Health-related quality of life in children with haemophilia in China: a 4-year follow-up prospective cohort study

A total of 42 subjects with haemophilia were recruited along with their parents from Nanjing Children’s Hospital affiliated with Nanjing Medical University, between August 2014 and January 2018 Parents (or legal guardians of the subjects) were interviewed in person to collect clinical information. Among these 42 subjects, 37 had haemophilia A and the others had haemophilia B. The median age of subjects interviewed was 5.48 (SD = 4.63). Patients with cognitive impairments, limited literacy skills, or other chronic illnesses such as cerebral palsy, diabetes, chronic immune thrombocytopenia, rheumatological disease and asthma were excluded from this study. According to the subjects’ disease history, no subjects were excluded. Besides, those who did not complete CHO-KLAT during the needed time-frame were considered ineligible for this research. Children aged 7 years old and older were asked to complete self-report versions of CHO-KLAT and self-report versions of SEC with little or no interruption from parents.

This is a 4-year prospective cohort follow-up study. Their early and recent responses were considered baseline data and follow-up data respectively. Events considered in event-free survival analysis were death, developing target joints (a target joint is a joint in which three or more spontaneous bleeds have occurred within a consecutive 6 month period [5]) and severe bleeding during the observation. CHO-KLAT is a disease-specific HRQoL measure that can be used in both clinical and home settings and has been updated to version 2.0 [8, 18]. It has been validated in several countries, including Canada, France, Germany, the Netherlands, Spain, the UK and China [14, 19] and consists of 2 questionnaires: one for children aged from 5 to 18 (child self-report), the other for parents on behalf of children aged from 2 to 18 (parent proxy report). This Chinese version includes all 35 original items from the CHO-KLAT. Scores have a range of 0 (worst) to 100 (best). There are no normative values of the CHO-KLAT. Higher scores mean better HRQoL. Children aged 7 years old and older, if literate, were asked to complete a self-report CHO-KLAT. Also, their parents were asked to complete a proxy CHO-KLAT on their behalf. The Chinese version has an additional unique SEC module, which consists of 9 questions covering information about stigma, opportunities for stable employment and access to treatment due to cost, etc. SEC scores have the same range as those of the original CHO-KLAT. The questionnaires were handed in hard copy to parents and patients in the clinic. Data were entered into an electronic database by researchers. All data were double entered to ensure accuracy. Data checks and cleaning were also performed. If more than 1/4 of answers were not given in one report, its score would not be shown and it was rejected for the research.

Statistical analyses here were performed using Statistical Package for Social Science (version 22.0; SPSS). Qualitative variables were described in number and percentages and mean ± SD, if quantitative. Preliminary analyses used paired sample t-tests to evaluate differences between baseline and follow-up patient-rated HRQL and one-sample t-tests to compare patient HRQL to normative values. ANOVA tests were used when comparing quantitative data between 2 groups or more. The Spearman coefficients were used when studying the relationship (direction and power) of quantitative variables simultaneously. The correlations between mean scores and bleeding rates were performed using correlation analysis. Kaplan–Meier survival curves were plotted, and the groups compared using the log-rank test to ascertain the relationship between treatment and prognosis. Cox regression model adjusted age, region of residence, severity and Family injection was used to calculate the P value. The events we considered in KM analysis were death, target joints developing and severe bleeding. Two-sided P values were selected and P < 0.05 was considered statistically significant.

The frequency distributions of all subjects of the selected variables (the demographic, clinical, and biochemical data at enrolment, as well as treatment during follow-up) are provided in Table 1. 18 cases (42.9%) had a history of severe bleeding; 15 (35.7%) suffered from intracranial hemorrhages and 11 (26.2%) had a single target joint at the time of recruitment while 4 cases had multiple target joints. 8 cases had restricted physical activity at baseline. The restricted physical activity is defined as impact on daily life activities in the past one month because the QoL questionnaires are filled according to participants’ situation during the past one month. None of the patients were known to have an inhibitor at enrolment. All the 4 mild hemophilia boys, 16 of 24 children with moderate disease, and 5 of 14 with severe disease received an on-demand therapy, while the others were subjected to prophylactic treatment at medium dose (15 - 30 U/kg, two or three times a week). During the follow-up, 4 patients were lost (2 cases had a history of target joints). Only 1 case suffered from severe bleeding (urinary system) and 4 cases developed target joints. 4 cases with target joints at baseline got better (the target joint stopped experiencing symptoms). Of all cases at the endpoint, 13 cases had target joints; however, merely 7 cases had restricted physical activity. One patient had an inhibitor.

The results of this comparison are shown in Table 2. At baseline, all parents on behalf of their children completed the parent version of the CHO-KLAT and SEC while only 12 patients were old enough to complete it independently. The mean scores for the CHO-KLAT child/proxy report were 60.69 (SD = 20.28) and 61.01 (SD = 12.14) respectively. For the SEC, the mean scores were 60.62 (SD = 14.32) and 56.43 (SD = 12.59) respectively. There were no significant differences between the scores of 12 proxy and child self-reports in CHO-KLAT or SEC (p = 0.440 and p = 0.275, respectively). At follow-up, 38 parents completed the scales. 18 children were old enough to complete the child self-reports scales independently. The mean score for the CHO-KLAT child self-report and parent proxy report were 64.93 (SD = 13.71) and 65.33 (SD = 15.78) respectively. The child/proxy report SEC scores were 65.09 (SD = 15.36) and 61.62 (SD = 17.46). Statistical significance was not achieved between the scores of 18 proxy and child self-reports in CHO-KLAT or SEC (p = 0.694 and p = 0.522, respectively). Scores at follow-up were higher than those at baseline in both CHO-KLAT and SEC but no statistical significance was observed.

No major relation was found between the two types of haemophilia and the scores of the child/proxy reports. Urban patients had significantly higher scores than rural patients in CHO-KLAT. Considering urban patients had higher insurance coverage than rural patients (85% vs 30–75%), financial status would be an explanation for the above difference, which was also proved by the SEC results in Table 3. As with severity of disease, the statistical difference appeared only in the CHO-KLAT parent proxy report group (p = 0.037) and SEC child self-report group (p = 0.040) at baseline. The mean scores of CHO-KLAT and SEC child/proxy reports in the prophylactic group were significantly higher than those in the on-demand group, especially for the parent proxy reports for CHO-KLAT and SEC. Patients who had received home injection got higher QoL at follow-up.

The mean of annual bleeding rate (ABR) and annual joint bleeding rate (AJBR) were 7.29 (SD = 9.73; ranging from 0 to 48), 2.86 (SD = 5.82; ranging from 0 to 24) at baseline, respectively, and 8.50 (SD = 10.16; ranging from 0 to 40), 4.50 (SD = 7.64; ranging from 0 to 40) at follow-up, respectively. The amount of change during this follow-up study in ABR (ΔABR) and AJBR (ΔAJBR) was 2.71 (SD = 9.40; ranging from − 20 to 30) and 1.89 (SD = 8.86; ranging from − 22 to 38).

Table 4 describes the correlations between mean scores and bleeding rates. As expected, a negative relation was apparent between ABR, AJBR, ΔABR, ΔAJBR and scores. The parent proxy report CHO-KLAT was most strongly related to the ABR (r = − 0.609, p = 0.000), AJBR (r = − 0.353, p = 0.030) and ΔABR (r = − 0.573, p = 0.000). Additionally, a significantly negative correlation was found between SEC parent proxy report scores and ABR, ΔABR, ΔAJBR (r = − 0.569, p = 0.000 and r = − 0.538, p = 0.000 and r = − 0.381 p = 0.018, respectively). There was no correlation between bleeding rates and child self-report scores of CHO-KLAT. Pearson bivariate correlations revealed positive significant association between SEC and patient HRQoL regardless of baseline or follow-up. The follow-up results showed that child self-reports had higher correlations with parent proxy reports in CHO-KLAT (r = 0.772) and SEC (r = 0.690). The correlation analysis indicated that SEC showed intimate correlations with the CHO-KLAT, suggesting the SEC module was necessary for Chinese patients.

The events we considered here were death, target joints developing and severe bleeding. Throughout the follow up, 5 special cases were recorded. One of them suffered from severe bleeding (urinary system) and the rest developed new target joints. The 5 cases observed were all haemophilia A, severe in 3, moderate in 1 and mild in 1. When it comes to treatment assessment, 4 cases received on-demand treatment whereas only 1 case received prophylaxis treatment. In a comprehensive multivariate Cox regression analysis including age and region of residence, the treating method remained an independent prognostic variable for haemophilia (P = 0.036, Fig. 1). Prophylaxis treatment is a key factor contributing to prognosis.