KINDL

English Children and Adolescents Revised Version

Questionnaire for Measuring Health-Related Quality of Life in

Manual

Ulrike Ravens-Sieberer & Monika Bullinger, 2000

Contents

Contents
Contents 1 Information about the KINDLR Questionnaire 1.1 Structure of the Questionnaire 1.1.1 Self-assessment and external assessment by age groups 1.1.2 Structure of the scales 1.1.3 Disease module 1.1.4 Kiddy-KINDLR interview version for children and additional questions for parents Structure of the Sub-Scales and Classification of Items 1.2.1 Self-report versions 1.2.2 Parents versions Psychometrics 1.3.1 Structure of sub-scales, reliability and factorial validity 1.3.2 Convergent validity 1.3.3 Discriminant validity 1.3.4 Sensitivity Execution Interpretation and Reference Values General Remarks on the Analysis Data Entry Recoding Items 2.3.1 Responses outside the valid range 2.3.2 Reversing items Dealing with Missing Data Calculation of Sub-Scale Scores 2.5.1 Kid-KINDLR and Kiddo-KINDLR 2.5.2 Formulae and examples for calculating sub-scale sum scores 2.5.3 Kiddy-KINDLR Contact 18 2 1 2 2 3 3 3 4 4 5 7 8 9 9 10 11 12 14 14 15 15 15 15 16 16 16 17 17 19 20

1.2 1.3

1.4 1.5 2.1 2.2 2.3 2.4 2.5

2 Analysing the KINDLR Questionnaire

2.6

3 Disc for Data Analysis 4 Further Reading 5 Appendix 21

Information about the KINDLR

1 Information about the KINDLR Questionnaire

While most quality of life measures for children have been developed in the English language and then translated in a further, methodologically elaborate step (Ravens-Sieberer and Cieza, 2000), the generic KINDLR Questionnaire for Measuring Health-Related Quality of Life in Children and Adolescents represents a German-language measure (originally developed by Bullinger et al. 1994, revised by Ravens-Sieberer & Bullinger 1998a, 1998b), for use in clinical populations but also with healthy children and adolescents. The KINDLR Questionnaire for Children and Adolescents described below was developed and tested with a view to remedying the discrepancy between the urgency of the subject Quality of Life in Children and Adolescents and the lack appropriate measures (Ravens-Sieberer, 2000). In developing it, the goal was to design a short, methodologically suitable and flexible set of instruments which could be completed both by children/adolescents and by their parents, which was available for different age groups and stages of development, which could be used for healthy and ill children (generic approach), which could be extended by means of specific modules (e.g. for different classes of diseases) and which could be used in different types of study, namely: a) epidemiological studies on the situation of children and adolescents in the Federal Republic of Germany, b) clinical studies dealing with the effects of therapeutic measures on the quality of life of acutely and chronically ill children, and c) in rehabilitation, looking into the effects of rehabilitation programmes and in each case from the perspective of the children and of their parents. The questionnaire has so far be used and tested in a number of studies over a period of up to three years involving over 3000 healthy and chronically ill children as well as their parents. The psychometric results reveal a high degree of reliability (Cronbachs .70 for most of the sub-scales and samples) and a satisfactory convergent validity of the procedure, beyond which the acceptance of the measure by children and adolescents is high (Ravens-Sieberer, 1998). The questionnaire has been able to distinguish between children with different physical disorders and under different types of strain. Overall, the KINDLR has proved to be a flexible, modular, psychometrically acceptable method of measuring quality of life in children by means of a central module covering generic aspects in childrens quality of life while also measuring the specific burdens associated with diseases in childhood via additional modules. Age-specific versions take into account the changes in the quality of life dimensions in the course of child development.

Ravens-Sieberer & Bullinger / 2000 / Page 1

Information about the KINDLR

1.1 Structure of the Questionnaire

The KINDLR questionnaire satisfies the demand for taking into account progress during child development and the principle of patient-generated data collection by providing different versions of the questionnaire for different age groups and both a self-report version and a proxy version. The common practice of modifying a measure originally designed for adults to make it suitable for children, was avoided here. With the KINDLR questionnaire an original German-language measure is available that was specifically designed and validated for children. The psychometric testing of the KINDLR shows that a procedure has been developed that can be used in epidemiological, clinical and rehabilitation research studies; the use of the KINDLR questionnaire in the context of health care planning is increasingly under discussion.

1.1.1 Self-assessment and external assessment by age groups

Three versions of the KINDLR questionnaire are available as self-report measures for different age groups:
Kiddy-KINDLR for children aged 4 to 7
Kiddy-KINDL Interview
R

Kid-KINDLR for children aged 8 to 12

Kid-KINDL
R

Kiddo-KINDLR
for adolescents aged 13 to 16
Kiddo-KINDL
R

In addition, the questionnaire is available in two proxy versions for parents (4 - 7-year-olds and 8 - 16-year-olds): Kiddy-KINDLR for Parents of children aged 4 to 7
4 - 7 years Paretns Questionnaire KindlR

KINDLR for Parents of children and adolescents

8 - 16 years Paretns Questionnaire KindlR

Furthermore, a short form of the KINDLR (12 items) has been developed, as well as a series of disease-specific modules (obesity, asthma, atopic dermatitis and diabetes). The questionnaire is also available in the languages English, French, Dutch, Russian, Turkish, Italian and Spanish. A Computer-Assisted Touch Screen version (CAT-Screen) of the questionnaire is available as well.

Ravens-Sieberer & Bullinger / 2000 / Page 2

Information about the KINDLR

1.1.2 Structure of the scales

The KINDLR questionnaire consists of 24 Likert-scaled items associated with six dimensions:
physical well-being, emotional well-being, self-esteem, family, friends and everyday functioning (school or nursery school/kindergarten). The sub-scales of these six dimensions can be combined to produce a total score.
Figure 1: Structure of the Sub-Scales Included in the KINDL Questionnaire
Physical 4 Items Emotional 4 Items
R

ITEMS

Self-esteem 4 Items Family 4 Items Friends 4 Items School 4 Items Total 24 Items

1.1.3 Disease module

All versions of the KINDLR contain an additional sub-scale entitled Disease, whose items can be completed in case of prolonged illness or hospitalisation. The additional sub-scale consists of a filter question and six items which measure the childs quality of life with respect to his or her illness. In addition, disease-specific modules are available for the illnesses obesity, bronchial asthma, atopic dermatitis and diabetes.

1.1.4 Kiddy-KINDLR interview version for children and additional questions for parents
On account of the particular difficulties associated with interviewing young children, the structure of the Kiddy-KINDLR differs from that of the other questionnaires (Kid/Kiddo). In the self-report version, it only consists of twelve items, two for each dimension. This means that no sub-scale scores can be calculated for the individual dimensions but only a total score. The additional questions on Disease are, on the other hand, included in full. The response categories of the Kiddy-KINDLR cover 3 levels (1 = never, 2 = sometimes, 3 = very often), the children are to be questioned in a face-to-face interview. The parents version of the Kiddy-KINDLR with its 24 items in 6 dimensions corresponds in structure to the parents version of the KINDLR for 8 to 16-year-old children and teenagers. However, in order to make up for the potentially lower information content of the self Ravens-Sieberer & Bullinger / 2000 / Page 3

Information about the KINDLR

reported responses by young children, the parents version of the Kiddy-KINDLR contains a further 22 items which can be treated as a sub-scale in their own right.

1.2 Structure of the Sub-Scales and Classification of Items

1.2.1 Self-report versions
Kiddy-KINDLR (4 to 7-year-olds) Childrens Version (Interview)
1. ... I felt ill 2. ... I had a headache or tummy-ache

Kid-KINDLR (8 to 12-year-olds) Childrens Version

Kiddo-KINDLR (13 to 16-year-olds) Teenagers Version

1. ... I felt ill 2. ... I was in pain 3. ... I was tired and worn-out 4. ... I felt strong and full of energy 5. ... I had fun and laughed a lot 6. ... I was bored 7. ... I felt alone 8. ... I felt scared or unsure of myself 9. ... I was proud of myself 10. ... I felt on top of the world 11. ... I felt pleased with myself 12. ... I had lots of good ideas 13. ... I got on well with my parents 14. ... I felt fine at home 15. ... We quarrelled at home 16. ... I felt restricted by my parents

Physical Well-Being
1. ... I felt ill 2. ... I had a headache or tummy-ache 3. ... I was tired and worn-out 4. ... I felt strong and full of energy

Emotional Well-Being
3. ... I had fun and laughed a lot 4. ... I was bored 5. ... I had fun and laughed a lot 6. ... I was bored 7. ... I felt alone 8. ... I was scared

Self-Esteem
5. ... I was proud of myself 6. ... I felt pleased with myself 9. ... I was proud of myself 10. ... I felt on top of the world 11. ... I felt pleased with myself 12. ... I had lots of good ideas

Family
7. ... I got on well with my parents 8. ... I felt fine at home 13. ... I got on well with my parents 14. ... I felt fine at home 15. ... We quarrelled at home 16. ... My parents stopped me from doing certain things

Friends
9. ... I played with friends 10. ... I got along well with my friends 17. ... I played with friends 18. ... Other kids liked me 19. ... I got along well with my friends 20. ... I felt different from other children 11. ... I coped well with the assignments set in nursery school/kindergarten 12. ... I enjoyed nursery school/ kindergarten 21. ... doing my schoolwork was easy 17. ... I did things together with my friends 18. ... I was a "success" with my friends 19. ... I got along well with my friends 20. ... I felt different from other people

Everyday Functioning (School or Nursery School/Kindergarten)

21. ... doing the schoolwork was easy

22. ... I enjoyed my lessons 23. ... I looked forward to the weeks ahead 24. ... I was afraid of bad marks or grades

22. ... I found school interesting 23. ... I worried about my future 24. ... I worried about getting bad marks or grades

Ravens-Sieberer & Bullinger / 2000 / Page 4

Information about the KINDLR

Kiddy-KINDLR (4 to 7-year-olds) Childrens Version

13. Are you staying in hospital just now or do you have some long-term illness? (Filter question) 14. ... I was afraid that my illness might get worse 15. ... I was sad because of my illness 16. ... I was able to cope well with my illness 17. ... my parents treated me like a baby because of my illness 18. ... I avoided others to notice my illness 19. ... I missed something at nursery school/kindergarten because of my illness

Kid-KINDLR (8 to 12-year-olds) Childrens Version Disease Module

25. Are you staying in hospital just now or do you have some long-term illness? (Filter question) 26. ... I was afraid that my illness might get worse 27. ... I was sad because of my illness 28. ... I was able to cope well with my illness 29. ... My parents treated me like a baby because of my illness 30. ... I avoided others to notice my illness 31. ... I missed something at school because of my illness

Kiddo-KINDLR (13 to 16-year-olds) Teenagers Version

25. Are you staying in hospital just now or do you have some long-term illness? (Filter question) 26. ... I was afraid that my illness might get worse 27. ... I was sad because of my illness 28. ... I was able to cope well with my illness 29. ... My parents treated me like a baby because of my illness 30. ... I avoided others to notice my illness 31. ... I missed something at school because of my illness

1.2.2 Parents versions

Kiddy-KINDLR (4 to 7-year-olds) Parents Version Physical Well-Being
1. my child felt ill 2. my child had a headache or tummy-ache 3. ... my child was tired and worn-out 4. ... my child felt strong and full of energy 5. ... my child had fun and laughed a lot 6. ... my child didn't feel much like doing anything 7. ... my child felt alone 8. ... my child felt scared or unsure of itself 9. ... my child was proud of himself 10. ... my child felt on top of the world 11. ... my child felt pleased with himself 12. ... my child had lots of good ideas 13. ... my child got on well with us as parents 14. ... my child felt fine at home 15. ... we quarrelled at home 16. ... my child felt that I was bossing him around 17. ... my child played with friends 18. ... my child was liked by other kids 19. ... my child got along well with his friends 20. ... my child felt different from other children 21. ... my child coped well with the assignments set in nursery school/ kindergarten 22. ... my child enjoyed the nursery school/ kindergarten 23. ... my child looked forward to nursery school/kindergarten 24. ... my child made lots of mistakes when doing minor assignments or homework 1. my child felt ill 2. my child had a headache or tummy-ache 3. ... my child was tired and worn-out 4. ... my child felt strong and full of energy

KINDLR (8 to 16-year-olds) Parents Version

Emotional Well-Being
5. ... my child had fun and laughed a 6. ... my child didn't feel much like 7. ... my child felt alone 8. ... my child felt scared or unsure of lot doing anything itself

Self-Esteem
9. ... my child was proud of himself 10. ... my child felt on top of the world 11. ... my child felt pleased with himself 12. ... my child had lots of good ideas

Family
13. ... my child got on well with us as parents 14. ... my child felt fine at home 15. ... we quarrelled at home 16. ... my child felt that I was bossing him around

Friends
18. ... my child did things together with friends 19. ... my child was liked by other kids 20. ... my child got along well with his friends 21. ... my child felt different from other children 22. ... my child easily coped with schoolwork 23. ... my child enjoyed the school lessons 24. ... my child worried about his future 25. ... my child was afraid of bad marks or grades

Everyday Functioning (School or Nursery School/Kindergarten)

Ravens-Sieberer & Bullinger / 2000 / Page 5

Information about the KINDLR

Kiddy-KINDLR (4 to 7-year-olds) Parents Version

Additional Items Kiddy Parents

25. ... my child was moody and whined a lot 26. ... my child had a healthy appetite 27. ... I managed to show patience and understanding towards my child 28. ... my child felt under pressure 29. ... my child slept soundly 30. ... my child romped around and was very active 31. ... my child kept bursting into tears 32. ... my child was cheerful and in a good mood 33. ... my child was alert and able to concentrate well 34. ... my child was easily distracted and absent- minded 35. ... my child enjoyed being with other children 36. ... I had to give my child a telling-off 37. ... I praised my child 38. ... my child had problems with teachers, kindergarten staff or other child-minders 39. ... my child was nervous and fidgety 40. ... my child was lively and energetic 41. ... my child complained of being in pain 42. ... my child was sociable and out- going 43. ... my child succeeded at everything he set out to do 44. ... my child became dissatisfied easily 45. ... my child cried bitterly 46. ... my child lost his temper quickly

Kiddy-KINDLR (4 to 7-year-olds) Parents Version

KINDLR (8 to 16-year-olds) Parents Version

Disease Module
47. Is your child staying in hospital just now or does it have a long-term illness? (Filter question) 48. ... my child was afraid that the illness might get worse 49. my child was sad because of the illness 50. ... my child was able to cope well with his illness 51. ... we treated our child as though he were younger, because of the illness 52. ... my child avoided others to notice his illness 53. ... my child missed something at nursery school/kindergarten because of his illness 25. Is your child staying in hospital just now or does it have a long-term illness? (Filter question) 26. ... my child was afraid that the illness might get worse 27. my child was sad because of the illness 28. ... my child was able to cope well with his illness 29. ... we treated our child as though he were younger, because of the illness 30. ... my child avoided others to notice his illness 31. ... my child missed something at school because of his illness

Ravens-Sieberer & Bullinger / 2000 / Page 6

Information about the KINDLR

1.3 Psychometrics
The psychometric tests conducted on the questionnaire and discussed here include the testing of its reliability (internal consistency, i.e. Cronbachs alpha) and validity (factorial validity, convergent validity, discriminant validity and sensitivity). The first steps in the psychometric calculations consisted of an item analysis and a reliability analysis of the KINDLR questionnaire using the Multitrait Analysis Program of the New England Medical Center at Tuffts University in Boston (MAP) (Hays et al., 1988). The MAP program uses Campbells multitrait approach and permits a confirmatory test of the questionnaires postulated scale structure. For each item it delivers the magnitude of the correlation between the item and the overall scale (adjusted for overlap) as well as the success statistics for the frequency of cases displaying a higher or significantly higher correlation of an item as compared with the correlation with other scales. These success statistics, also known as scale fit, can take on values between 0% and an optimum 100%, and indicate the factorial validity of the questionnaire on an item level. Beyond this, the internal consistency coefficient Cronbach's alpha is given. The convergent validation was done by correlating the sub-scales of the KINDLR questionnaire with comparable well-being scales. The discriminant validity was determined by distinguishing subgroups within the survey population according to criteria that suggested a difference in their quality of life assessment. The analysis of the psychometric quality of the questionnaire was performed using two samples:

Sample 1
1501 pupils in the fourth and eighth grade completed the short version (12 items) of the KINDLR in the course of the regular school medical check-ups in Hamburg. The fourth graders (n=918) were an average of 9.7 years old, the eight grade teenagers (n=583) 14.1 years old. The sample consisted of 48.3 % girls and 51.7 % boys (Ravens-Sieberer et al., 2000a).

Sample 2
1050 children and adolescents from 7 German rehabilitation clinics completed the KINDLR questionnaire before embarking on a rehabilitation programme. The sample consisted of 50.7 % girls and 49.3 % boys. Their average age was 12.6 years; the sample comprised three diagnostic groups: bronchial asthma (n=254), atopic dermatitis (n=163) and obesity (n=633) (Ravens-Sieberer et al., 2000b).

Ravens-Sieberer & Bullinger / 2000 / Page 7

Information about the KINDLR

1.3.1 Structure of sub-scales, reliability and factorial validity

The analysis of the data from the KINDLR questionnaires completed by chronically ill children and adolescents and their parents (Sample 2, n=1050) using the MAP program revealed a good utilisation of the scale width with floor and ceiling effects generally below 10%. The scale fit was above 80% for all sub-scales. The reliability was checked by confirmatory testing, Cronbachs alpha as a measure of internal consistency reached values of around =.70 for most sub-scales, while the overall scale displayed a consistency coefficient of over =.80.
Table 1: Structure of Sub-Scales and Reliability (children, data transformed 0-100) Sub-scale Physical Emotional Self-esteem Family Friends School Total Disease module n 915 915 915 915 915 915 915 915 No. of items 4 4 4 4 4 4 24 6 70.63 80.31 57.88 77.69 71.44 64.56 70.58 77.71 Mean Standard deviation 17.31 14.88 20.56 17.13 18.25 21.88 11.94 17.96 Floor % .1 .0 .4 .0 .3 .1 .0 .0 Ceiling % 4.8 9.7 1.7 10.5 7.7 3.4 .0 11.9 Scale fit % 87.5 91.7 100.0 100.0 91.7 91.7 94.2 91.7 Internal consistency .63 .68 .75 .76 .74 .64 .84 .66

Table 2: Structure of Sub-Scales and Reliability (parents, data transformed 0-100) n No. of items Physical Emotional Selfesteem Family Friends School Total Disease module 899 899 899 899 899 4 4 4 24 6 79.38 67.94 65.63 56.58 69.83 17.75 18.13 18.06 13.15 18.88 .1 .0 .1 .0 .0 5.9 4.1 3.6 .0 6.1 100.0 87.5 95.8 95.8 88.9 .81 .74 .62 .89 .72 899 899 899 4 4 4 64.00 69.44 56.06 Mean Standard deviation 18.44 17.75 19.50 Floor % .0 .0 .2 Ceiling % 2.6 4.2 2.1 Scale fit % 83.3 100.0 100.0 Internal consistency .70 .76 .77

Ravens-Sieberer & Bullinger / 2000 / Page 8

Information about the KINDLR

1.3.2 Convergent validity

In terms of convergent validity, the KINDLR sub-scales were correlated both with the subjective health dimensions of the Child Health Questionnaire (Landgraf et al., 1999) and the SF-36 (Bullinger & Kirchberger, 1998) and with the Life Satisfaction questionnaire adapted for children (FLZM; Herschbach & Henrich, 2000). It was found that the KINDLR total score displayed a high correlation particularly with the General Well-Being sub-scale of the Child Health Questionnaires and with the Vitality and Emotional Well-Being subscales of the SF-36 as well as the FLZM (r>.60). Since the KINDLR aims more at the psychosocial than the physical aspects of health-related quality of life, the correlation of the KINDLR sub-scales with physical aspect of quality of life (Global Health and Physical Activities scale of the Child Health Questionnaire) are lower, as expected.
Table 3: Convergent Validity (Pearson r) KINDLR sub-scale Physical Emotional Self-Esteem Family Friends School Total Disease module CHQ Global health .34 .26 .32 .15 .19 .28 .40 .36 CHQ General well-being .52 .59 .50 .37 .48 .42 .72 .51 CHQ Physical activities .37 .31 .23 .12 .37 .15 .39 .37 .55 .48 .46 .26 .42 .31 .62 .41 SF-36 Vitality SF-36 Emotional well-being .44 .56 .44 .32 .41 .39 .64 .43 FLZ Life satisfaction .40 .45 .53 .47 .46 .41 .69 .45

1.3.3 Discriminant validity

The initial results of the use of the KINDLR questionnaire on chronically ill children show that the measure is able to distinguish between differences in the impairment of healthrelated quality of life in children with different diseases (asthma, atopic dermatitis, obesity) both on a sub-scale level and in terms of its total score (see Figure 2).
Table 4: Sub-scale Means for Different Samples Physical Healthy (n=1501) Asthma (n=254) Atopic dermatitis (n=163) Obesity (n=633) 75.56 71.02 74.94 70.1 Emotional 83.01 82.35 81.44 79.28 Selfesteem 66.6 63.68 62.63 54.71 83.99 79.33 80.76 76.78 78.18 76.8 77.72 70.84 73.13 67.34 68.55 62.43 76.75 73.38 74.41 68.93 82.28 78.29 75.45 Family Friends School Total Disease

Ravens-Sieberer & Bullinger / 2000 / Page 9

Information about the KINDLR

QoL sub-scale scores of healthy and chronically ill children

90

85

* **

** ** **

80

75

70

**

65

60

55

50 0

Physical
Healthy (n=1501)

Emotional

Self-Esteem

F a m ily

Friends

School

Total

Desease Module

Asthma (n=254)

Atopic dermatitis (n=163)

Obesity (n=633)

Figure 2: Discriminant Validity of the KINDLR Questionnaire (*p .05, **p .01)

1.3.4 Sensitivity
In a study dealing with the rehabilitation of chronically ill children (cf. Sample 2, Section 1.3) 1050 children completed the KINDLR questionnaire at the beginning and the end of a sixweek in-patient rehabilitation programme. Four of the six KINDLR sub-scales could be applied both before and after rehabilitation (School and Family had to be omitted because they could not be assessed). For all three groups of chronically ill children (asthma, atopic dermatitis and obesity), significant changes between before and after rehabilitation were apparent, particularly on the KINDLR dimensions Physical, Self-Esteem, Friends and Total Score, as well as on disease-specific asthma, atopic dermatitis and obesity modules and the additional Disease module.

100

90

**
80

** * ** **

**

** **

70

60

50 Physical Emotional SelfEsteem Friends Total Sub-scales Disease Asthma Atopic derm. Obesity

t-Test; * p < .05, ** p < .01

Before

After

Figure 3: KINDL Sub-Scale Differences by Time of Measurement . T-Test; * p

.05, ** p

.01

Ravens-Sieberer & Bullinger / 2000 / Page 10

Information about the KINDLR For the overall group, the Kazis effect size varied between d=.02 and .69, the changes between before and after rehabilitation were particularly pronounced for the dimension SelfEsteem and also in terms of the Total Score. Looking at the individual diagnostic groups, asthma and atopic dermatitis displayed effects in the field of the changes in chronic-generic and the asthma-specific module. In the obesity group, distinct effects were observed in the field of changes in Self-Esteem, Total Score and the disease-specific obesity module d=.69.
Table 5: Strength of Effect t1 t2 (KAZIS effect size) KINDL sub-scale Physical Emotional Self-Esteem Family Friends School Total Disease module Disease-specific module No. of items 4 4 4 4 4 4 24 6 12 .24 .24 .06 .07 .51 .11 .08 .50 .23 .18 .20 .26 d total .10 .03 .40 d asthma .08 .00 .14 d neurod. .02 .11 .12 d adipos. .13 .02 .56

.33
.33 .82

1.4 Execution
The KINDLR questionnaire was designed in order to obtain a self-assessment and an external assessment of health-related quality of life in children and adolescents in a wide range of ages and independently of current health status. For each item, the children and teenagers are required to mark the response that comes closest to their own personal experiences. Parents are asked to complete the KINDLR questionnaire as proxies, i.e. they are asked to judge the childrens quality of life from their own point of view (Ravens-Sieberer et al., in press a). In developing the KINDLR questionnaire, a great deal of emphasis was placed on ensuring that the questions and instructions are straightforward and easy to understand. Both the selfassessment version and the version for external assessment include the necessary instructions for completion and can therefore be answered by children, adolescents and parents without any further assistance from other persons. It is not necessary for a supervisor to be constantly present while the KINDLR questionnaire is being completed although this may be advisable in the case of children whose reading skills are perhaps not of a necessary standard. In such cases, the KINDLR questionnaire can be administered in the form of an interview (face-to-face or by telephone). Experiences and results so far indicate a high practicability of telephone interviews (Bullinger u. Bahner, 1997).
Ravens-Sieberer & Bullinger / 2000 / Page 11

Information about the KINDLR

The time needed to complete the KINDLR questionnaire, as measured empirically in a series of studies, varies between 5 and 15 minutes, depending on the age of the children/teenagers. The average time for completion is 10 minutes, whereby younger children usually require longer to complete the questionnaire. There is no time limit for completing the questionnaire, however for the self-assessed version it is advisable to have the individual questionnaires checked by an authorised person for completeness, since the analysis of the questionnaire may be jeopardised if only a few items are answered. If both the self-assessment and external assessment versions of the KINDLR questionnaire are used, it is crucial to ensure that the children/teenagers and parents complete the questionnaires independently of one another! One means of conducting surveys using the KINDLR questionnaire, beyond the simple paper-and-pencil questionnaire form, is to include the KINDLR questionnaire in computeraided diagnoses and to instruct the children/teenagers on how to enter their own data using a laptop computer. This option is available via the CAT-Screen program (Ravens-Sieberer et al., in press b). In the long term, the approach of a computer-assisted, multimedia version of the KINDLR questionnaire would seem sensible particularly in the clinical field, as a routine means of documenting initial and final check-ups (on admission and prior to being discharged) in children and adolescents.

1.5 Interpretation and Reference Values

The scores achieved on the individual KINDLR sub-scales and the KINDLR total score represent a quantification of the subjects health-related quality of life from the respondents point of view. There are three ways of interpreting these scores: First of all, the values within the individual sub-scales can be studied directly. The distance from the possible limits (maximum and minimum achievable values) can give a first indication of a respondents self-assessment. The second means of interpretation consists in comparing the sub-scale scores of individuals or populations with the reference values for corresponding age-groups and sexes. This can be done both in terms of the healthy comparison group and, where appropriate, with reference to specific diseases. Here the relative deviation of the measured value from the expected value can be quoted. In a third possible means of interpretation, changes in the patients clinical condition can be related to changes in his or her self-reported health status based on clinical measurements and quality of life data collected at the same time.

Ravens-Sieberer & Bullinger / 2000 / Page 12

Information about the KINDLR Until the data from a standard sample is available for the KINDLR questionnaire, the results of a large sample of Hamburg school children (n = 1501) can be used as a preliminary reference for healthy children (Sample 1, cf. Section 1.3). For the following reference values of the sub-scales transformed to a base of 100, the items missing from the short version have been estimated using regression analysis. The scores for the Disease module are based on a sample of chronically ill children (Sample 2, cf. Section 1.3). Here again, the scale has been transformed to a range of 0 to 100.
Children (8 - 12 years old) n = 918 Girls Boys mean KINDL - Total Quality of Life Score -100 KINDL Physical Well-being-100 KINDLR Emotional Well-being -100 KINDLR Self-Esteem -100 KINDL Family -100 KINDL Friends -100 KINDL - School -100 KINDLR Disease -100
R R R R R

Adolescents (13 - 16 years old) n=583 Girls Boys mean 70,78 68,24 79,41 58,14 75,51 78,06 65,19 60,10 s.d. 10,01 17,38 12,89 19,06 17,68 13,47 13,21 14,80 mean 73,54 77,18 79,49 63,27 79,56 78,43 63,58 64,91 s.d. 8,83 13,07 11,80 19,34 17,05 11,96 14,04 12,90

s.d. 8,63 14,19 11,33 17,83 12,85 13,78 12,29 15,25

mean 76,67 76,68 82,89 66,52 83,58 78,21 72,35 64,17

s.d. 8,66 13,03 10,67 18,95 13,14 12,78 12,88 13,75

76,83 74,43 83,11 66,68 84,40 78,10 74,10 60,56

Ravens-Sieberer & Bullinger / 2000 / Page 13

2 Analysing the KINDLR Questionnaire

The following instructions for analysing the six sub-scales that make up the KINDLR questionnaire and for determining the total score, contain general information about the analysis and describe the necessary steps from data entry through to analysis. These steps are the same for all forms Kiddy, Kid and Kiddo and for the childrens/teenagers and parents versions of the KINDLR questionnaire. Formulas will then be described for summarising the items and for converting the results into sub-scale scores. Finally, the possible ways of dealing with missing data will be dealt with. As with all standardised instruments, the KINDLR sub-scales can only be interpreted meaningfully if the contents and the method of analysis are standardised. Changing the content of the questionnaire or the individual steps in its analysis can destroy the reliability and validity of the results. Minor changes can affect the results to a point where comparisons with reference data have to be called into question, and the results of different studies can no longer be compared.

2.1 General Remarks on the Analysis

The KINDLR questionnaire is analysed by adding the item responses marked on each sub-scale, with certain items being reversed beforehand. Only sub-scales in which less than 30% of the items are missing can be analysed, whereby mean value replacement is used to deal with such missing values. A computerised analysis program exists for the KINDLR questionnaire, which carries out both item reversal and the summarisation of the sub-scales and their addition (see Chapter 3, Disc for Data Analysis). The items and sub-scales of the KINDLR questionnaire are calculated such that a higher score corresponds to a higher health-related quality of life. Once the data have been entered, analysis of the items and sub-scales is carried out in four steps: 1. Recoding items. This is necessary for 10 (in certain versions for 11) items. 2. Calculating sub-scale scores by adding the items in each sub-scale (raw scores). 3. Combining these to form a total score; and 4. transformation of the sub-scale scores to values between 0 and 100. These steps may be conducted using the algorithms presented in this Manual, or else using the program on the enclosed floppy disc for the SPSS statistics package.

Information about the KINDLR

2.2 Data Entry

The answers to the items in the KINDLR questionnaire should be entered exactly as coded in the questionnaire, in other words the number that the respondent has ticked, checked, circled or marked in some other way (raw value). In doing so, the response never is always assigned the value 1, the alternative response all the time the value 5. Occasionally, there may nevertheless be some confusion as to which number to enter. The following guidelines suggest how the most common coding problems should be dealt with: If two possible responses are marked for a single question and these responses are adjacent to one another, then one response is chosen according to a random procedure and entered. If two possible responses are marked for a single question and these responses are not adjacent to one another, then the item is coded as a missing value. If three or more possible responses are marked for a single question, the item is coded as a missing value.

2.3 Recoding Items

The next step following data entry is to recode the responses. Recoding is the process by which item scores are deduced which will then be used in calculating the sub-scale scores. This process consists of several steps: 1. Values that lie outside the valid range are converted into missing values; 2. Scores are reversed; and 3. Missing values are replaced by specific scores which are estimated for each individual.

2.3.1 Responses outside the valid range

Before the final item scores are assigned, all items should be checked to see whether answers occur that lie outside the possible range. Answers outside the valid range are values which are lower than the minimum score of 1 or higher than the maximum score of 5. Scores outside this range are usually due to errors made during data entry, and should where possible be replaced by the correct values by referring to the original questionnaire. If the questionnaire is not available, all values that lie outside the permitted range should be recoded as missing values.

2.3.2 Reversing items

10 (in some cases 11) KINDLR items are worded in such a way that a higher item score implies a poorer health-related quality of life. Reversing the values of these items is necessary in order to ensure that higher scores correspond to a higher health-related quality of life for all the KINDLR items and sub-scales. Chapter 2.5.1 gives details of which items need to be reversed L.
Ravens-Sieberer & Bullinger / 2000 / Page 15

Information about the KINDLR

2.4 Dealing with Missing Data

In rare cases (1-2% of respondents, or fewer), respondents fail to answer one or several items on a sub-scale. An advantage of sub-scales consisting of several items (like the sub-scales of the KINDLR) is that a sub-scale score can still be estimated even when some items are missing. The algorithm recommended on the floppy disc replaces each missing value by an estimate made for that specific individual, provided the respondent has answered at least 70% of the items on the sub-scale. This step in the analysis is integrated into the program for data analysis included on the disc.

2.5 Calculation of Sub-Scale Scores

2.5.1 Kid-KINDLR and Kiddo-KINDLR
When analysing the KINDLR questionnaire on the quality of life of children and adolescents in the age range of 8 to 16-year-olds, the following six sub-scale scores can be calculated: 1. 2. 3. 4. 5. Physical Well-being (Items 1L, 2L, 3L, 4) Emotional Well-being (Items 5, 6L, 7L, 8L) Self-esteem (Items 9, 10, 11, 12) Family (Items 13, 14, 15L, 16L) Friends (Items 17, 18, 19, 20L)

6. School (Items 21, 22, 23, 24L)

With the School sub-scale, it should be noted that in the KINDLR Parents (8-16 y) and the Kiddo version, Item 23 also has to be reversed. A Total Score is formed for all the items. Finally, if necessary an additional sub-scale can be calculated using the six question in the Disease module: 7. Disease (Items 26L, 27L, 28, 29L, 30L, 31L) The values are as follows: 1 = never 2 = rarely 3 = sometimes 4 = often 5 = all the time missing value = blank Important! The items marked with a L have to be reversed, i.e. 1=5, 2=4, 3=3, 4=2, 5=1. Response value 5 (all the time) must be the positive end of the item. !!! Data entered into the data base must always be in the form of raw data !!!
Ravens-Sieberer & Bullinger / 2000 / Page 16

Information about the KINDLR The sub-scale scores can be quoted in various ways. In the SPSS syntax used on the disc for data analysis, the following quantities are calculated: sum score sub-scale score (corresponding to the mean of the item scores) sub-scale score transformed to a range of 0 to 100

2.5.2 Formulae and examples for calculating sub-scale sum scores

Sum score = Sum of sub-scale items

Sub-scale score

Sum of sub - scale items Number of sub - scale items

=

Example: Physical well-being sub-scale score

Sum of Items 1, 2, 3, 4 4

Total score

Sum of all items 24 (Sub - scale score) - (lowest possible score) x 100 Possible range of raw score

Sub-scales transformed to 100

The calculations described here may be found in the SPSS file kid_kindl_k.sps (or the corresponding files for other versions) on the disc for data analysis.

2.5.3 Kiddy-KINDLR
The calculation of sub-scale scores for the parents version of the Kiddy KINDLR is essentially the same as described above for the other KINDLR versions. However the 22 additional items (Items 25 to 46) form a separate sub-scale known as Kiddy Parents. Here the following items need to be reversed: 25, 28, 31, 34, 36, 38, 39, 41, 44, 45, 46, 48, 49, 51, 52, 53. In the self-assessment version of the Kiddy interview, only the total score is calculated, and where necessary the additional sub-scale Disease. The values for the childrens version are as follows: 1 = never 2 = sometimes 3 = very often

Ravens-Sieberer & Bullinger / 2000 / Page 17

Information about the KINDLR

2.6 Contact
If you have any questions about the KINDLR questionnaire, its short form, the disease-specific modules, translations or the computer questionnaire CAT-Screen, we will be happy to assist you. You can contact us at the following addresses: Dr. Ulrike Ravens-Sieberer MPH Head of the Epidemiological Research Group "Child and Adolescent Health" Robert Koch Institute Stresemannstr. 90 D-10963 BERLIN Tel: +49-30-4547-3436 (-3434 Office) Fax: +49-30-4547-3555 E-mail: ravens-siebereru@rki.de Dr. Ulrike Ravens-Sieberer MPH Hamburg University Department for Medical Psychology Martinistr. 52 D-20246 Hamburg Tel: +49-40-42803-6206 (-6430 Office) Fax: +49-40-42803-4940 E-mail: ravens@uke.uni-hamburg.de

Ravens-Sieberer & Bullinger / 2000 / Page 18

Disk for Analysis

3 Disc for Data Analysis

After sending in the completed KINDLR user form, we will send you the disc for data analysis, which allows you to enter and analyse your KINDLR data simply and in a standardised form. The disc contains three types of files: 1. SPSS masks for entering KINDLR data (recognisable by the .sav file extension) 2. SPSS syntax files for forming scales (recognisable by the .sps file extension) 3. Read Me files containing advice on using the SPSS files Each KINDLR version has its own SPSS mask and its own syntax file: Mask Kiddy_Kindl_k_3.sav Kid_Kindl_k_3.sav Kiddo_Kindl_k_3.sav Kiddy_Kindl_e_3.sav Kid_Kindl_e_3.sav Syntax kiddy_kindl_k.sps kid_kindl_k.sps kiddo_kindl_k.sps kiddy_kindl_e.sps kid_kindl_e.sps

Parent and children versions may be identified by the ending _k (Kinder = children) and _e (Eltern = parents). The ending _3 indicates that the masks are designed for entering up to three separate time points. Because of the differences between the Kiddy-KINDLR and the other KINDLR versions, there are two Read Me files: Read Me_Kindl and Read Me_Kiddy (see Appendix).

Ravens-Sieberer & Bullinger / 2000 / Page 19

4 Further Reading
Bullinger, M., Mackensen, S., Kirchberger, I. (1994): KINDL - ein Fragebogen zur gesundheitsbezogenen Lebensqualitt von Kindern. Zeitschrift fr Gesundheitspsychologie 2: 64-67 Bullinger, M., Bahner, U. (1997). Zeitschrift fr Gesundheitswissenschaft. Bullinger, M., Kirchberger, I. (1998). SF-36 Fragebogen zum Gesundheitszustand. Hogefre, Gttingen. Hays, R.D., Hayashi, T., Carson, S. & Ware, J.E. (1988). Users Guide for the Multitrait Analysis Program (MAP). Santa Monica, CA: The RAND Corporation, N-2786-RC. Herschbach, P., Henrich, G. (2000). Fragen zur Lebenszufriedenheit (FLZM). In: U. RavensSieberer, & A. Cieza (Hrsg.). Lebensqualitt und Gesundheitskonomie in der Medizin Konzepte, Methoden, Anwendung. Mnchen: Ecomed-Verlag, 98-110. Landgraf, J. M., Abetz, L., Ware, J. E. (1999). Child Health Questionnaire (CHQ): A Users Manual. Boston, MA: HealthAct. Ravens-Sieberer, U. (1998). LQ- Messung bei Kindern psychometrische Ergebnisse zum KINDL. In: M. Bullinger, M. Morfeld, U. Ravens-Sieberer, U. Koch (Hrsg.). Medizinische Psychologie in einem sich wandelnden Gesundheitssystem: Identitt, Integration & Interdisziplinaritt. Pabst Verlag, 1998. Ravens-Sieberer, U. & Bullinger, M. (1998a). Assessing the health related quality of life in chronically ill children with the German KINDL: first psychometric and content-analytical results. Quality of Life Research, Vol. 4, No 7. Ravens-Sieberer, U. & Bullinger, M. (1998b). News from the KINDL-Questionnaire A new version for adolescents. Quality of Life Research, 7, 653. Ravens-Sieberer, U. (2000). Verfahren zur Erfassung der gesundheitsbezogenen Lebensqualitt bei Kindern und Jugendlichen Ein berblick. Bundesgesundheitsblatt Gesundheitsforschung Gesundheitsschutz, 43, 198-209. Ravens-Sieberer, U. & Cieza, A. (Hrsg.) (2000). Lebensqualitt und Gesundheitskonomie in der Medizin Konzepte, Methoden, Anwendung. Mnchen: Ecomed-Verlag. Ravens-Sieberer, U., Grtler, E. & Bullinger, M. (2000a). Subjektive Gesundheit und Gesundheitsverhalten von Kindern und Jugendlichen Eine Befragung Hamburger Schler im Rahmen der schulrztlichen Untersuchung. Gesundheitswesen, 62, 148-155. Ravens-Sieberer, U., Redegeld, M. & Bullinger, M. (2000b). Lebensqualitt chronisch kranker Kinder im Verlauf der statioren Rehabilitation. In J. Neuser & J. T. de Bruin (Hrsg.), Verbindung und Vernderung im Fokus der Medizinischen Psychologie, 2000 (S. 89). Lengerich: Pabst Science Publishers. Ravens-Sieberer, U., Grtler, E., Schwarzlmueller, M., Bullinger, M. (in press a). The proxy-byparent problem a comparison of direct and circular proxy QoL Ratings with childrens selfreport. Quality of Life Research. Ravens-Sieberer, U. Heilmann, M., Wallese,S. (in press b). Assessment of Quality of Life in Young Children with a Computer Assisted Touch Screen Program (Cat-Screen) Reliability, Validity and Feasibility. Quality of Life Research.

Appendix

5 Appendix
1. Read Me_Kindl File 28.08.2000 Dear KINDLR User, On our data analysis disc we have included a template for a KINDLR database (children and parents, different age versions) and a data analysis file for the KINDLR questionnaire. The calculations can be done using the SPSS program suite. Please pay special attention to the following points: The following comments concern the Kid-KINDLR (for children between the ages of 8 and 12) and the Kiddo-KINDLR (for adolescents between the ages of 13 and 16). If you are using the Kiddy-KINDLR (for children between the ages of 4 and 7) please refer to the corresponding Read Me file. The file kid_kindl_k_3.sav can be used to enter all the Kid-KINDLR data supplied by children. Analogously, the file kid_kindl_e_3.sav is for the parents data. The KiddoKINDLR data of adolescents can be entered into the file kiddo_kindl_k_3.sav. For all SPSS files on this disc, the files for the child/adolescent versions can be identified by a _k, while the parents versions bear an _e. !!! Data entered into the data base must always be in the form of raw data !!! The values of the KINDLR items are as follows: 1 = never, 2 = seldom, 3 = sometimes, 4 = often, 5 = all the time, blank is treated as a missing value. The values of all the other variables are recorded in the SPSS file. Further important variables in the file are: id = identification number (five digits), beginning with a 9 for children, with a 7 for parents. Apart from this, children and parents must have the same ID number so that data sets can subsequently be assigned to each other in pairs! sex (1=boy, 2=girl) name of the child age at the time point in question (enter years and months as decimal figures: age = years + (months/12), e.g. 8 years and 6 months = 8.5) time point in question

sex = name = alter =

geschw= number of siblings at the particular (0=0, 1=1, 2=2, 3=3, 4=4, 5=5, 6=more than 5)

Ravens-Sieberer & Bullinger / 2000 / Page 21

 schule = type of school (1=primary school, 2=lower secondary, 3=intermediate secondary, 4=comprehensive, 5=upper secondary, 6=special school, 7=private tuition) The parents version does not include the last two variables, instead it has the following additional variables: eltern = elt2 = parent completing questionnaire (1=mother, 2=father, 3=both together, 4=other) completing parent, other details (text field)

The instructions for reversing the raw scores are to be found in the analysis files kid_kindl_k.sps, kid_kindl_e.sps, and kiddo_kindl_k.sps, as are the commands for forming scales. The program file kid_kindl_k.sps also shows how six sub-scales and a total score can be formed. Forming sub-scales is carried out in an analogous way to the psychometric testing of the 24-item KINDL. The six sub-scales correspond to the arrangement in the questionnaire: Physical well-being (Items 1 to 4) Emotional well-being (Items 5 to 8) Self-esteem (Items 9 to 12) Family (Items 13 to 16) Friends (Items 17 to 20) School (Items 21 to 24)

In addition, where necessary an additional sub-scale can be formed from the six questions in the Disease module (Items 26 to 31). For each sub-scale, the results are calculated in three forms: a summed score, a mean score and finally a score transformed to a scale of 0 to 100. If you have any questions about data analysis, please do not hesitate to contact us at (Tel: +49 - 30 - 4547 - 3436 or - 3434). Yours sincerely

Dr. Ulrike Ravens-Sieberer M.P.H.

Tel: +49-30-4547-3434 ravens-siebereru@rki.de ravens@uke.uni-hamburg.de

Ravens-Sieberer & Bullinger / 2000 / Page 22

2. Read Me_Kiddy File 28.08.2000 Dear KINDLR User, On our data analysis disc we have included a template for a Kiddy-KINDLR database (children and parents) and a data analysis file for the KINDLR questionnaire. The calculations can be done using the SPSS program suite. Please pay special attention to the following points: The file kiddy_kindl_k_3.sav can be used to enter all the Kiddy-KINDLR data supplied by children. Analogously, the file kiddy_kindl_e_3.sav is for the parents data. For all SPSS files on this disc, the files for the child/adolescent versions can be identified by a _k, while the parents versions bear an _e. !!! Data entered into the data base must always be in the form of raw data !!! The values of the Kiddy-KINDLR items are as follows: 1 = never, 2 = sometimes, 3 = very often. For the parents version, the following values apply: 1 = never, 2 = seldom, 3 = sometimes, 4 = often, 5 = all the time. Blank is treated as a missing value. The values of all the other variables are recorded in the SPSS file. Further important variables in the file are: id = identification number (five digits), beginning with a 9 for children, with a 7 for parents. Apart from this, children and parents must have the same ID number so that data sets can subsequently be assigned to each other in pairs! sex (1=boy, 2=girl) name of the child age at the time point in question (enter years and months as decimal figures: age = years + (months/12), e.g. 8 years and 6 months = 8.5) time point in question

sex = name = alter =

geschw= number of siblings at the particular (0=0, 1=1, 2=2, 3=3, 4=4, 5=5, 6=more than 5)

schule = type of school (1=primary school, 2=lower secondary, 3=intermediate secondary, 4=comprehensive, 5=upper secondary, 6=special school, 7=private tuition) The parents version does not include the last two variables, instead it has the following additional variables:
Ravens-Sieberer & Bullinger / 2000 / Page 23

 eltern = elt2 =

parent completing questionnaire (1=mother, 2=father, 3=both together, 4=other) completing parent, other details (text field)

The instructions for reversing the raw scores are to be found in the analysis files kiddy_kindl_k.sps and kiddy_kindl_e.sps, as are the commands for forming scales. The program file kiddy_kindl_k.sps also shows how a total score can be formed from the twelve items in the childrens version. In the parents version, on the other hand, (kiddy_kindl_e.sps) six sub-scales can be formed. Forming sub-scales is carried out in an analogous way to the psychometric testing of the 24-item KINDL. The six sub-scales correspond to the arrangement in the questionnaire: 1. Physical well-being (Items 1 to 4) 2. Emotional well-being (Items 5 to 8) 3. Self-esteem (Items 9 to 12) 4. Family (Items 13 to 16) 5. Friends (Items 17 to 20) 6. School (Items 21 to 24) In addition, a specific sub-scale for parents of children up to the age of 7 is formed: Kiddy Parents ( (Items 25 to 46). Finally, where necessary an additional sub-scale can be formed from the six questions in the Disease module (children: Items 14 to 19; parents: 48 to 53). For each sub-scale, the results are calculated in three forms: a summed score, a mean score and finally a score transformed to a scale of 0 to 100. If you have any questions about data analysis, please do not hesitate to contact us at (Tel: +49 - 30 - 4547 - 3436 or - 3434). Yours sincerely

Dr. Ulrike Ravens-Sieberer M.P.H.

Tel: +49-30-4547-3434 ravens-siebereru@rki.de ravens@uke.uni-hamburg.de

Ravens-Sieberer & Bullinger / 2000 / Page 24