Health-related quality of life assessment in Indonesian childhood acute lymphoblastic leukemia

Indonesia is a multi-island state, and has a population of approximately 220 million of whom 37% are children under 15 years. A childhood leukemia incidence of 2.5 to 4.0 new cases per 100 000 children leads to an estimated 2000 to 3200 new childhood ALL cases each year [22]. Our study was conducted at the pediatric department of Dr. Sardjito Hospital in Yogyakarta, Indonesia. Dr. Sardjito Hospital is a teaching hospital of the medical faculty of the Gadjah Mada University, and a tertiary-care referral hospital of the Yogyakarta and Central Java Provinces. Annually aproximately 30–40 children are diagnosed with ALL. The pediatric department consists of a clinic (VIP, 1^st, 2^nd, and 3^rd class) and a policlinic (VIP and general).

After cultural linguistic validation, a cross-sectional study was conducted with childhood ALL patients and their guardians in various phases of treatment using the Pediatric Quality of Life Inventory™ (PedsQL™) 4.0 Generic Core Scale and the Pediatric Quality of Life Inventory™ (PedsQL™) 3.0 Cancer Module. The cultural linguistic validation process consists of 3 steps: forward translation, backward translation, and patient testing. In the first step, the instruments were translated from English version to Bahasa Indonesia version, and it was done independently by two local professional translators, native Indonesia speakers, and bilingual in English language. The second step, the Bahasa Indonesia version was translated back to English version and it was done independently by two local professional translators, native English speakers, and bilingual in Bahasa Indonesia language. In this step a comparison between the backward version with the original version was assessed in order to detect any misunderstandings, mistranslations or inaccuracies in the intermediary forward version of the questionnaire. Finally, the translated questionnaire was tested in 5 cancer families, both in children and their guardians, to determine whether the translation (instructions, items and response choices) is acceptable, whether it is understood in the way it is supposed to be, and whether the language used is simple and appropriate. All PedsQL™ translations were conducted in close ongoing collaboration with Mapi Research Institute in Lyon, France and with Dr. James W. Varni.

Participants were childhood ALL patients who were hospitalized or visited the policlinic at Dr Sardjito Hospital, Yogyakarta, Indonesia, from October 2006 until January 2008. Eligible patients were children of 2–16 years at time of diagnosis. Patients with severe conditions who were not able to answer the questionnaire were excluded. During this period 2 ALL dexametasone-based protocols were used: Wijaya Kusuma ALL protocol and Indonesia ALL protocol. According to both protocols there are 2 risk groups: standard risk (SR) and high risk (HR). Both protocols consist of an induction, consolidation, re-induction (only for HR) and maintenance phase. We classified the phase as either intensive (induction, consolidation, re-induction) or non-intensive (maintenance). Interviews of eligible patients and their guardians, who were interviewed separately, were performed individually by a trained psychologist. It took 10–20 minutes for guardians and 20–45 minutes for patients to complete the questionnaire. Demographic data regarding information about age, relationship of respondent to child, number of children in the family, parental educational status, and socio-economic status were collected from guardians. Parental educational status was categorized into low educational status (no education, elementary school, junior high school) and high educational status (senior high school, academy, university). We classified SES as either poor or prosperous. This classification was based on 2 determinants: monthly income level of parents and assigned hospital class during the diagnostic process. Both determinants are obtained routinely during admission to the clinic and recorded in the medical records. The threshold per month income for poor versus prosperous families was set at 1.000.000 Indonesian Rupiah (about 100 US dollar). Patients attending VIP or first class wards and VIP policlinic were classified as prosperous. In case there is discordance between family income and hospitalization class or data of family income is not available, we used hospitalization class as SES determinant. Medical data regarding diagnosis and treatment status were obtained from the patients' MR. The study was approved by the Medical Ethics Committees of Faculty medicine, Gadjah Mada University. Informed consents were also obtained from the guardians.

The PedsQL™ 4.0 Generic Core Scale (Generic Scale) is a multidimensional instrument developed by Varni et al [12]. It consists of 23 items categorized into 4 subscales: physical functioning (8 items), emotional functioning (5 items), social functioning (5 items) and school functioning (5 items). For interpretation, three scores can be obtained: physical health (score of the physical functioning subscale); psychosocial health (combined scores of the emotional functioning, social functioning and school functioning subscales), and total score (combined score of physical health and psychosocial health).

The PedsQL™ 3.0 Cancer Module (Cancer Module) is a multidimensional instrument developed by Varni et al. to assess the impact of disease and treatment on the HRQOL of pediatric cancer patients [13]. It consists of 27 items distributed to 8 subscales: pain and hurt (2 items), nausea (5 items), procedural anxiety (3 items), treatment anxiety (3 items), worry (3 items), cognitive problems (5 items), perceived physical appearance (3 items) and communication (3 items). Both instruments were available for children aged: 5–7, 8–12 and 13–18 years old; as well as for the guardians of children aged: 2–4, 5–7, 8–12 and 13–18 years old.

The scale has five Likert response options: never, almost never, sometimes, often and almost always. To simplify the interpretation, all Likert scales were converted to 0–100. The higher scores indicate a higher HRQOL. For the versions adapted to children between the ages of 5 and 7 years, there are only three response options: never, sometimes and almost always (corresponding to 100, 50, 0). For this age, a Face Scale was used, comprising of 3 pictures of facial expressions varying from a smiling face to a very sad face to indicate no problem/difficulty/pain to a lot of problems/difficulty/pain.

The internal consistency of the Generic Scales and the Cancer Scale were measured by Cronbach's Alpha Coefficient. Values ≥ 0.70 were considered acceptable for comparisons between groups [23]. The correlation between the self-reports and the proxy-reports was assessed by interclass correlation coefficient. A comparison score between different demographic characteristics: age, gender, parental education status, SES, position of the child in the family, and medical groups of treatment (risk stratification, phase of treatment) were measured using Independent-Sample T test. All analyses were performed with two-sided tests and a value of p < 0.05 was considered significant. The SPSS for Windows (version 12) programs were used for the data analysis.

During October 2006 – January 2008, 110 ALL patients were hospitalized or visited the pediatric department of Dr. Sardjito Hospital. Ninety-eight patients were recruited according to the inclusion and exclusion criteria. There were no significant differences between the studied group and those who did not participate with regard to demographic characteristics. All guardians completed the proxy-reports questionnaire. Fifty-five patients participated in patient self-report, 3 (10%) patients of 5–7 year-old age group did not wish to answer the questionnaire. There was no significant difference between proxy-reports and self-reports with regard to age, gender, parental educational status, SES, ALL risk stratification, and phase of treatment. The average age of all patients was 6.6 years (SD = 3.7), median 6 years, range 2–16 years. Most respondents of proxy reports were mothers (n = 74, 76%), and most families (n = 76, 78%) belonged to low SES (Table 1).

Cronbach's alpha for total scale of the Generic Scales and the Cancer Scale were all above 0.7 in both self-reports and proxy-reports. A positive correlation between self-reports and proxy-reports was found on all subscales and total scale of both the Generic Scales and the Cancer Scale, except the perceived physical appearance. Physical functioning, procedural anxiety and communication subscales showed a good correlation between self-reports and proxy-reports (ICC> 0.61), but other subscales only showed a weak-moderate correlation (ICC 0.31–0.60). Guardians tended to report lower HRQOL than did patients in Cancer Scale. In both self report and proxy report, psychosocial health subscale was better than physical health subscale (table 2).

Gender, position of the child in the family, SES, parental educational status, and risk stratification did not correlate to any scales of HRQOL (table 3 and additional file 1). However, HRQOL of 2–5 year-group was significantly lower than HRQOL in the older group regarding procedural anxiety, treatment anxiety and communication subscales (p < 0.05) (table 4). The total score of the Generic Scales and the Cancer Scale in the non-intensive phase were significantly better than in the intensive-phase, in both patient self-reports as well as proxy-reports. Procedural anxiety subscale showed the greatest significant difference between the different phases. In the intensive phase, both children and guardians reported lower scores in procedural anxiety and communication. In the non-intensive phase, both children and guardians reported the lowest score in the worry subscale (table 5).