Quality of life Quality of life and health in children following ... - Nature

Bone Marrow Transplantation (2005) 36, 171–176 & 2005 Nature Publishing Group All rights reserved 0268-3369/05 $30.00

www.nature.com/bmt

Quality of life Quality of life and health in children following allogeneic SCT U Forinder1,2, C Lo¨f1 and J Winiarski1 1

Department of Paediatrics, Karolinska Institutet, Karolinska University Hospital, Huddinge, Stockholm, Sweden; and 2Department of Social Work, Stockholm University, Stockholm, Sweden

Summary: A total of 52 children, age 9 or over and at least 3 years (median ¼ 8) beyond SCT for leukaemia (n ¼ 32) or nonmalignant diseases, participated in a single-centre study of health and quality of life (QoL). QoL and self-esteem were assessed with SCHQ-CF87, a generic multidimensional self-report instrument, and with ‘I think I am’. As a group, the children had good QoL, but were below norm in the bodily pain (Po0.05), general health and self-esteem dimensions (Po0.01). Lansky or Karnofsky function levels were at a median of 90. Sense of coherence (SOC-13) was normal and correlated with SCHQ-CF87. Most children were subjectively and objectively in good health according to a self-assessment symptom inventory or by a medical record-based scoring of late effects, although pain was commonly reported. A total of 25% of the patients were rated as having moderate to severe late effects, without considering cataracts or infertility. Neither age at SCT, gender, malignant vs nonmalignant disease, nor stature influenced QoL significantly. Children with moderate to severe chronic graft-versus-host disease or cognitive deficits had lower QoL in some dimensions. No correlation was, however, found between the physician-rated total late effects score and overall QoL. Contrarily, QoL was clearly related to the degree of self-rated symptoms. Bone Marrow Transplantation (2005) 36, 171–176. doi:10.1038/sj.bmt.1705021 Published online 6 June 2005 Keywords: allogeneic; children; health; late effects; stem cell transplantation; quality of life

Stem cell transplantation (SCT), entails major strains on both children and their families. The 60–70%1 of children who make up the group of long-term survivors also have a number of late effects to contend with. These include both somatic and psychological effects that impact on the lives of these families.2 The objective of this study was to shed

light on these consequences for the quality of life (QoL) and health of long-term surviving children. The multidimensional concept of QoL is, in medicine, often used synonymously with health-related quality of life (HRQoL).3 HRQoL researchers generally subscribe to the WHO’s definition of health as a state of complete physical, mental and social well-being rather than merely the absence of disease or infirmity.4 In Syrjala’s5 recent overview of OoL studies on SCT patients, mainly adults, she found that they generally rate their QoL as high, although certain problem areas could be identified. She also observed that not much is known about QoL among children after SCT.5 In 1994, Andrykowski6 noted that it was adults who had been the focus of post-SCT QoL research and that increased attention should be devoted to children. This observation still remains true.5 The aim of this study was thus to explore aspects of QoL and health in children and adolescents following SCT. It employs instruments whereby the children themselves rate their QoL. To gain insight into the health situation, we used both a self-assessment form and a rating based on the patients’ medical records. This article thus focuses on the results from the children’s point of view.

Patients and methods During 2003, 54 children aged 9 or older for whom at least 3 years had elapsed since SCT and who visited Huddinge Hospital for their yearly post transplant follow-up were asked to participate in the study. We chose the 9 years age limit to ascertain that the participants would be able to independently respond to the instruments. Children with severe neurodegenerative disorders that were manifest prior to SCT were not included. Two children declined to take part. Leukaemia and myelodysplastic syndrome were the dominant indications for SCT, accounting for a total of 62%. Patient characteristics are presented in Table 1. Details concerning SCT protocols have been described elsewhere.1 Approval for this study was obtained from the Human Research Ethics Committee of the Karolinska Institutet.

Instruments used to rate health and disability Correspondence: Dr U Forinder, Department of Social Work, Stockholm University, 106 91 Stockholm, Sweden; E-mail: [email protected] Received 5 November 2004; accepted 12 April 2005 Published online 6 June 2005

Late effects score. The medical records of each patient were reviewed in detail and the existence and severity of late effects in each of eight predefined problem categories was assessed by a physician (JW) who knew all patients but was

QoL in children following SCT U Forinder et al

172

unaware of the outcome of the QoL studies. The following eight categories were defined: neurological symptoms (mainly seizures), cognitive problems, chronic graft-versushost disease (GVHD), pulmonary function based on symptoms and spirometry, problems involving appearance and alopecia, skeletal pain/osteonecrosis, growth impairment and renal dysfunction (subnormal GFR). Infertility and cataracts were not included in the assessment. The scale used, specified in detail for each category, was briefly as follows: no symptoms ¼ 0; minor, infrequent symptoms ¼ 1; moderately severe symptoms ¼ 2; very severe, uncontrolled or debilitating disease ¼ 3. Cognitive function was scored according to the level of supportive measures needed in school and psychologist’s reports. For example, a score of 1 meant that the child had minor school aptitude problems, while a score of 2 indicated the need of special school. Growth impairment was graded a score of 1 when stature was p1 s.d., according to growth charts, a score of 2 for p–2 s.d. and 3 for p–3 s.d. Uncomplicated well-controlled epilepsy was graded a score of 1, while more than one seizure yearly corresponded to a score of 2 in the Table 1

neurological symptoms category. A GFR at 60–80% of normal received a score of 1. A total late effects score was calculated for each patient as the sum of the eight category scores (see Table 2). Subjective health and symptom inventory checklist. A selfassessment checklist was designed with 13 items subdivided into four blocks (symptoms, pain, activity, school). The items requested children to grade the severity and frequency of symptoms on a scale of 0–3, a higher score indicating better health. The mean score for the items in each block was calculated. The activity block focused on participation in physical exercise both in and out of school and fatigability. The school block asked about cognitive problems, school aptitude and bullying. The pain block included items relating to pain levels, while the subjective symptom block included items involving chronic GVHD, as well as problems relating to late effects, such as dryness of skin, mucous membranes and eyes, alopecia, nausea, gastrointestinal problems, sleep disturbances, etc. Lansky and Karnofsky scales. The Karnofsky scale7 was used in the age group 16 and over. The scale measures the level of patient activity and medical care requirements. The Lansky scale8 is a similar instrument based on the level of play activity and was used below the age of 16. Both scales range from 10 (moribund) to 100 (normal function). A paediatrician assessed the scores in consultation with the patient or family.

Patient characteristics

Total number of patients, n Gender F/M, n (%)

52 23 (44%)/29 (56%) 9–22 (15) 16/14

Age at interview (years), range (mean) Mean age at interview according to gender (years), F/M Age at SCT (years), range (mean) Years since SCT, range (mean) Indications for SCT Malignant disorders, total, n (%) ALL/AML/MDS/CML, n Nonmalignant disorders, total, n (%) SAA/FA/Thal/CGD/Gaucher/ALD/HLH,a n

1–16 (7) 3–19 (8)

Sense of coherence. Sense of coherence (SOC) is a measure of a global tendency to view life situations as comprehensible, manageable and meaningful.9 In adults,9 and possibly in adolescents,10 a high SOC score may function as a stress moderator. We used SOC-13, a shortened version that has proved reliable.11,12 In children under the age of 10, the same dimensions of SOC were assessed by the version ‘How do I feel?’13

32 (62%) 18/8/5/1 20 (38%) 7/4/3/2/2/1/1

a

SAA ¼ severe aplastic anaemia; FA ¼ Fanconi anaemia; Thal ¼ thalassaemia major; CGD ¼ chronic granulomatous disease; ALD ¼ adrenoleukodystrophy; HLH ¼ haemophagocytic lymphohistiocytosis; n ¼ number of patients.

Table 2

Late effects score

Late effects scorea according to category

0 (none)

Neurological symptoms, n (%) Cognitive problems, n GVDH, n Impaired pulmonary function, n Skeletal pain, osteonecrosis, n Appearance, alopecia, n Growth impairment, n Renal function, n

48 41 45 38 44 44 25 51

Total late effects scoreb Patients, n (%) Total late effects score, mean and median a

(92%) (79%) (87%) (73%) (85%) (85%) (48%) (98%)

1 (minor) 3 6 2 5 8 3 9 1

2 (moderate)

(6%) (12%) (4%) (10%) (15%) (6%) (17%) (2%)

0

1

2

12 (23%)

7 (13%)

9 (17%)

1 5 2 7 0 4 10 0

(2%) (10%) (4%) (13%) (0%) (8%) (19%) (0%)

3 (severe) 0 0 3 2 0 1 8 0

(0%) (0%) (6%) (4%) (0%) (2%) (15%) (0%)

3

4–10

11 (21%)

13 (25%)

2/2

Late effects score: Based on the medical records of each patient, the presence and severity of predefined late effects in each of eight categories were assessed on a scale of 0–3 (0 ¼ no symptoms, 1 ¼ minor symptoms, 2 ¼ moderate symptoms, 3 ¼ very severe or debilitating disease). An individual total late effects score was calculated as the sum of the category scores. The number (n) and percentage of patients (%) at each of five total score levels are described, where a total score of 4–10 represents either several or more severe sequelae. b

Bone Marrow Transplantation


173

Instruments for measuring QoL SCHQ-CF87. The Swedish version of the Child Health Questionnaire (SCHQ-CF)14 is an adapted version of a North American, crosscultural, self-administered instrument for assessing the physical and psychosocial health status of children. It was developed in the US for children aged 5–18 years and is used as a proxy-report and a selfreport. The self-report contains 12 multi-item scales and two single-item scales. A higher score indicates better QoL. The instrument has been validated in many countries,15–17 including Sweden.18,19 The scales in SCHQ-CF87 have good internal consistency.19 I think I am. ‘I think I am’ (ITIA)20 is a self-report scale standardized on a sample of Swedish children, used in both research21,22 and clinical contexts for assessing selfesteem in children and adolescents.20–22 The scale includes descriptions of a child’s conceptions of himself/herself, a higher score reflecting higher self-esteem. There are two versions for the age groups 7–10 and 11–16. Correlations have been demonstrated between ITIA scores and mental health and peer popularity while internal consistency is adequate.20 Procedure. Patients and their parents were informed about the study by letter well ahead of their visits. The symptom inventory checklist was also sent to them for completion at home prior to their visit. During the visit, both patient and parents (when present) were informed together, and the parents were then asked to sit in another room so as to avoid situations in which patients and parents influenced each other. Comparison groups and norm values. As comparison groups we used, for SCHQ-CF87, the US norm values,14 as well as the study group from Norrby et al’s19 report of Swedish children with juvenile chronic arthritis (JCA). The US norm group included 232 children aged 5–18 years. Norrby et al’s study group consisted of 199 children 9–16 years of age (boys 47%/girls 53%). For ITIA and SOC, we used the study and control groups from Ra¨ty et al,12 who investigated 158 adolescents (male 44%/female 56%) with uncomplicated epilepsy in the 13–22 age group and a random sample of 282 Swedish adolescents matched for gender, age and residential areas. For the youngest SCT children, we used norm group values for the instruments.13,20

Statistical analysis The data were analysed using the Statistical Package for Social Sciences (SPSS version 12.0 for Windows). Unpaired t-test analysis was performed to explore any statistically significant differences between the SCT group and the historic comparison groups,12–14,19 as well as to investigate gender- and age-related differences. The Pearson’s product moment correlation coefficient was used to assess construct validity by correlating SCHQ and ITIA with SOC-13. Intercorrelations between health variables and the QoL instrument variables were also investigated, as was the

correlations with age at the time of SCT. Cronbach’s alpha was used to determine the internal consistency of the instruments. One problem is that of multiple comparisons, increasing the risk of getting significant results by chance. Thus, it cannot be precluded that some differences observed could be due to chance.

Results Health factors Late effects. As shown in Table 2, relatively few children suffered severe sequelae. The median total score was 2, indicating that the average patient had either one moderately severe problem or two minor problems in any of the eight categories. A total of 54% of the children had total scores of 2 or less. A total of 13 patients (25%) had total scores ranging from 4 to 10, signifying more pronounced sequelae. The most prevalent late effect was short stature, defined as stature 1 s.d. or below (n ¼ 27, score ¼ 1). Eight patients were –3 s.d. or shorter (score ¼ 3). Although cognitive problems were noted in 11 children, none had dysfunctions severe enough (score 3) to hinder school attendance (including special school), social activities or work. Seven patients had ongoing GVHD-related symptoms 3–9 years after their transplants. Two of the eight patients presently reporting minor skeletal pains had a previous history of severe osteonecrosis, which had required the use of a wheelchair and reconstructive orthopaedic surgery. Eight children had extensive alopecia (all had received busulphan conditioning) or disfiguring dermal symptoms due to GVHD. There were no secondary malignancies. Subjective health and symptom inventory checklist. A total of 48 patients returned completed checklists. The inventory checklist had a good internal consistency, with a Cronbach’s alpha of 0.90. The median scores calculated for the items in each of the four blocks, except the pain block, were between 2.3 and 2.6, indicating few problems (score 3 ¼ well, no symptoms). While the symptoms block scores had a range of 1.9–2.9 (median 2.6), the pain block revealed that pain-related problems are common. It contained lower scores than any of the other blocks (range 1–3, median ¼ 2.0). According to the school block scores (range 1.5–3, median 2.4), most children experienced no problems in this area, or only minor ones, although there are three outliers with scores ranging 1.5–1.75 that represent considerable difficulties. Lansky/Karnofsky scales. All the older teenagers (416 years, n ¼ 24) had Karnofsky scores of 80 or above (median ¼ 90), except two who scored 70. The younger children (n ¼ 28) had Lansky scores of 90 or above in all but five cases, scoring 80 (median ¼ 90). SOC. The scores obtained on SOC-13 for adolescents (aged 13–22), and on the version for younger children (aged 9–12), ‘How do I feel?’, showed that children in the SCT group have an SOC well on par with that of both the norm Bone Marrow Transplantation


174

groups and other chronically ill children. The mean value for SOC-13 in the SCT group (n ¼ 33, M ¼ 64.0, s.d. ¼ 13.8, Cronbach’s alpha ¼ 0.87) was on par with that of the healthy norm group (n ¼ 282, M ¼ 63.7, s.d. ¼ 12.07) and with that of patients with uncomplicated epilepsy (n ¼ 158, M ¼ 64.05, s.d. ¼ 13.05).12 The mean value for the younger children in the SCT group (n ¼ 7, M ¼ 48.9, s.d. ¼ 2.3) was on par with that of the norm group of children aged 9 (M ¼ 48.7, s.d. ¼ 5.7).13 Correlation between health variables. There was a positive correlation (0.48, Po0.05) between SOC-13 and the symptom inventory checklist as a whole, as well as with its activity (0.41, Po0.05) and symptoms (0.59, Po0.01) subscales, but not with the pain and school subscales. Moreover, SOC-13 and the total late effects score showed no correlation. For the youngest children (n ¼ 8) using ‘How do I feel?’, there was a positive correlation between ‘How do I feel?’ and the ‘Pain’ block (r ¼ 0.97, Po0.05), as well as a strong correlation with the school subdivision (r ¼ 0.96, Po0.01). No correlation was found with the other item blocks of the inventory checklist.

The QoL instruments – SCHQ-CF and ITIA The scores of the study group (SCT) differ significantly on some of the SCHQ-CF scales from those of the comparison groups (Table 3). The SCT group scored higher than Norrby et al’s JCA group on the subscale role socially due to physical difficulties (RP), but lower on the general health (GH) scale. Compared to the US norm sample, the SCT group scored higher on the general behaviour (BE) subscale, with a similar trend on the role and behavioural difficulties (RB) subscale, but lower on the bodily pain, selfesteem and general health subscales. Cronbach’s alpha ranged from 0.75 to 0.91. The subscales met or exceeded the minimum standard for group level analysis (0.70) and the values are comparable to those obtained from the US normal sample.14 For ITIA, there were no significant

Table 3

differences between the SCT and the comparison group scores. Cronbach’s alpha was 0.88 for the total score and 0.63–0.84 for the subscales. ITIA 7–10 years version. The small SCT sample (n ¼ 4) using the children’s form of ITIA rated their self-esteem slightly but not significantly higher (M ¼ 20.5, s.d. ¼ 9.8) than the comparative norm group.20

Relationship between health and QoL Late effects and QoL. There was no correlation between the QoL variables (ITIA, SCHQ-CF87) and the total late effects score. The groups suffering from late effects in the form of moderately severe cognitive impairments or moderate to extensive chronic GVHD were too small for statistical analyses to be relevant. We therefore visually compared the values of these children with the means obtained on ITIA and SCHQ-CF87 for the entire SCT group. Briefly, most of the children with moderately severe cognitive impairment (n ¼ 5) had lower values on the psychosocial subscales of SCHQ-CF87 than the average of the SCT group as a whole. The average levels for the five cognitively impaired patients in, for example, the RE, RB and BE subscales were 67, 74 and 66 compared to 86, 92 and 82, respectively, for the SCT group as a whole. The children suffering from moderate to severe GVHD (n ¼ 5) had lower values on the GH subscale (an average of 32.5 compared to 53.8 for the whole SCT group) and self-esteem (50.5 compared to 73.2). In other respects, they had high values on the remaining subscales. Both the GVHD and the cognitive impairment groups reported lower values than the whole SCT group using the ITIA instrument, which measures self-esteem. QoL values were not affected by growth impairment. Subjective health and symptom inventory checklist and QoL. There was a positive correlation between ITIA and the symptoms block of the symptom inventory

QoL as assessed by SCHQ-CF87

a

SCHQ Subscale PF RE RB RP BP BE MH SE GH CH

SCT sample (1) n ¼ 51 Mean7s.d.

JCA sample (2) n ¼ 45 Mean7s.d.

(1)–(2) t-Value

US norm (3) n ¼ 232 Mean7s.d.

(1)–(3) t-Value

87.1712.9 86.2722.7 92.3714.8 90.6716.2 66.1728.2 81.8713.7 73.4714.9 73.2717.6 53.8716.1 3.871.0

80.3724.2 87.3723.3 90.4718.4 81.0727.3 61.4729.0 80.6713.1 75.3716.3 76.5718.2 62.1720.8 3.171.3

1.75 0.24 0.24 2.10* 0.80 0.44 0.59 0.88 2.12* 0.50

88.8713.9 85.9721.0 86.5721.5 88.3721.0 74.4723.1 76.6714.6 72.7716.0 81.8715.8 66.4714.6 —

0.79 0.09 1.83 0.74 2.23* 2.33* 0.28 3.45** 5.61** —

a Scale: SCHQ-CF87:14 PF ¼ physical functioning, RE ¼ limitations in role and social behaviour due to emotional difficulties, RB ¼ limitations in role socially due to behavioural difficulties, RP ¼ limitations in role socially due to physical difficulties, BP ¼ bodily pain, BE ¼ general behaviour, MH ¼ mental health, SE ¼ self-esteem, GH ¼ general health, CH ¼ change in health. *Po 0.05, **Po 0.01. (2) Norrby,19 (3) Landgraf.14 Mean7s.d. for the scores obtained in the SCT study group (1) and in the historic comparison samples of children with chronic disease, JCA (2) and in the US norm group (3). t-Tests were performed on the SCT sample (1) and each of the two comparison groups (2 and 3).



175 Table 4 Intercorrelation (Pearson’s product moment correlation coefficient) between the various subscales of the children’s form, SCHQ-CF87, and the total scale scores on SOC-13 (n ¼ 33) and ITIA (n ¼ 49) SCHQ

SOC-13

ITIA

PF RE RP RB BP BE MH SE GH FA FC SOC-13

0.47** 0.63** 0.58** 0.46** 0.41* 0.72** 0.73** 0.62** 0.70** 0.39* 0.40*

0.14 0.13 0.14 0.13 0.34* 0.36* 0.35* 0.44** 0.35* 0.01 0.31* 0.34

**Po0.01, *Po0.05. Scales: SCHQ-CF87:14 PF ¼ physical functioning, RE ¼ limitations in role socially due to emotional difficulties, RB ¼ limitations in role socially due to behavioural difficulties, RP ¼ limitations in role socially due to physical difficulties, BP ¼ bodily pain, BE ¼ general behaviour, MH ¼ mental health, SE ¼ self-esteem, GH ¼ general health, CH ¼ change in health, FA ¼ family activity, FC ¼ family cohesion. ITIA:20 version 11–16 y for adolescents; SOC-13:11 short version of sense of coherence.

checklist (r ¼ 0.48, n ¼ 32, Po0.01), with a higher selfesteem score associated with fewer subjective symptoms (r ¼ 0.36, Po0.05) and higher participation in physical activities (r ¼ 0.51, Po0.01). All four blocks, particularly the symptoms and activity blocks, correlated positively with different subscales of SCHQ-CF87. The strongest correlation was between the symptoms block and subscales general behaviour (BE; r ¼ 0.58, Po0.01), mental health (MH; r ¼ 0.52, Po0.01) and between the activity block and self-esteem (SE; r ¼ 0.51, Po0.01). The pain (r ¼ 0.36, Po0.01) and school blocks (r ¼ 0.33, Po0.05) correlated with mental health. SOC and QoL. There were positive correlations between SOC-13 and all SCHQ-CF87 subscales and likewise between ITIA and the SCHQ subscales bodily pain, mental health, self-esteem, general health and family cohesion. No correlation was found between SOC-13 and ITIA (Table 4). Gender, age and primary disease vs health and QoL. There was no gender or age effect in any of the instruments, with the exception of the subscale self-esteem (SE) that correlated negatively with age at the time of the study in SCHQ (r ¼ 0.33, n ¼ 46, Po0.05). Overall, patients with malignant disease did not report significantly different QoL than those with nonmalignant disorders.

Discussion The study shows that the majority of young SCT patients felt their health and QoL to be good. Normal SOC scores, high Karnofsky and Lansky functional scores and the circumstance that 75% of the patients have no or limited physician-rated late effect-related symptoms further sup-

port the notion that most paediatric SCT survivors have a good health-related QoL. Pain symptoms were common, however. Interestingly, there was no overall association between the physician-rated total late effects score and QoL, while patients who themselves had reported many symptoms also scored lower in terms of QoL. This indicates the specific importance to QoL of the patient’s subjective perception of his or her health. It is important to consider the various dimensions of QoL as well. In areas that rate bodily pain (BP), general health (GH) and self-esteem (SE), the SCT group had lower QoL scores than the US norm group, as they did on GH compared with Swedish JCA children. One possible explanation is that, unlike the comparison groups, these children have suffered a life-threatening disease. In the areas of general behaviour (BE) and role behavioural difficulties (RB), the SCT patients scored higher than the US norm and on par with the JCA group. In areas that rate limitation in role socially due to physical difficulties (RP), the SCT group scored higher than the JCA comparison group. Ideally, cancer patients or organ transplant patients would be preferable for comparison,6 but as there were no such Swedish samples studied with these instruments, we were limited to the JCA group. Another limitation relates to the cross-sectional study design and the size of our study group. The sample does, however, consist of consecutive, unselected patients, and is representative of a paediatric SCT population. Being younger than 3 at the time of SCT has been mentioned as a risk factor for developing psychosocial difficulties, due to greater cognitive impairment.23 Our study was limited due to the small number of very young children it includes, and it failed to corroborate the existence of a significant relationship between low QoL and low age at SCT. On the other hand, when we look at the five individuals with moderately severe cognitive impairment, they do form a group that had lower QoL scores, particularly on the psychosocial subscales. Moreover, four of them were aged around 1 year at the time of SCT. The group with chronic GVHD also had lower QoL scores, particularly on the scales rating self-esteem and general health. Both the GVHD and the cognitive impairment samples were too small to support statistically valid conclusions, and these limited observations, although suggestive, need confirmation in future studies. There is also a negative association between QoL and increasing age at the time of the QoL assessment, in that self-esteem tends to decline with age. Problems in terms of body image can also lead to lower QoL.24 Short stature has been considered to cause lower QoL, although this was not confirmed either by our study or in a recent review.25 There are several studies showing that surviving childhood cancer not necessarily is associated with serious psychological or social problems for the majority of patients, although there is an important minority (15– 30%) who experience ongoing psychological and/or social difficulties.26,27 With the exceptions of the children scoring a low SOC or severely affected by late effects, that is, cognitive impairment and, in line with experience from adult SCT,5 chronic GVHD, Bone Marrow Transplantation


176

this study does not point to other consistent risk factors from the QoL perspective, such as age at transplant, gender, stature or primary disease, while there are areas in which these patients require support. An obvious clinical conclusion might thus be to focus more on psychosocial support on the group with apparent late-effect-related problems and a low SOC. Moreover, we should not only consider the risk factors in the lives of these patients, but also the role of protective factors, such as family cohesion and SOC.28 A further analysis might provide an answer to the question of whether this is true of the adolescents in our study.

Acknowledgements We are grateful to Professor Stig Elofsson, Department of Social Work, Stockholm University, for his valuable advice regarding the statistics. The study was supported by the Swedish Children’s Cancer Foundation and The Mayflower Foundation.

References 1 Gustafsson A˚, Remberger M, Winiarski J, Ringdeń O. Unrelated bone marrow transplantation in children, outcome and a comparison with sibling donor grafting. Bone Marrow Transplant 2000; 25: 1059–1065. 2 Forinder U. Bone marrow transplantation from a parental perspective. J Child Health 2004; 8: 134–148. 3 Lindstrom B. Quality of life: a model for evaluating health for all. Conceptual considerations. Soz Pra¨ventivmed 1992; 37: 301–306. 4 WHO. Constitution of the World Health Organisation 1948. WHO: Geneva. 5 Syrjala K. Assessment of Quality of Life in hematopoietic cell transplantation recipients. In: Blume K, Forman S, Appelbaum F (eds). Thomas’ Hematopoietic Cell Transplantation. Blackwell Science: Malden, 2004, pp 507–518. 6 Andrykowski MA. Review of psychosocial factors in bone marrow transplantation. Bone Marrow Transplant 1994; 13: 357–375. 7 Karnofsky DA, Burchenal JH. The clinical evaluation of chemotherapeutics in cancer. In: MacLeod CM (ed.). Evaluation of Chemotherapeutic Agents. Columbia University Press: New York, 1949, pp 191–205. 8 Lansky SB, List MA, Lansky LL. The measurement of performance in childhood cancer patients. Cancer 1987; 60: 1651–1656. 9 Antonovsky A. Unravelling the Mystery of Health. How People Manage Stress and Stay Well. Josey Bass Publ: San Francisco, 2001. 10 Torsheim T, Aaroe LE, Wold B. Sense of coherence and school-related stress as predictors of subjective health complaints in early adolescence: interactive indirect or direct relationship? Soc Sci Med 2001; 53: 603–641.


11 Hansson K, Olsson M. Ka¨nsla av sammanhang – ett ma¨nskligt stra¨vande. Nord Psykol 2001; 53: 238–255. 12 Ra¨ty LK, Wilde Larsson BM, Soderfeldt BA. Health-related quality of life in youth: a comparison between adolescent and young adults with uncomplicated epilepsy and healthy controls. J Adolesc Health 2003; 33: 252–258. 13 Margalit M. Loneliness and coherence among preschool children and children with learning disabilities. J Learn Disabil 1998; 331: 173–180. 14 Landgraf JM, Ware JE. The CHQ User’s Manual, 2nd printing. The Health Institute, New England Medical Center: Boston, 1996. 15 Selvaag AM, Ruperto N, Asplin L. The Norwegian version of the Child Health Assessment Questionnaire and the Health Questionnaire. Clin Exp Rheumatol 2001; 19 (4 Suppl. 23): S116–S120. 16 Raat H, Landgraf JM, Bonsel GJ. Reliability and validity of child health questionnaire-child form (CHQ-CF87) in Dutch adolescent population. Qual Life Res 2002; 11: 575–581. 17 Waters E, Salmon L, Wake M. The health and well-being of adolescents: a school-based population study of the self-report Child Health Questionnaire. J Adolesc Health 2001; 29: 140–148. 18 Andersson-Ga¨re B, Ruperto N, Berg S. The Swedish version of the Child Health Assessment Questionnaire (CHAQ) and the Child Health Questionnaire (CHQ). Clin Exp Rheumatol 2001; 19 (4 Suppl. 23): S146–S150. 19 Norrby U, Nordholm L, Fasth A. Reliability and validity of the Swedish version of Child Health Questionnaire. Scand J Rheumatol 2003; 32: 101–107. 20 Ouvinen-Birgerstam P. Jag tycker jag a¨r (I Think I am) Manual. Psykologifo¨rlaget: Stockholm, 1999. 21 Arvidson J, Larsson B, Lo¨nnerholm G. A long-term follow-up study of psychosocial functioning after autologous bone marrow transplantation in childhood. Psychooncology 1999; 8: 123–134. 22 Essen v L, Enska¨r K, Kreuger A. Self-esteem, depression and anxiety among Swedish children and adolescents on and off cancer treatment. Acta Paediatr 2000; 89: 229–236. 23 Smedler AC, Nilsson C, Bolme P. Total body irradiation: a neuropsychological risk factor in paediatric bone marrow transplant recipients. Acta Paediatr 1995; 84: 325–330. 24 Madan-Swain A, Brown RT, Sexon SB. Adolescent cancer survivors, psychosocial and family adaptation. Psychosomatics 1994; 35: 453–459. 25 Sandberg DE, Voss LD. The psychosocial consequences of short stature: a review of the evidence. Best Pract Res Clin Endocrinol Metab 2002; 16: 449–463. 26 Calminus G, Kiebert G. Studies on health-related quality of life in childhood cancer in the European setting. An overview. Int J Cancer 1999: (Suppl. 12): 83–86. 27 Zebrack BJ, Chesler MA. Quality of life in childhood cancer survivors. Psychooncology 2002; 11: 132–141. 28 Phipps S, Mulhern RK. Family cohesion and expressiveness promote resilience to the stress of pediatric bone marrow transplant: a preliminary report. J Dev Behav Pediatr 1995; 16: 257–263.