shown to be overinflated in comparison to the perceptions of other informants, as parents and teachers [27]. These children tend to inflate their self-perception mostly in the domains of greatest deficit, such as conduct problems. This phenomenon is known as “the positive illusory bias”. A recent study, on the other hand, showed that self-perceptions of academic competence of adolescents positively predicted academic achievements at follow-up [28]. Thus, the academic discussion with regard to positive illusory bias in self-reports is still inconclusive. Regarding proxy reports, they also need to be handled cautiously due to biased connected to several parental factors, for example maternal depression [29]. Concerning people with disability agreement and reliability of proxy responses tend to be best for relatives, lower for friends, and lowest for health care proxies [30]. In the child welfare system, this so called “proxy problem” can be even more complicated, as biological or foster parents are not always available or appropriate as informants [15]. Davidson-Arad [17] illustrated the challenge of using multi informants in QoL assessment in the child welfare system. The author compared differences between QoL in 30 Israeli children and adolescents, aged 10 to 17 years, as reported by multi-informants. Child QoL was evaluated by the welfare caseworker, a professional uninvolved in the decision, the child, and the parents. Parents consistently rated their child’s QoL higher than the professionals did. At present, there is a lack of research comparing QoL adult proxy reports with child self-reports in the child welfare system, and to our knowledge, there are no studies using staff as substitute proxy informants of child QoL. Therefore, we wanted to explore the grade of association between adolescent self-report and the proxy primary contact report.

The existing research gap, methodological limitations, and inconsistent findings in the few existing QoL studies among adolescents in RYC, make it difficult to generate hypotheses about their well-being. One would expect that removing adolescents from a living condition with gross neglect and abuse, and provide care in RYC would considerably improve their QoL. However, there has been documented a high level of emotional and behavioral problems among adolescent in RYC [31‐33], which one would expect to impact the residents’ QoL in a negative way. It has also been documented that patients in Child and Adolescent Mental Health Services (CAHMS) have poor QoL [11, 34, 35] due to their referral for emotional and behavioral problems. There are no studies of adolescents in RYC where their peers in CAHMS were used as a meaningful reference group. Thus, we do not know if, and to which degree adolescents’ QoL in RYC is reduced compared to patients in CAHMS. Such a study would provide important information to decision makers, personnel in the child welfare sector, and personnel in the health services, so as to improve quality of care for adolescents in RYC.

The present study is the first large scale, nationwide study assessing QoL among adolescents in RYC, by both self- and proxy reports, compared to the general population and outpatients in CAHMS.