PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of archdischArchives of Disease in ChildhoodVisit this articleSubmit a manuscriptReceive email alertsContact usBMJ
 
Arch Dis Child. 2007 October; 92(10): 855–860.
Published online 2007 May 23. doi:  10.1136/adc.2006.098947
PMCID: PMC2083244

Adaptation of the Manchester‐Minneapolis Quality of Life instrument for use in the UK population

Abstract

Introduction

The availability of health‐related quality of life (HRQL) measures that are reliable, valid, brief and comprehensible and appropriate for use with UK children is limited. We report the validation of a HRQL measure suitable for UK use in healthy children, children with chronic disease conditions and socially disadvantaged children.

Patients

A total of 1238 children took part in the study, including healthy children as controls (n = 824) and five exemplar groups: children diagnosed with asthma (n = 87), diabetes (n = 103) or inflammatory bowel disease (IBD; n = 69), children in remission from cancer (n = 68) and children in public care (n = 87).

Methods

In phase I, the Manchester‐Minneapolis Quality of Life instrument (MMQL) Child Form was translated into UK English. In phases II and III, the questionnaire was shortened and validated.

Results

MMQL was anglicised and shortened to five components comprising 29 items. Good internal reliability was found with α reaching at least 0.69 for all subscales. Construct validity was established through moderate correlations with comparable PedsQL subscales (Pearson's r ranged from 0.38 to 0.58, p<0.01). Discriminant validity was also demonstrated in children with asthma and IBD, children in remission from cancer and children in public care, all of whom reported significantly lower HRQL than healthy children. Children with diabetes showed similar HRQL to their healthy peers. Good reproducibility and moderate responsiveness were demonstrated for the new measure.

Conclusions

The anglicised and shortened MMQL was shown to be valid and reliable and could be a valuable new tool for the assessment of HRQL in children.

Keywords: health‐related quality of life, paediatrics, chronic illness, public care

The value of measuring health‐related quality of life (HRQL) is now well recognised within medicine. Measuring HRQL includes recording patient perceptions of their physical, emotional and social health and function that can supplement clinical information on their health status. There has been a rapid growth in the number of scales available for the measurement of HRQL.1 Disease‐specific questionnaires allow the detection of relatively small changes in patients with a specific condition, whereas generic instruments allow comparisons of patients with different conditions or with healthy peers.

New HRQL scales have been mainly developed for adult populations; less attention has been given to the development of paediatric HRQL scales.2,3 Existing adult HRQL measures for various reasons are not appropriate for use in paediatric populations.2,3,4 The subscales measured may not be appropriate to children, for example issues relating to sexuality, employment or income may not be relevant. Children's HRQL measures have to take account of change and the stage of development of the child. Children and adults do not always share the same views about their illness, they may interpret questions differently and have different perspectives regarding disease course and impact.5,6

Proxy assessments of HRQL, completed by clinicians, parents or carers, have been used to assess a child's HRQL.3,7,8,9,10 Although proxy assessments can provide useful information on a child's health status,6,11 especially where a child is too ill or young to provide their own HRQL, it is now widely recognised that children should as far as possible rate their own HRQL.12,13 It has been documented that children as young as 5 can provide valid judgements on aspects of their health such as pain, but older children would be required to provide more valid feedback on emotional aspects of functioning.3,13 In addition, it has been documented that parental reports of a child's HRQL can differ substantially from the child's own self‐reported HRQL.8,9,10

Thus reliable, well‐validated child‐specific measures designed to assess the problems experienced by children are required. The aim of this study was therefore to anglicise, refine and validate a child HRQL measure that had been developed in the USA (the Minneapolis‐Manchester Quality of Life instrument, MMQL14) for use with healthy UK children, children in public care and children with chronic health conditions.

Methods

The study was carried out in three phases. Phase I involved anglicisation of the existing measure and refinement of MMQL coding categories and questionnaire layout. In phase II a short form of the MMQL was developed and in phase III the psychometric properties of the short form were tested.

Minneapolis‐Manchester Quality of Life instrument (MMQL)

This questionnaire was originally developed in the US by a UK paediatrician for use with cancer survivors.14 The original version has three different forms, one for children aged 8–11 years (Child Form, MMQL‐CF), one for 12–18 year olds (Youth Form, MMQL‐YF) and one for young people aged 19–25 years. The questionnaire measures the subscales of sports and activities, activity, feelings, physical appearance, friendships, intimate relationships, school and outlook.

PedsQL version 4.0 core module

The PedsQL measurement model is a modular approach to measuring HRQL in children and adolescents which is rapidly becoming established in the US.15,16 Anglicised versions are also available.17 It consists of a brief, practical generic core module, which is complemented by a number of condition‐specific measures. The child self‐report is suitable for children between 5 and 18 years of age. The subscales measured by the PedsQL are Physical Functioning, Emotional Functioning, Social Functioning and School Functioning.

Phase I

Anglicisation of the MMQL questionnaire followed the guidelines defined by Varni18,19and Quittner20 and was carried out by three of the research team (PU, CE and MJ). Any phrases and terms that were not used or understood by a sample of UK children were removed and possible alternatives were suggested. Some refinement of coding categories and questionnaire layout was also undertaken. An opportunistic sample of school children (age range 10–18 years) was recruited from a local school where interviews and completion of the MMQL took place.

Phase II

A total of 660 children aged between 8 and 18 and meeting the inclusion criteria were approached to take part in phase II. Five groups were chosen as exemplars representing four chronic conditions (asthma, diabetes, chronic inflammatory bowel disease (IBD) and cancer (long‐term survivors)). In addition children looked after in public care were recruited as the fifth exemplar. Children with a chronic health problem were identified with the guidance of collaborating clinicians and completed the questionnaires in clinic or at home. Children in public care were identified by a consultant paediatrician following a routine clinical assessment of a child in state care. Children in public care completed the questionnaires at home.

Controls were recruited from schools in Swansea, Neath, Port Talbot and Bridgend (Wales, UK). Permission to approach schools was obtained from the directors of education in each local education authority. School children completed the questionnaires in the classroom.

In both the exemplar and control groups, children with moderate to severe learning difficulties and children and parents for whom English was not their first language were excluded from the study.

Phase III

A total of 1347 children were approached to take part in phase III. Eligible subjects were recruited from the same five exemplar groups as in phase II. Inclusion criteria for the cancer group were altered slightly to make this a more homogeneous group. Specifically, only children aged 8–18 years who had received an allogeneic bone marrow transplant (BMT) following acute lymphoblastic leukaemia (ALL) and had completed therapy at the time of assessment were included in the study. Controls were recruited from schools using the same method as in phase II.

In phase III, all children completed the shortened and anglicised version of the MMQL‐YF. Non‐attenders at clinic visits were sent letters and offered the opportunity to take part in the study as it was felt that this group were not having an equal opportunity to be recruited into the study.

Subjects in this phase of the study were asked to complete the MMQL‐YF twice (initial and follow‐up clinic visit) for the purpose of evaluating test re‐test reliability and responsiveness.

The study was approved by the Welsh Multi‐centre Research Ethics Committee (MREC). Informed consent was sought from all children and their parents or those with parental responsibilities (for children less than 16 years of age), following oral and written explanation of the study.

Analysis

Data were analysed using the Statistical Package for Social Sciences (SPSS, Chicago, IL) version 11.4.

Assessing underlying dimensions and internal consistency

The internal consistency and underlying dimensions of the MMQL were assessed with the subjects from phase II and phase III. In phase II, principal component analysis21 was performed to determine the relevant components. A “direct oblimin” oblique rotation method was then applied to determine the structure of the questionnaires being shortened and produce a pattern of components that could be interpreted more easily. A component was considered important if its eigenvalue exceeded 1.1.22 The eigenvalue of a component represents the amount of variance explained by that component. Each component also had to exhibit face validity, that is, it appeared at face value to be measuring a clinically recognisable aspect of the patient's health. Questions were considered as contributing to a component if they had a factor loading of at least 0.4 on that component. Questions not contributing to any of the important components were considered for removal.

The internal consistency of the shortened MMQL components were assessed by item‐total correlations and Cronbach's alpha (α).23 Questions yielding item‐total correlations below 0.4 were considered for rejection.24 Questions were also considered for rejection if more than 75% of individuals gave the same response, because such questions are not sensitive enough to discriminate between different levels of severity.24 Questions were also considered for exclusion if they were disliked or considered difficult to answer by the children completing them. Finally, Cronbach's α for each of the resulting components should exceed 0.7.25

In phase III, principal components analysis was performed again without any restrictions on the data and the structure was compared to the initial emergent structure from phase II. This was carried out in order to validate the underlying components of the short form of the MMQL. Internal consistency was assessed by Cronbach's α.

Assessing validity

The construct validity of the shortened MMQL components was assessed in phase III by comparing them with the appropriate PedsQL subscales. If the components were valid measures of childhood HRQL, they would be expected to show significant small to moderate levels of correlation with each of the PedsQL scales, with the largest correlations being seen between the PedsQL scales measuring physical, social, emotional and school function and the comparable MMQL components.

In phase III, the discriminant validity was assessed by comparison of components scores between exemplar and control groups. If the shortened components were valid measures of the effects of childhood HRQL, the exemplar groups would be expected to score lower on the MMQL components than the control group. Independent samples t tests with Bonferroni corrections were used to compare control and exemplar groups.

Assessing reproducibility

Reproducibility of the MMQL was assessed in phase III. Following completion of the initial shortened MMQL in clinic, the exemplar groups were asked to complete a second “retest” questionnaire on their next visit to clinic. Children who were not due to return to clinic for 3–4 months were asked to complete the retest questionnaire at home. In addition to completing the MMQL, children were asked to rate whether their health had changed (improved, got worse or stayed the same) since the first questionnaire was completed. Those reporting no changes were included in the reproducibility analysis. Reproducibility was assessed using the intra‐class correlation coefficient.26

Assessing responsiveness

The responsiveness of the MMQL was assessed in phase III by using the scores of children who reported a change in their health. The response ratio (mean change in scores for subjects reporting a change divided by the standard deviation of the subjects reporting no change) was used to quantify the responsiveness. The larger the ratio, the more responsive the instrument.

Results

Phase I

Thirty children completed the anglicised MMQL and were interviewed. Children reported no problems with the language in the new versions of the MMQL. Minor comments were made on the length, amount of repetition and personal nature of some of the questions. However, no further modification of the MMQL was deemed necessary.

Phase II

A total of 390 children completed questionnaires in phase II.

The MMQL‐CF performed poorly in phase II and principal component analysis could not identify a reasonable component structure. The child form was therefore withdrawn from subsequent analysis and field work in phase III. All children (aged 8–18 years) were administered the shortened YF in phase III.

The principal component analysis of the data for the MMQL‐YF found five meaningful underlying components (table 11).). A total of 30 items were selected for inclusion in the shortened MMQL (MMQL‐YF30). Fourteen items were excluded because of a factor loading below 0.4 and one item was excluded because all children gave the same response to the item. Table 11 illustrates the five new proposed components.

Table thumbnail
Table 1 Selection of questions for the shortened MMQL‐YF30

Cronbach's α for the proposed MMQL‐YF30 components exceeded the accepted criterion of 0.70 (see table 22).

Table thumbnail
Table 2 Internal consistency of the MMQL‐YF30 components

Phase III

A total of 865 children completed questionnaires in phase III. Principal component analysis of the phase III data confirmed the underlying structure of the MMQL‐YF30 identified in phase II. The item “feeling uncomfortable with body development” was found to be redundant (factor loading less than 0.40) and was removed from the final version of the MMQL short form (MMQL‐YF29).

Internal reliability of the MMQL‐YF29 was confirmed by Cronbach's α values for each of the components, all but one of which reached or exceeded 0.70 (Appearance 0.89, School Functioning 0.86, Social Functioning 0.84, Emotional Functioning 0.84, Physical Functioning 0.69).

Validity of the MMQL‐YF29 was demonstrated by significant correlations between the MMQL‐YF components and the PedsQL subscales. This was most apparent in those components measuring similar constructs (table 33).

Table thumbnail
Table 3 Correlation between the MMQL‐YF29 components and PedsQL self‐report subscales

Table 44 illustrates significant differences between the mean component scores for the control and exemplar groups. Children with cancer, IBD and asthma reported lower physical functioning than control groups. Children in public care reported lower school scores than the control group. Children with diabetes reported higher scores than controls for all the component scores, although none of these reached significance once the Bonferroni corrections were applied.

Table thumbnail
Table 4 Mean scores (SD) on the MMQL‐YF29 components for different patient groups and controls

The reproducibility of the MMQL‐YF29 was good and was demonstrated by strong test‐retest intra‐class correlations in those children with no change in their health status (Appearance 0.86, School Functioning 0.83, Social Functioning 0.66, Emotional Functioning 0.74, Physical Functioning 0.87).

The responsiveness of the MMQL‐YF29 was tested in those children reporting a change in their health (table 55).). All children in this group reported an improvement in their health. Moderate levels of responsiveness were demonstrated for all components except social functioning, where children had lower HRQL despite reporting improved physical health. Physical functioning was the only component with a statistically significant difference in scores.

Table thumbnail
Table 5 Responsiveness of the MMQL‐YF29 components for children reporting an improvement in their health

Discussion

The MMQL was successfully shortened to five components comprising 29 items (MMQL‐29). In addition the coding categories and questionnaire structure were refined, and the language was made more appropriate for use with a UK population. It appears that the items selected for the MMQL‐29 reflect those of universal concern for both children and young people as this version of the MMQL was found to be suitable for 8–18 year olds. One benefit of this feature is that it will facilitate the evaluation of differences in HRQL across and between age groups, as well as tracking of HRQL longitudinally. The PedsQL is the only other empirically validated paediatric HRQL instrument to span a broad age range.27

Good internal reliability was found for the MMQL, with α exceeding 0.70 for all but one component.25 The only component not reaching 0.70 was Physical Functioning, which at 0.69 approached the recommended minimum standard. Construct validity of the MMQL‐29 was established through moderate correlations with the PedsQL core module. PedsQL is an established generic instrument which has been shown to be reliable and valid16 and is often heralded as one of the more promising paediatric measures of HRQL currently available.6

Discriminant validity was demonstrated for the MMQL‐29, with differences being noted in HRQL between healthy children and those with chronic health conditions. In most cases, these differences were in the expected direction, with children with chronic health conditions reporting lower HRQL than healthy children. However, this was not always the case, with perhaps the most notable exception being children with diabetes. This group reported better HRQL than healthy children on all components, although this did not always reach statistical significance. The similarity in HRQL between children with diabetes and their healthy peers has been highlighted previously.17,28

What is already known on this topic

  • There are no well developed measures to assess health‐related quality of life (HRQL) that are suitable for children in the UK.
  • HRQL measures need to be brief, reliable, multi‐dimensional and include a proxy rating.
  • The Minneapolis‐Manchester Quality of Life instrument (MMQL) was developed from focus groups with children and adolescents and has established psychometric properties.

What this study adds

  • We report a relatively brief (29 items), reliable measure of HRQL for UK children aged 8–18 years.
  • The measure distinguished between children with different chronic conditions as would be expected.
  • Validity was confirmed through predicted relationships with subscales of the PedsQL.
  • Responsiveness was confirmed through ratings made by children experiencing a change in health.

Other than the PedsQL, the MMQL‐29 is the only anglicised generic measure which fulfils the recommendations advocated for a childhood HRQL scale, that is, it is brief, comprehensive, reliable and valid.29 It is also recommended that a parent proxy form be available for child HRQL measures, a condition which the PedsQL also fulfils. A parent version of the MMQL was also developed in parallel with our anglicisation and shortening of the child MMQL (results of which will be reported elsewhere). Furthermore MMQL has the advantage of considering self‐image as well as physical, social, emotional and school functioning. It therefore has excellent potential value as a reliable and valid outcome measure for use in both research and clinical settings.

Acknowledgements

We would like to thank the authors of the US MMQL (Smith Bhati and Robert L Kane) for giving us permission to use the MMQL. We would also like to thank all the collaborators in local clinics, social services, local education authorities and participating schools without whom this research would not have been possible.

Abbreviations

HRQL - health‐related quality of life

IBD - inflammatory bowel disease

MMQL - Manchester‐Minneapolis Quality of Life instrument

MMQL‐CF - MMQL Child Form

MMQL‐YF - MMQL Youth Form

Footnotes

We would like to thank the Welsh Office of Research and Development (WORD) for funding the research.

Competing interests: None.`

References

1. Garratt A, Schmidt L, Mackintosh A. et al Quality of life measurements: bibliographic study of patient assessed health outcome measures. BMJ 2002. 2341417–1419.1419 [PMC free article] [PubMed]
2. Connolly M A, Johnson J A. Measuring quality of life in paediatric patients. Pharmacoeconomics 1999. 16(6)605–625.625 [PubMed]
3. Eiser C, Morse R. Quality‐of‐life measures in chronic diseases of childhood. Health Technol Assess 2001. 5(4)1–157.157 [PubMed]
4. Carpay J A, Arts W F M. Outcome assessment in epilepsy: available rating scales for adults and methodological issues pertaining to the development of scales for childhood epilepsy. Epilepsy Res 1996. 24127–136.136 [PubMed]
5. Eiser C. Children's quality of life measures. Arch Dis Child 1997. 77350–354.354 [PMC free article] [PubMed]
6. Eiser C, Morse R. A review of measures of quality of life for children with chronic illness. Arch Dis Child 2001. 84205–211.211 [PMC free article] [PubMed]
7. Puhan M A, Behnke M, Devereaux P J. et al Measurement of agreement on health‐related quality of life changes in response to respiratory rehabilitation by patients and physicians ‐ a prospective study. Respir Med 2004. 98(12)1195–1202.1202 [PubMed]
8. Guyatt G, Juniper E, Feeny D. et al Children and adult perceptions of childhood asthma. Pediatrics 1997. 99165–168.168 [PubMed]
9. Powers P M, Gerstle R, Lapey A. Adolescents with cystic fibrosis: family reports of adolescent health‐related quality of life and forced expiratory volume in one second. Pediatrics 2001. 107(5)70–74.74
10. Stancin T, Drotar D, Taylor H G. et al Health‐related quality of life of children and adolescents after traumatic brain injury. Pediatrics 2002. 10934–41.41 [PubMed]
11. Jokovic A, Locker D, Guyatt G. How well do parents know their children? Implications for proxy reporting of child health‐related quality of life. Qual Life Res 2004. 131297–1307.1307 [PubMed]
12. Ronen G M, Streiner D L, Rosenbaum P. Health‐related quality of life in childhood epilepsy: moving beyond ‘seizure control with minimal adverse effects'. Health Qual Life Outcomes 2003. 136
13. Matza L S, Swensen A R, Flood E M. et al Assessment of health‐related quality of life in children: a review of conceptual, methodological, and regulatory issues. Value Health 2004. 7(1)79–92.92 [PubMed]
14. Bhatia S, Jenney M E M, Bogue M K. et al The Minneapolis‐Manchester Quality of Life Instrument: reliability and validity of the Adolescent Form. J Clin Oncol 2002. 20(24)4692–4698.4698 [PubMed]
15. Varni J W, Seid M, Rode C A. The PedsQL™: measurement model for the pediatric quality of life inventory. Med Care 1999. 37(2)126–139.139 [PubMed]
16. Varni J W, Seid M, Kurtin P S. PedsQL™ 4. 0: reliability and validity of the Pediatric Quality of Life Inventory™ version 4, 0 generic core scales in healthy and patient populations. Med Care 2001. 39(8)800–812.812 [PubMed]
17. Upton P, Eiser C, Cheung I. et al Measurement properties of the UK‐English version of the Pediatric Quality of Life Inventory™ 4.0 (PedsQL™) generic core scales. Health Qual Life Outcomes 2005. 322
18. Varni J W. PedsQL™ Translations. http://www.pedsql.org/translations.html (accessed 24 May 2007)
19. Varni J W, Katz E R, Seid M. et al The Pediatric Cancer Quality of Life Inventory (PCQL) I: instrument development, descriptive statistics, and cross‐informant variance. J Behav Med 1998. 21(1)179–204.204 [PubMed]
20. uittner A L, Sweeny S, Watrous M. et al Translation and linguistic validation of a disease‐specific quality of life measure for cystic fibrosis. J Pediatr Psychol 2000. 25(6)403–414.414 [PubMed]
21. Joliffe I T. Principal component analysis. New York: Springer, 1986
22. Joliffe I T, Morgan B J T. Principal component analysis and exploratory factor analysis. Stat Methods Med Res 1992. 169–75.75 [PubMed]
23. Cronbach L J. Coefficient alpha and the internal structure of tests. Psychometrika 1951. 16297–334.334
24. Streiner G L, Norman R D. Health measurement scales. A practical guide to their development and use. Oxford: Oxford University Press, 1995
25. Nunnally J C, Bernstein I R. Psychometric theory. 3rd ed. New York: McGraw‐Hill, 1994
26. Deyo R A, Diehr P, Patrick D L. Reproducibility and responsiveness of health status measures. Statistics and strategies for evaluation. Control Clin Trials 1991. 12142S–158S.158S [PubMed]
27. Varni J W, Burwinkle T M, Jacobs J P. et al The PedsQL™ in type 1 and type 2 diabetes: reliability and validity of the Pediatric Quality of Life Inventory™ generic core scales and type 1 diabetes module. Diabetes Care 2003. 26631–637.637 [PubMed]
28. Laffel L M B, Connell A, Vangsness L. et al General quality of life in youth with type 1 diabetes. Diabetes Care 2003. 263067–3073.3073 [PubMed]
29. Mulhern R K, Horowitz M E, Ochs J. et al Assessment of quality of life among pediatric patients with cancer. Psychol Assess 1989. 1130–138.138

Articles from Archives of Disease in Childhood are provided here courtesy of BMJ Publishing Group