OUTCOMES IN PRACTICE

Pediatric Health-Related Quality of Life

Measurement Technology: A Guide for Health
Care Decision Makers
James W. Varni, PhD, Michael Seid, PhD, and Paul S. Kurtin, MD

Defining Health-Related Quality of Life

M
anaged care organizations, health care providers,
and purchasers of health care services are con- Most HRQOL measures have evolved from the World Health
cerned not only with improving health and well- Organization’s definition of health: a state of complete physi-
being but also with documenting the value of health care cal, mental, and social well-being, not merely the absence of
services provided. Outcomes assessment is essential to disease or infirmity [4]. For patients with chronic health con-
achieving these goals. Health outcomes assessment refers ditions, the goal of health care is to restore them to the fullest
to the evaluation of health care products, services, or pro- health possible by improving symptom management, treat-
grams and the consequences of their use [1]. Outcomes ment adherence, and their ability to cope with the impact of
assessment provides critical information regarding the effec- the condition. For this reason, HRQOL may be more impor-
tiveness and quality of health care and is an essential com- tant than biomedical measures when assessing patients with
ponent of performance improvement projects. chronic health conditions [5]. The value of HRQOL measure-
Comprehensive outcomes evaluation encompasses clini- ment in patients with chronic health conditions has repeated-
cal, economic, and patient-based outcomes. Patient-based ly been demonstrated with adults [6,7]. It follows that if we
outcomes are outcomes best described by the patient, such want to ensure that children receive the best possible health
as satisfaction with care and perceptions of quality of life. care in the most appropriate settings from the most qualified
Although quality of life can be construed to encompass all professionals, we need to assess their health status.
aspects of an individual’s life (eg, housing, neighborbood,
work, school), health-related quality of life (HRQOL) gener- Self-Report versus Proxy-Report
ally refers to those domains of health that can potentially be It is considered self-evident that patient self-report of
influenced by the health care system. Health-related quality HRQOL is the “gold standard” for measurement in adults. In
of life, health status, and functional status are terms often pediatrics, several issues, including cognitive-developmental
used interchangably to describe patients’ perceptions of their considerations, complicate the decision regarding the best
health, but HRQOL is considered the more comprehensive respondent for HRQOL assessment. While some measures
term [2]. allow for pediatric patient self-report, others rely on a proxy,
The importance of assessing the patient’s perception of such as a parent, to rate the child’s HRQOL; however, self-
his or her HRQOL has been well established [3], and a num- report and proxy-report often do not agree. This imperfect
ber of standard adult measures have been developed in concordance has been consistently noted in research compar-
recent years. In contrast, standard measures validated for ing adult patients’ self-reports and the reports of their health
use in evaluating the health care received by pediatric care providers and significant others (eg, spouses) [3]. In
patients are few in number and in the early stages of devel- pediatrics, a lack of congruence has been documented among
opment. In order to generate an evidence-based pediatric
health care database for use in determining the appropriate-
ness of health care services on a local and national basis, James W. Varni, PhD, Senior Scientist, Center for Child Health Out-
pediatric HRQOL instruments with excellent measurement comes, Children’s Hospital and Health Center, San Diego, CA, Pro-
precision are required. fessor of Psychiatry, University of California, San Diego, School of
Medicine, San Diego, CA, e-mail: [email protected]; Michael Seid, PhD,
This paper will discuss conceptual and methodological
Research Scientist, Center for Child Health Outcomes, Children’s
issues related to pediatric HRQOL measurement, evaluate Hospital and Health Center, San Diego; and Paul S. Kurtin, MD,
several existing instruments, and offer some guidance on Director, Center for Child Heath Outcomes, Children’s Hospital and
instrument selection. Health Center, San Diego.

Vol. 6, No. 4 JCOM April 1999 33

 PEDIATRIC QUALITY OF LIFE

the reports of physically healthy children and the reports based on health preferences among the general population.
of their parents, teachers, and health care professionals in The feasibility and usefulness of this approach with pediatric
assessments of the childrens’ functional status [8]. Agreement samples is doubted given the cognitive-developmental
among observers has been found to be lower for subjective sophistication required [30] and will not be discussed here.
experiences (eg, pain, nausea, depression, anxiety) than for An alternative measurement model to the generic versus
observable events (eg, emesis, walking, activities of daily liv- disease-specific dichotomy is an assessment strategy that
ing, behavior problems). This lack of agreement among re- combines both approaches. In this modular measurement
porters of pediatric patient functioning has been termed strategy, a generic core measure designed for use in pediatric
“cross-informant variance” [9] and has been observed in patients with acute and chronic health conditions as well as in
HRQOL assessment across multiple pediatric health condi- healthy populations is paired with supplemental condition-
tions [10–13]. Cross-informant variance is hardly surprising specific modules designed for use with designated patient
given that HRQOL derives from perceptions of the impact of samples. This strategy enables decision makers to measure
disease and treatment [14], and it clearly underscores the disease- and treatment-related symptoms or health problems
need for pediatric patient self-report instruments with defin- of discrete patient subgroups as well as compare diverse
itive measurement precision across the developmental stages groups of patients. The supplemental modules assess specific
of childhood and adolescence [12]. disease or treatment effects and other relevant HRQOL issues
Although reliable and valid self-report instruments are not sufficiently covered in the core measure, whereas the core
essential for the accurate determination of pediatric HRQOL, measure affords the opportunity to make comparisons across
parent proxy-report is important for several reasons. Because disease groups and with healthy population norms. The avail-
children are rarely in a position to refer themselves for treat- ability of physically healthy child and adolescent norms pro-
ment, even when they are experiencing symptoms and vides an essential benchmarking role not only in interpreting
health-related problems, parents’ perceptions of a child’s the meaning of scale scores but also as a metric with which to
HRQOL influences the likelihood that health care will be evaluate the outcome of interventions for pediatric patients
sought for the child. Further, the use of parent proxy-report to with chronic health conditions. Initial research on the utility of
estimate pediatric patient HRQOL may be necessary when this measurement strategy is very promising [12,13].
the patient is unable or unwilling to complete the HRQOL
measure because of young age or illness variables. Attributes of HRQOL Instruments
Three basic dimensions can be used to evaluate a potential
Generic versus Disease-Specific Measurement HRQOL instrument: conceptualization, measurement prop-
Two measurement strategies typically are used in health sta- erties (psychometrics), and practicality. Conceptualization
tus assessment: (1) generic population/epidemiologic assess- refers to the theoretical underpinnings of the measurement:
ment and (2) disease- or condition-specific assessment [15]. In How do the test constructors think of HRQOL? Is there a
generic assessment, the goal is breadth, or the ability to mea- theoretical or empirical basis for this conceptualization? Is
sure general health status across diverse samples including this conceptualization consistent with the potential user’s
screening healthy populations for specific problems related to ideas of how HRQOL should be measured in a particular
their health and well-being. Generic measures are designed to case? Is pediatric HRQOL to be measured by child self-
be broadly applicable and can be used to make comparisons report, proxy-report, or both? If there is a proxy-report form,
across diverse patient populations. In disease-specific assess- is it parallel to the child self-report?
ment, the goal is to assess health status within a circumscribed Similar to laboratory tests for biological disease, stan-
clinical sample, such as patients with asthma [16,17], cystic dardized tests for HRQOL assessment must have excellent
fibrosis [18], cancer [19–22], chronic headache [23], skin prob- measurement properties (reliability and validity) [10]. Re-
lems [24], spina bifida [25], or specific congenital or acquired liability refers to how well a measure reflects true scores (as
chronic physical impairments such as limb deficiencies opposed to error) [31]. In practice, reliability is often deter-
[26–28]. Disease-specific measures have the advantage of mined via internal consistency, or how well each test item
being more sensitive to clinical change in designated patient correlates with the scale of which it is intended to be part. A
groups, a quality that is often not found in generic measures general rule is that internal consistency should be at least
[29]. However, these specific measures are limited in their use- 0.70 for group comparisons and at least 0.90 for individual
fulness in making comparisons across diverse patient popu- comparisons [31]. Validity refers to how well an instrument
lations, including benchmarking with healthy population measures what it purports to measure. Validity is not an all-
norms. Thus, a trade-off between clinical usefulness and com- or-nothing quality. Most tests have some degree of validity,
parative data often exists in selecting one type of measure over and virtually no HRQOL test has perfect validity in all situ-
another. A third type of measurement, utility assessment, is ations.

34 JCOM April 1999 Vol. 6, No. 4

 OUTCOMES IN PRACTICE

Table. Comparison of Pediatric HRQOL Instruments

Instrument

Attribute PedsQL CHQ* CHIP FSII(R)

Conceptualization
Multidimensional Yes Yes Yes Yes
Scales assessing physical health Yes Yes Yes Yes
Scales assessing mental health Yes Yes Yes No
Meaning of high scores Perception of Minimal problems, Able to participate Few behavioral
minimal problems, health does not fully in manifestations of
high well-being interfere with role, appropriate tasks problems
high well-being
Modular format Yes No No No
Reliability and validity
Peer-reviewed psychometric data Yes Yes, for parent form Yes Yes
Internal consistency† Group Mixed‡ Group Group/individual
Validity – factor structure Yes Yes Yes Yes
Validity – known groups Yes No Yes Yes
Validity – hospital days, days lost Yes No Yes Yes
Practicality (ease of use)
Mode of administration Self, interviewer Self, interviewer Self, interviewer Interviewer
Length 30 items 87 items 175 items 43 items,
14-item short form
Time to complete Less than 5 minutes No data 45 minutes No data
Scored easily Yes No Yes Yes
Patient-report age range (yrs) 5 to 18 10 to 17 11 to 17 No self report
Proxy-report age range (yrs) 2 to 18 10 to 17 No proxy report 0 to 16
Parallel patient- and proxy-reports Yes No No No
Languages English, Spanish English, Spanish, others English English, Spanish

PedsQL = Pediatric Quality of Life Inventory; CHQ = Child Health Questionnaire; CHIP = Child Health and Illness Profile; FSII(R) = Functional Status
Measure (updated).

*Pediatric self-report format, unless otherwise indicated.

†Internal consistency reliability should exceed 0.70 for group comparisons, 0.90 for individual comparisons.

‡Some scales fall above 0.70, although others do not.

Practicality refers to the test’s usefulness in real-world (or both) and how these reports are related (Table). In cases in
settings as well as the ease with which it is administered, which both a self-report and proxy-report exist, the discussion
scored, and interpreted. focuses on the properties of the self-report form.

Properties of Select Instruments The PedsQL™ (Pediatric Quality of Life Inventory™)

Although there are numerous disease- and condition-specific During the past 15 years, Varni and associates have conduct-
pediatric HRQOL instruments, this discussion will consider ed a programmatic research effort in measurement instru-
only instruments that can be used in more than one popula- ment development, which has resulted in the items contained
tion. The instruments discussed vary along many dimensions, in the PedsQL (Figure). The PedsQL has been designed as a
including whether the measure is self-report or proxy-report modular HRQOL measurement instrument for pediatric

Vol. 6, No. 4 JCOM April 1999 35

 PEDIATRIC QUALITY OF LIFE

health conditions, with generic core scales and condition- Practicality. The PedsQL is a 30-item measure and takes less
specific disease and treatment related modules [12,13]. It con- than 5 minutes to complete. Missing data rates are about
sists of parallel child self-report and parent proxy-report of the 0.01% of item responses for the pediatric cancer, arthritis,
child’s HRQOL such that the parent-report items are third- and asthma samples [12,13]. Scoring is very straightforward,
person verbatim copies of the items found on the child’s with all items scored on a 0 to 4 scale and easily converted to
report. Child and parent reports of the child’s HRQOL can be the 0 to 100 scale for standardized interpretation. The age
compared across the age range. range for the PedsQL is 5 to 18 years for patient self-report
and 2 to 18 years for parent proxy-report. The instrument has
Conceptualization. The PedsQL conceptualizes pediatric been field-tested in English and Spanish, is currently being
HRQOL as the patient’s perceptions of the impact of disease field-tested in German, and is being adapted internationally.
and treatment in a variety of health and well-being domains The measure is self-administered for parents and older chil-
[2,32], including physical, emotional, social, and school func- dren, with children aged 5 to 7 years completing the measure
tioning and general perceptions of well-being [4,33,34]. This with the aid of an interviewer.
conceptualization is consistent with recent writings [35] sug-
gesting that it is reasonable to expect that “someone in the Child Health Questionnaire (CHQ)
best possible physical and emotional state compatible with The CHQ parent and child questionnaires encompass seve-
their medical condition has the best chance of achieving a ral domains of physical health, mental health, and role func-
high quality of life, whatever this may mean to the individ- tioning [36]. The parent-report and child-report forms are
ual concerned.” The PedsQL multidimensional generic core similar but not parallel. The CHQ user’s manual indicates
scales encompass the essential domains for pediatric that the questionnaires are generic health measures for use in
HRQOL assessment: physical functioning (8 items), emo- general survey research [36].
tional functioning (5 items), social functioning (5 items),
school functioning (5 items), well-being (6 items), and a glo- Conceptualization. The CHQ conceptualizes HRQOL as con-
bal perception of overall health status (1 item). Scales are sisting of physical and psychosocial well-being measured
scored from 0 to 100, with 100 indicating highest HRQOL. along the dimensions of status, disability, and personal eval-
uation [36]. Summary scales for the CHQ include physical
Reliability and validity. The PedsQL generic core scales were functioning, self-esteem, mental health, general health per-
designed to be used across various pediatric health condi- ceptions, behavior, bodily pain, role/social–physical (the
tions, with PedsQL cancer, asthma, and arthritis modules impact of physical functioning on role performance), role/
developed and field-tested thus far [12,13]. The PedsQL social–emotional/behavioral (the impact of emotional/
generic core scales and condition-specific modules have behavioral functioning on role performance), parental
demonstrated good internal consistency reliability, with co- impact/time, parental impact/emotional, family activities,
efficient alpha generally ranging from 0.70 to 0.92 for patient and family cohesion.
self-report of health and well-being [12,13]. Construct validity
has been demonstrated for both item-level and scale-level Reliability and validity. Item-level analyses show a wide
analyses [12,13]; clinical validity has been established by range in the performance of individual items and scale scores,
demonstrating that PedsQL scores distinguish between with internal consistency alpha coefficients ranging from 0.62
pediatric cancer patients on- and off-treatment [12] and that to 0.97. Construct validity for the child self-report CHQ is
PedsQL scale scores are associated in the expected direction documented via item-level analysis. The CHQ user’s manual
with disease- and treatment-related symptoms in patients documents the clinical validity of the parent-report form, but
with asthma and arthritis [13] and are strongly related to days it does not address the child-report form. Norms are available
missed from school for patients and days missed from work for the parent-report form but not for the child-report form.
for parents [13]. PedsQL modules are currently being devel-
oped for pediatric diabetes, cystic fibrosis, sickle cell disease, Practicality. The parent-report form is available in 50- and
cerebral palsy, cardiology, and medically fragile patients. 28-item versions, but the child-report form is 87 items long.
The PedsQL is continuously being field-tested nationally The length of the child-report form may be a factor in data
in pediatrician offices, hospital specialty clinics, community completeness. For example, in the middle-school sample,
settings, and schools, as well as in international field trials. It 53% to 60% of 10 to 12 year olds and 72% to 74% of older chil-
is anticipated that 5000 to 10,000 pediatric respondents and dren completed all 87 items [36]. In the clinical groups, com-
their parents will be accrued by 1999, including physically pletion rates ranged from 63% to 77% [36]. Scoring is some-
healthy school children and children with a wide diversity of what complicated by the fact that different items have
chronic health conditions. different numbers of responses and that item values on some

36 JCOM April 1999 Vol. 6, No. 4

 OUTCOMES IN PRACTICE

Below is a list of things that might be a problem for you. Please tell us how much of a problem each one has been for you during the past
ONE month.
There are no right or wrong answers. If you do not understand a question, please ask for help.
In the past ONE month, how much of a problem has this been for you . . .

About My Health And Activities Never Almost Never Sometimes Often Almost Always
It is hard for me to walk more than one block 0 1 2 3 4
It is hard for me to run 0 1 2 3 4
It is hard for me to do sports activity or exercise 0 1 2 3 4
It is hard for me to lift something heavy 0 1 2 3 4
It is hard for me to take a bath or shower by myself 0 1 2 3 4
It is hard for me to do chores around the house 0 1 2 3 4
I hurt or ache 0 1 2 3 4
I have low energy 0 1 2 3 4

About My Feelings Never Almost Never Sometimes Often Almost Always

I feel afraid or scared 0 1 2 3 4
I feel sad or blue 0 1 2 3 4
I feel angry 0 1 2 3 4
I have trouble sleeping 0 1 2 3 4
I worry about what will happen to me 0 1 2 3 4

How I Get Along With Others Never Almost Never Sometimes Often Almost Always
I have trouble getting along with other kids 0 1 2 3 4
Other kids do not want to be my friends 0 1 2 3 4
Other kids tease me 0 1 2 3 4
I cannot do things that other kids my age can do 0 1 2 3 4
It is hard to keep up when I play with other kids 0 1 2 3 4

About School Never Almost Never Sometimes Often Almost Always

It is hard to pay attention in class 0 1 2 3 4
I forget things 0 1 2 3 4
I have trouble keeping up with my schoolwork 0 1 2 3 4
I miss school because of not feeling well 0 1 2 3 4
I miss school to go to the doctor or hospital 0 1 2 3 4

Please tell us how much each sounds like you during the past ONE month.
In the past ONE month, how much does this sound like you . . .

About Me Never Almost Never Sometimes Often Almost Always

I feel happy 0 1 2 3 4
I feel good about myself 0 1 2 3 4
I feel good about my health 0 1 2 3 4
I get support from my family and friends 0 1 2 3 4
I think good things will happen to me 0 1 2 3 4
I think my health will be good in the future 0 1 2 3 4

In the past ONE month . . .

In General Bad Fair Good Very Good Excellent

In general, how is your health? 0 1 2 3 4

Figure. Pediatric Quality of Life Inventory (PedsQL). Version shown is adapted from child report for ages 8 to 12 years. © 1998 by JW Varni. All rights reserved.

Vol. 6, No. 4 JCOM April 1999 37

 PEDIATRIC QUALITY OF LIFE

items have to be recoded based on weightings. The parent appropriate activities. The developers viewed behavior as
proxy-report form is available for children aged 5 to 18 years, the final common pathway of health and defined the healthy
and the child self-report form is available for patients aged child as one who exhibits age-appropriate physical, psycho-
10 to 18 years. The instrument is available in English, Spanish, logical, intellectual, and social behaviors [43]. This measure
French, and other languages. Both parent and child reports attempts to isolate health-related problems from other be-
can be self-administered or interviewer-administered. haviors by asking, for each question, whether the behaviors
are due to a health problem.
The Child Health and Illness Profile (CHIP)
The CHIP [37] is a broad measure of health for use in epi- Reliability and validity. The FSII(R) 43-item and 14-item
demiologic surveys of general health with adolescents aged parent-report forms have excellent internal consistency reli-
11 to 17 years [38]. The measure consists of an adolescent ability, whereas the 7-item form has adequate to marginal
self-report form. Work is underway on a child self-report reliability. The self-report 14-item form has good internal
form, but this has yet to be disseminated; there is no parent consistency reliability. For the interviewer form, construct
proxy-report form. validity has been demonstrated via factor analysis, with a
general health factor emerging for all ages as well as a stage-
Conceptualization. The CHIP is defined as “a comprehensive specific factor (responsiveness for children younger than
instrument that broadly defines health as the ability to partic- 2 years, activity for children 2 to 3 years old, and interper-
ipate fully in developmentally appropriate physical, psycho- sonal functioning for children aged 4 years and older). The
logical, and social tasks” [37]. The measure assesses 6 broad measure distinguishes between well and ill children across
domains of health functioning: activity, comfort, satisfaction the age range and correlates with days in the hospital and
with health (perceived well-being), disorders, achievement, days lost from school [43]. The self-report form has low to
and resilience. Each domain has various subdomains. moderate correlations with measures of use, severity, or
function [44].
Reliability and validity. The CHIP has been tested in several
thousand adolescents, including well, acutely ill, and chron- Practicality. The measure is a proxy interview for ages 0 to
ically ill patients. Overall, the evidence shows that internal 16 years. As an interview measure, the FSII(R) takes approx-
consistency reliability is sufficient for use in group compar- imately 30 minutes [44] and requires an interviewer to
isons [37,39]. Content validity has been addressed via focus administer. Thus, it is probably more expensive than the
groups and expert consultation [37]. Construct and clinical paper-and-pencil measures, but missing data are unlikely to
validity has been demonstrated in the ability of the instru- be a problem. The measure lacks specific assessment of men-
ment to distinguish among known groups of adolescents tal health and disease- and treatment-specific symptoms,
[37,39,40]. The CHIP also has been validated for describing and the factors are based on strictly behavioral indices of
adolescent reports of health needs [41,42]. functioning from the parent’s perspective. Although the
measure is intended for use up to age 16, it does not appear
Practicality. The CHIP is self-administered, but it is 175 items to have age-appropriate activity and development items
long and takes approximately 45 minutes to complete [37]. It for adolescents. The measure is available in English and
has been used almost exclusively in research settings and Spanish.
epidemiologic studies.
Conclusion
FSII(R) Pediatric HRQOL measurement technology is still at an
The Functional Status Measure has been updated as the early stage of development. As such, consensus has yet to be
FSII(R) [43]. It was developed to assess the health status of reached among users of this technology regarding which
chronically ill children. This measure is available only as instrument should be considered the gold standard. Health
parent-report and is administered via interview. The long care decision makers interested in applying HRQOL mea-
form is 43 items. A shorter, 14-item core measure has been surement technology will benefit to the extent that they
developed but still requires a clinical interview format. A choose the technology appropriate for their needs. To facili-
self-report format has been reported in the literature [44]. tate the decision-making process, the authors offer the fol-
lowing guidelines:
Conceptualization. The FSII(R) conceptualizes health status
as the behavioral manifestation of functioning. The measure 1. The instrument should be practical, ie, reasonably
inventories behavioral manifestations of illness that interfere short, easy to administer, easy to score, and easy to
with an individual’s performance of the full range of age- interpret.

38 JCOM April 1999 Vol. 6, No. 4

 OUTCOMES IN PRACTICE

2. The instrument should have excellent reliability children with newly diagnosed cancer: cross-informant vari-
and validity. Evidence from peer-reviewed journal ance. J Psychosoc Oncol 1995;13:23–38.
articles should be the standard for evaluation of 10. Varni JW, Setoguchi Y. Screening for behavioral and emo-
tional problems in children and adolescents with congenital
the instrument’s measurement properties.
or acquired limb deficiencies. Am J Dis Child 1992;146:103–7.
11. Guyatt GH, Juniper EF, Griffith LE, Feeny DH, Ferry PJ.
3. The instrument should demonstrate utility in di- Children and adult perceptions of childhood asthma.
verse pediatric populations. Alarge normative data- Pediatrics 1997;99:165–8.
base that includes the HRQOL scores of physically 12. Varni JW, Seid M, Rode CA. The PedsQL: measurement
healthy children and adolescents is desirable for model for the Pediatric Quality of Life Inventory. Med Care
benchmarking and quality improvement purposes. 1999;37:126–39.
13. Varni JW, Seid M, Jacobs JR, Rode CA. The PedsQL: II.
4. The instrument should permit patient self-report for Development of the asthma and arthritis disease-specific
a wide range of ages. Research has demonstrated modules for the Pediatric Quality of Life Inventory. Qual
that children as young as 5 years old can self-report Life Res. Under review 1999.
their health and well-being [45]. Parent proxy-report 14. Schipper H, Clinch JJ, Olweny CLM. Quality of life studies:
should parallel child self-report. definitions and conceptual issues. In: Spilker B, editor. Quality
of life and pharmacoeconomics in clinical trials. 2nd ed. Phila-
delphia: Lippincott-Raven; 1996:11–23.
5. It is preferable that the instrument employ a mod-
15. Patrick DL, Deyo RA. Generic and disease-specific measures
ular format that integrates a generic core scale with
in assessing health status and quality of life. Med Care 1989;
disease-specific modules, enabling decision mak- 27(Suppl 3):S217–32.
ers to make comparisons with healthy population 16. Juniper EF, Guyatt GH, Epstein RS, Ferrie PJ, Jaeschke R,
norms and to measure disease- and treatment- Hiller TK. Evaluation of impairment of health related qual-
related symptoms with greater sensitivity and ity of life in asthma: development of a questionnaire for use
specificity. in clinical trials. Thorax 1992;47:76–83.
17. Juniper EF, Guyatt GH, Feeny DH, Ferrie PJ, Griffith LE,
References Townsend M. Measuring quality of life in children with asth-
1. Mariner WK. Outcomes assessment in health care reform: ma. Qual Life Res 1996;5:35–46.
promise and limitations. Am J Law Med 1994;20:37–57. 18. Czyzewski DI, Mariotto MJ, Bartholomew LK, LeCompte
2. Guyatt GH, Feeny DH, Patrick DL. Measuring health-related SH, Sockrider MM. Measurement of quality of well being in
quality of life. Ann Intern Med 1993;118:622–9. a child and adolescent cystic fibrosis population. Med Care
3. Sprangers MA, Aaronson NK. The role of health care pro- 1994;32:965–72.
viders and significant others in evaluating the quality of life 19. Varni JW, Katz ER, Seid M, Quiggins DJ, Friedman-Bender A,
of patients with chronic disease: a review. J Clin Epidemiol Castro CM. The Pediatric Cancer Quality of Life Inventory
1992;45:743–60. (PCQL). I. Instrument development, descriptive statistics,
4. World Health Organization. Constitution of the World Health and cross-informant variance. J Behav Med 1998;21:179–204.
Organization basic document. Geneva, Switzerland: The 20. Goodwin DAJ, Boggs SR, Graham-Pole J. Development and
Organization; 1948. validation of the Pediatric Oncology Quality of Life Scale.
5. Coons SJ, Kaplan RM. Quality of life assessment: understand- Psychol Assess 1994;6:321–8.
ing its use as an outcome measure. Hosp Formulary 1993;28: 21. Varni JW, Katz ER, Seid M, Quiggins DJ, Friedman-Bender
486–98. A. The Pediatric Cancer Quality of Life Inventory-32 (PCQL-
6. Sprangers MA, Cull A, Bjordal K, Groenvold M, Aaronson 32): I. Reliability and validity. Cancer 1998;82:1184–96.
NK. The European Organization for Research and Treatment 22. Varni JW, Rode CA, Seid M, Katz ER, Friedman-Bender A,
of Cancer. Approach to quality of life assessment: guidelines Quiggins DJ. The Pediatric Cancer Quality of Life Inventory-
for developing questionnaire modules. EORTC Study Group 32 (PCQL-32): II. Feasibility and range of measurement.
on Quality of Life. Qual Life Res 1993;2:287–95. J Behav Med. In press 1999.
7. McHorney CA, Ware JE Jr, Lu JF, Sherbourne CD. The MOS 23. Langeveld JH, Koot HM, Loonen MC, Hazebroek-
36-item Short-Form Health Survey (SF-36): III. Tests of data Kampschreur AA, Passchier J. A quality of life instrument
quality, scaling assumptions, and reliability across diverse for adolescents with chronic headache. Cephalalgia 1996;16:
patient groups. Med Care 1994;32:40–66. 183–96.
8. Achenbach TM, McConaughy SH, Howell CT. Child/adoles- 24. Lewis-Jones MS, Finlay AY. The Children’s Dermatology
cent behavioral and emotional problems: implications of cross- Life Quality Index (CDLQI): initial validation and practical
informant correlations for situational specificity. Psychol Bull use. Br J Dermatol 1995;132:942–9.
1987;101:213–32. 25. Parkin PC, Kirpalani HM, Rosenbaum PL, Fehlings DL, Van
9. Varni JW, Katz ER, Colegrove R, Dolgin M. Adjustment of Nie A, Willan AR, King D. Development of a health-related

Vol. 6, No. 4 JCOM April 1999 39

 PEDIATRIC QUALITY OF LIFE

quality of life instrument for use in children with spina bi- 36. Landgraf J, Abetz L, Ware J E. The CHQ user’s manual. Bos-
fida. Qual Life Res 1997;6:123–32. ton: The Health Institute, New England Medical Center;
26. Pruitt SD, Varni JW, Seid M, Setoguchi Y. Prosthesis satisfac- 1996.
tion outcome measurement in pediatric limb deficiency. Arch 37. Starfield B, Bergner M, Ensminger M, Riley A, Ryan S, Green
Phys Med Rehabil 1997;78:750–4. B, et al. Adolescent health status measurement: development
27. Pruitt SD, Varni JW, Setoguchi Y. Functional status in chil- of the Child Health and Illness Profile. Pediatrics 1993;91:
dren with limb deficiency: development and initial valida- 430–5.
tion of an outcome measure. Arch Phys Med Rehabil 1996; 38. Starfield B, Riley AW, Green BF, Ensminger ME, Ryan SA,
77:1233–8. Kelleher K, et al. The adolescent child health and illness pro-
28. Pruitt SD, Seid M, Varni JW, Setoguchi Y. Toddlers with limb file. A population-based measure of health. Med Care 1995;
deficiency: conceptual basis and initial application of a func- 33:553–66.
tional status outcome measure. Arch Phys Med Rehabil. In 39. Starfield B, Forrest CB, Ryan SA, Riley AW, Ensminger ME,
press 1999. Green BF. Health status of well vs ill adolescents. Arch Pediatr
29. Murawski MM, Miederhoff PA. On the generalizability of Adolesc Med 1996;150:1249–56.
statistical expressions of health related quality of life instru- 40. Forrest CB, Starfield B, Riley AW, Kang M. The impact of
ment responsiveness: a data synthesis. Qual Life Res 1998; asthma on the health status of adolescents. Pediatrics 1997;
7:11–22. 99:E1.
30. Revicki DA, Kaplan RM. Relationship between psychometric 41. Riley AW, Forrest CB, Starfield B, Green BF, Kang M, Ens-
and utility-based approaches to the measurement of health- minger ME. A taxonomy of adolescent health need: reliabili-
related quality of life. Qual Life Res 1993;2:477–87. ty and validity of the adolescent health and illness profiles.
31. Nunnally JC, Bernstein IH. Psychometric theory. 3rd ed. Med Care. In press 1999.
New York: McGraw-Hill; 1994. 42. Riley AW, Green BF, Starfield B, Forrest CB, Kang M,
32. Pal DK. Quality of life assessment in children: a review of Ensminger ME. A taxonomy of adolescent health need: de-
conceptual and methodological issues in multidimensional velopment of the adolescent health and illness profiles. Med
health status measures. J Epidemiol Community Health 1996; Care. In press 1999.
50:391–6. 43. Stein RE, Jessop DJ. Functional Status II (R). A measure of
33. Hays RD, Stewart AL. The structure of self-reported health child health status. Med Care 1990;28:1041–55.
in chronic disease patients. Psychol Assess 1990;2:22–30. 44. Lewis CC, Pantell RH, Kieckhefer GM. Assessment of chil-
34. Vivier PM, Bernier JA, Starfield B. Current approaches to dren’s health status. Field test of new approaches. Med Care
measuring health outcomes in pediatric research. Curr Opin 1989;27(3 Suppl):S54–65.
Pediatr 1994;6:530–7. 45. Varni JW, Waldron SA, Gragg RA, Rapoff MA, Bernstein BH,
35. Leplege A, Hunt S. The problem of quality of life in medi- Lindsley CB, Newcomb MD. Development of the Waldron/
cine. JAMA 1997;278:47–50. Varni Pediatric Pain Coping Inventory. Pain 1996;67:141–50.

Copyright 1999 by Turner White Communications Inc., Wayne, PA. All rights reserved.

40 JCOM April 1999 Vol. 6, No. 4