In recent years, measurement of HRQL has become more important in the assessment of several pediatric diseases, including ITP. The lack of evidence that any of the current treatments can modify either the short or long-term outcomes in children with ITP generated interest in assessing HRQL in this population. This study aimed to determine the impact of ITP on children’s HRQL from diagnosis to six months after diagnosis and to identify any risk factors.
Considering the first objective, we observed a slightly lower score for baseline emotional functioning in parent-proxy and self reports as well as the school functioning in self-reports when compared to original norm data. Surprisingly, this difference was not found when the comparison was performed with the French healthy population : results were generally even higher than in the healthy population for most items. The fact that children enrolled in the French reference population were older than those of our study could be one of the possible explanation for this difference, as younger children are less likely to worry. But this does not explain all the differences observed in parent-proxy reports. These differences were not sufficient to be considered as an impairment, nor a minimal clinically important in terms of quality of life. We did not observe any difference of HRQL between self or parent-proxy-reports and the original norm data after 6 months of ITP evolution either. The results are not surprising as ITP is a blood disease which causes considerable worry about both rare but serious bleeding and having a more serious condition. Over the following months, most patients (children and parents) experienced an improvement of HRQL and at M6 quality of life was the same as the healthy population. On all major parameters, HRQL scores increased between D1 and M6 and, considering the global score, the majority of patients (54.5%) perceived a change in quality of life as measured by MCID. This was an expected result as it coincides with the natural favorable evolution of the disease [8–27]. In our cohort, as in previous reports, about two thirds of children had recovered a normal platelet count at 6 months and bleeding scores improved with time in the majority of the patients. We can also hypothesize that the knowledge concerning ITP that parents and children develop over time, helps reduce their anxiety . Interestingly, significantly lower ICC was noted for children of 5 to 7 years old, especially for emotional functioning at diagnosis and for social and school functioning at M6, as already mentioned in the literature . More generally this difference is observed in the global score with better agreement for children of more than 13 years old. Although MCID does not significantly differ depending on age, these data could suggest a trend for young children to be more affected by ITP. The strict limitation of the child’s activities enforced by the parents to protect them from serious injuries could be one of the major reasons for their stress and their parents’ concern, which lead to a lower score. On the other hand, in older children, one can also speculate that HRQL self-report results are a representation of the sense that teenagers might believe themselves to be invincible.
Considering the second end-point, from our results, a lower score in social functioning seemed to be associated with an initial hospitalisation and the quality of life perceived in term of global score at M6 was associated with a viral infection before ITP (Table 5). Other variables such as age, platelet count at diagnosis and 6 months, initial bleeding scores or initial treatment did not influence scores of children or parents. This result is consistent with the findings of Neunert et al. who found only weak and non-significant negative correlations between bleeding severity or platelet count at diagnosis and HRQL . In our series about two thirds of patients had an initial ICIS score ≥ III with a platelet count < 10.109/L, with both variables improving with time, lowering anxiety of parents and children. Conversely, a recent viral infection could potentially have symptoms affecting the child’s HRQL, with a subsequent improvement in HRQL once the infection resolves, explaining the significant change in the MCID between D1 and M6. Similarly, the initial hospitalization would impact on HRQL, regardless of the reason for hospitalization. So, these two factors (viral infection and initial hospitalization) appear to effect the MCID independently of ITP.
Some limitations in our study should be pointed out. First, the limited size of our cohort did not allow us to analyse clinically relevant subgroups of patients such as those with persistent ITP at M6, with regular physical activity or patients from different socio-professional origins. Second, patients were recruited from seven pediatric treatment centers. These various origins may have induced a clustering effect of HRQL according to the medical and paramedical speech at diagnosis and through-out the follow-up. Third, we did not differentiate assessments of mothers and fathers. We can speculate that parents can have different evaluations of their own child quality of life. Fourth, a significant number of patients had an initial ceiling effect at 100: in such a condition, we cannot detect any improvement in quality of life during follow-up. Finally, although PedSQL™ 4.0 is an international tool, translated and linguistically validated in French, there is, to our knowledge, only one study establishing references in a French healthy population of 185 children . We compared our results to those published in the larger series published by Varni et al. : differences observed (Table 3) raise the question of the validity of some of our comparisons especially as the validation performed by Tessier et al.  was conducted among school children (over 5 years). As emphasized by these authors, further work would probably be required in the French population to determine real reference values for PedSQL™ 4.0.
However, despite these limitations, we think it important to note that ITP does not significantly affect HRQL as measured by the PedSQL™ 4.0. Even if we observe an initial decrease in scores (especially in the psychosocial field), it remains moderate and quickly corrected within 6 months after diagnosis. We can wonder if the PedsQL™ 4.0 Generic Core Scales is a sufficient tool to measure HRQL in children with ITP. Recently, Klaassen et al.  performed a cross-culturally translated KIT (Kids’ ITP Tools) validation. This tool is specifically designed to measure HRQL in children with ITP. Its validity was assessed by a recent study in four countries and French newly diagnosed patients, different from those reported here, were enrolled in this study . In the whole study population, the authors note that the KIT scores are moderately correlated with the PedsQL scores. Taking into account these results and our own results, we think it would be interesting to study the quality of life of these patients with a larger cohort.