This is the first study to assess multidimensional HRQoL in children with SCD using a self-report measure and a control group of healthy siblings. In general, the HRQoL of children with SCD appeared similar to the HRQoL of healthy siblings, with the exception of lower scores on the physical domain for children with SCD, and, specifically in adolescents, lower scores on the autonomy domain. However, the HRQoL of both children with SCD and healthy siblings was considerably lower than the Dutch norm population on several domains, specifically in children aged 6-11. This implies that reduced HRQoL in children with SCD is primarily related to the low SES of this patient population, although there are disease specific effects on the physical and autonomy domain.
A large study in urban elementary schoolchildren from poor socioeconomic areas formerly established the essential impact that SES can have on HRQoL. These children reported lower HRQoL than chronically diseased children [24
]. Nevertheless, prior studies in children with SCD suggested that the HRQoL of children with SCD is impaired, even after considering the potential detrimental effect of low SES [15
]. The contrast between these previous findings and the present results may be explained by the use of different instruments, different comparison groups and particularly the use of proxy- versus self-report. In our study, a self-report questionnaire was administered, while other researchers have mainly used parental proxy-report. Parents of chronically diseased children generally tend to report worse HRQoL for their children than children themselves [25
], which was confirmed by previous studies in children with SCD [13
]. This may have led to an underestimation of the HRQoL of children with SCD. Parents could underestimate their children's HRQoL as a result of parental distress [27
]. Alternatively, parents and children could have different perceptions: while children may be unaware of the potential consequences of the disease, parents may have greater concerns for their future [28
] and take on a protective attitude.
Parental (over-)protection could also underlie our finding of lower scores of adolescents with SCD on the autonomy domain, compared to healthy siblings. This finding is congruent with results of a previous study in our hospital [13
] as well as other studies on adjustment in adolescents with SCD [29
], and seems to be associated with consequences of the disease. Therefore, we recommend that parental overprotection should be considered and targeted in intervention programs for adolescents with SCD. As higher levels of self-efficacy were previously found to be associated with fewer SCD symptoms [30
], interventions to increase autonomy could possibly lead to a decrease in SCD symptomatology in this age group.
Interestingly, adolescents did not differ significantly from healthy siblings or the Dutch norm population on any of the other domains except for autonomy, while children aged 6-11 did have significantly lower scores than the norm population on 7 of the 10 domains. Although adolescents have been described to generally experience more problems in psychosocial adaptation than younger children [7
], HRQoL scores of adolescents were within the normative range in a previous study as well [27
]. The resilience of adolescents could be a consequence of developmental growth and adjustment, possibly causing a better coping style with increasing age. This should be investigated further longitudinally.
Surprisingly, children with SCD did not report any social problems. This is a remarkable contrast to the results of previous findings [6
], including results of our own work in which parents rated their children with SCD to show less competent social behavior than healthy siblings [31
]. Furthermore, it is striking that children with SCD did not report a lower HRQoL on the School environment domain, as studies on neurocognitive sequelae of SCD have shown a decrease in general intellectual ability as well as deficits in specific neurocognitive domains in children with SCD [32
]. Children with SCD may have adjusted to their social and cognitive deficits and therefore do not subjectively experience HRQoL problems in these domains. In future research behavioural and neurocognitive data should be correlated to HRQoL, to gain more insight into the relation between objectively identified problems and the subjective HRQoL of children with SCD.
While interpreting the results of this study, strengths and limitations should be taken into account. Findings of this study are strengthened by the use of a well-standardized self-report measure and robust statistical methods to take within family correlations into account. Furthermore, by including a control group of healthy siblings with the same age, ethnicity and SES, the differential effect of low SES and disease specific consequences of SCD on HRQoL could be determined.
However, not all participating children with SCD had an eligible sibling, causing us to recruit healthy siblings from the entire SCD patient population. This led to an overrepresentation of children from Surinam descent in the healthy sibling group. Nevertheless, demographic differences between the groups were not statistically significant and explorative within-group analyses revealed no significant differences between children from Surinam or African descent on any of the HRQoL domains. Another limitation was the small sample size, mitigating statistical power. The results of our exploratory analyses need to be interpreted within this limitation. Moreover, the current study design did not allow us to investigate if there is an interaction effect of SES and sickle cell disease on HRQoL. Further limitations concerned the instrument we used to assess HRQoL. The KIDSCREEN-52 is not yet validated in children aged 6 and 7. However, young participants received cognitive debriefing to ascertain reliable self-report, and this issue was taken into account in statistical analyses. Children as young as 5 years of age have been found to be able to reliably report HRQoL [26
]. Furthermore, data were collected in a clinical setting, which may have affected the response of the participants. Finally, the KIDSCREEN-52 assesses HRQoL of the past week. As SCD has a very unpredictable course, it would be more appropriate to assess HRQoL over longer time periods, e.g. the past month. These issues should be taken into account in future research.