Validity and reliability of the Iranian version of the Pediatric Quality of Life Inventory(TM) 4.0 (PedsQL(TM)) Generic Core Scales in children.

0.73 MB·English

by  Amiri, Parisa

Amirietal.HealthandQualityofLifeOutcomes2012,10:3
http://www.hqlo.com/content/10/1/3
RESEARCH Open Access
Validity and reliability of the Iranian version of
™
the Pediatric Quality of Life Inventory 4.0
™
(PedsQL ) Generic Core Scales in children
Parisa Amiri1, Ghazaleh Eslamian1,2, Parvin Mirmiran1, Niloofar Shiva3, Mohammad Asghari Jafarabadi4 and
Fereidoun Azizi3*
Abstract
Background: This study aimed to investigate the reliability and validity of the Iranian version of the Pediatric
Quality of Life Inventory™ 4.0 (PedsQL™ 4.0) Generic Core Scales in children.
Methods: A standard forward and backward translation procedure was used to translate the US English version of
the PedsQL™ 4.0 Generic Core Scales for children into the Iranian language (Persian). The Iranian version of the
PedsQL™ 4.0 Generic Core Scales was completed by 503 healthy and 22 chronically ill children aged 8-12 years
and their parents. The reliability was evaluated using internal consistency. Known-groups discriminant comparisons
were made, and exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) were conducted.
Results: The internal consistency, as measured by Cronbach’s alpha coefficients, exceeded the minimum reliability
standard of 0.70. All monotrait-multimethod correlations were higher than multitrait-multimethod correlations. The
intraclass correlation coefficients (ICC) between the children self-report and parent proxy-reports showed moderate
to high agreement. Exploratory factor analysis extracted six factors from the PedsQL™ 4.0 for both self and proxy
reports, accounting for 47.9% and 54.8% of total variance, respectively. The results of the confirmatory factor
analysis for 6-factor models for both self-report and proxy-report indicated acceptable fit for the proposed models.
Regarding health status, as hypothesized from previous studies, healthy children reported significantly higher
health-related quality of life than those with chronic illnesses.
Conclusions: The findings support the initial reliability and validity of the Iranian version of the PedsQL™ 4.0 as a
generic instrument to measure health-related quality of life of children in Iran.
Keywords: Health-related quality of life, PedsQL™, Iran, Children
Background cognitive), and social health dimensions delineated by
Health-related quality of life (HRQOL) measures are the World Health Organization (WHO) [6].
increasingly being used in an effort to continually There are numerous of well-developed generic and
improve the quality of the healthcare for pediatric disease specific HRQOL measures for children and ado-
patient health in clinical trials [1], population health [2], lescents [7]. To integrate the merits of generic and dis-
clinical improvement [3], and among purchasers of ease-specific instruments for children and adolescents,
health care services [4]. Today, most descriptions of aged 2-18 years old, the Pediatric Quality of Life Inven-
HRQOL refer to it as a multidimensional construct [5] tory™ (PedsQL™) was designed and developed in the
that focuses on individuals’ subjective evaluation of their US [8]. The PedsQL™ 4.0 Generic Core Scales and dis-
physical, psychological (including emotional and ease-specific questionnaires have resulted from iterative
process and are applicable for healthy schools [9] and
community populations [10], as well as pediatric popula-
*Correspondence:[email protected]
3EndocrineResearchCenter,ResearchInstituteforEndocrineSciences, tions with acute [11] and chronic health conditions,
ShahidBeheshtiUniversityofMedicalSciences,Tehran,Iran such as cancer, cerebral palsy, diabetes, rheumatologic
Fulllistofauthorinformationisavailableattheendofthearticle
©2012Amirietal;licenseeBioMedCentralLtd.ThisisanOpenAccessarticledistributedunderthetermsoftheCreativeCommons
AttributionLicense(http://creativecommons.org/licenses/by/2.0),whichpermitsunrestricteduse,distribution,andreproductionin
anymedium,providedtheoriginalworkisproperlycited.
Amirietal.HealthandQualityofLifeOutcomes2012,10:3 Page2of9
http://www.hqlo.com/content/10/1/3
diseases, and end-stage renal disease [12-16]. The aged 8-12 years, were recruited from university hospitals
PedsQL™ 4.0 Generic Core Scales include child self- with identified chronic health conditions including
report and parent proxy-report forms and can be com- asthma (n = 3), renal failure (n = 8), and cancer (n =
pleted easily [10]; the US English version of the PedsQL 11). The study protocol was approved by the ethics
has been linguistically validated in many non-English- committee of the Obesity Research Center, Research
speaking countries [17-20]. Institute for Endocrine Sciences, Shahid Beheshti Uni-
Childhood is the crucial phase for overall develop- versity of Medical Sciences.
ment, including physical, psychological, and social devel-
opment, throughout an individual’s lifespan [21]. Measures
Health-related quality of life assessment for children PedsQL™ 4.0 Generic Core Scales
may be useful in targeting interventions and directing The 23-item PedsQL™ 4.0Generic CoreScales isaself-
resources to individuals and communities. Moreover, as administeredquestionnairethatincludeschildself-reports
cultural differences may exist in the assessment of andparentproxy-reports,whichencompassthefollowing
HRQOL, nation-specific information is required to subscales:PhysicalFunctioning(8items),EmotionalFunc-
enable national and international evaluation and tioning(5items),SocialFunctioning(5items)andSchool
benchmarking. Functioning(5items).A5-pointLikertresponsescaleran-
We have previously reported the initial reliability and gingfrom0(neveraproblem) to4(almostalwaysapro-
validity of the Iranian version of the PedsQL™ 4.0 as a blem) is used across child self-reports for ages 8-18 and
genericinstrumenttomeasureHRQOLofageneralpopu- parent proxy-reports. According to the manual of the
lation of Iranian adolescents, aged 13-18 years [22]. instrument,ifmorethan50%oftheitemsinthescaleare
Anotherstudyconductedonattentiondeficit/hyperactivity missing, the scale score is not computed. The total scale
disorders in Iranian children and adolescents, aged 8-17 scoresforbothchildself-reportandparent proxy-report
years, reported the psychometric properties of the were also calculated [8,10]. In addition to the PedsQL™
PedsQL™[23];giventhelimitedsamplesizeofthestudy 4.0questionnaires,allfamilieswererequiredtocompletea
mentionedandconsidering thatthe PedsQL™originally familyinformationformonsocio-demographicandchild
has two separate scales for children (8-12 years old) and healthcharacteristics.
adolescents(13-18yearsold)thatmakesasinglestatistical
analysis and conclusions difficult and vague, the current Procedure
study, aimed to investigate reliability and validity of the Translation
Iranian version of the PedsQL™ 4.0 generic core scale TheIranian(Persian)translationandlinguisticvalidation
among a large number of Iranian children, aged 8-12 of the PedsQL™ 4.0 questionnaire followed recom-
years. Basedon previous studies from international back mendedguidelines[24].Thisprocessincludedusingtwo
translationsofthePedsQL™4.0,wehypothesizedthatthe translators,whoareahealtheducatorandaclinical psy-
PedsQL™4.0couldalsodemonstratesatisfactorypsycho- chologist independently. To produce a conceptual
metricpropertiesinIranianchildrenandwouldhencedif- equivalence of the translation to the original English
ferentiateHRQOLbetweenahealthypediatricpopulation questionnaire, both translators discussed any disparities
andonewithchronichealthconditions. andagreedonasingleversion.Thebackwardtranslation
of the first reconciled forward version of the PedsQL™
Methods 4.0questionnairetotheoriginalU.S.Englishversionwas
Participants performedbytwolocalprofessionaltranslatorswhowere
Participants were 649 children, aged 8-12 years, who not associated with the first translation phase with
were recruited from primary and secondary schools in experience of living in English-speaking countries. In a
Tehran, and their parents. The inclusion criteria were pre-test, the PedsQL™ was given to 50 children and
obtaining of agreement from both the children and their their parents to ensure confidence in the linguistic and
parents, who were required to give their written conceptual equivalence of the translations. Cognitive
informed consent to participate. Overall 525 children interviewingtechniquewasalsousedtofindandcorrect
and their parents agreed to take part in the study, giving errors introduced through the translation process. The
a response rate of 80%. No significant differences were relevantchangesinthetranslationprocesswerereviewed
observed in age, gender, health status and their residen- forconceptualequivalenceandauthorizedbytheprinci-
tial area between participants and non-participants. paldeveloperofthePedsQL™(Dr.Varni).
Three hundred and thirty-two (63.2%) of the children
were girls and 503 (95.8%) were healthy (did not suffer Data collection
from any chronic health condition). All questionnaires Participants were selected from four primary and sec-
were completed anonymously. Twenty-two children, ondary schools, located in the north of Tehran. All the
Amirietal.HealthandQualityofLifeOutcomes2012,10:3 Page3of9
http://www.hqlo.com/content/10/1/3
schools were selected using stratified random sampling using exploratory factor analysis (EFA), utilizing princi-
methods, considering level of education and gender. pal component analysis and varimax rotation. To assess
Participants from two schools from each sex and level how well the EFA extracted model fits observed data,
were recruited for the study. Trained research personnel we conducted confirmatory factor analysis (CFA), using
visited each classroom and distributed a package includ- the method of weighted least squares for estimation.
ing a written consent form, cover letter, family informa- Asymptomatic covariance matrix was considered a
tion form and the PedsQL™ for the parents to fill out weighted matrix. Input matrix was covariance matrix of
at home. The cover letter explained the study and guar- data. Fit indices and reasonable values of these indices
anteed the confidentiality of data, assuring that even the for CFA were considered as c2/df < 5, Root Mean
school staff would not see the information. The partici- Square Error of Approximation (RMSEA) < 0.08 and
pants could contact the researchers to get further infor- also, Comparative Fit Index (CFI), Goodness of Fit
mation and guidance. After the research team had Index (GFI), Adjusted Goodness of Fit Index (AGFI) >
collected the questionnaires which were returned to the 0.9 [32]. Given previous PedsQL™ CFI findings, 5- and
school, project staff revisited each class and adminis- 6-factor models were tested [33-35]. Construct validity
tered the PedsQL™ 4.0 to those children, whose parents was tested performing the known-groups method which
had completed the questionnaires at home and signed compares scale scores across groups known to differ in
consent forms. Subjects who suffered chronic health the health construct being surveyed. We hypothesized
conditions were recruited from two university hospitals. that healthy children would report higher scores than
After receiving informed consents from parents, the children with a chronic health condition. Student’s t test
questionnaires were completed by children and their was performed to determine gender differences between
parents separately. Trained research personnel assisted parent proxy-report and child self-report. Statistical ana-
participants in completing the questionnaires. lysis was performed using SPSS 15.0 (SPSS Inc., Chi-
cago, IL) and LISREL 8.80 (Scientific Software
Statistical analysis International Inc., 2007). P-Values less than 0.05 were
The total score of each scale was computed by summing considered significant.
up items related to the scale and used in the analysis.
The data were presented as “Mean ± SD” for the vari- Results
ables. To determine whether univariate normality exists, Parent proxy-reports were completed by 397 (76.2%)
we examined the distribution of each observed variable mothers, 114 (21.9%) fathers and 10 (1.9%) by other
for skewness and kurtosis. For the skewness index, abso- caregivers such as grandparents. Missing responses for
lute values above than 3.0 are extreme [25]. Absolute items were rare and ranged from 0.0 to 1.9 percent for
values higher than 10.0 for the kurtosis index, suggest a both the child self-report and parent proxy-report. In
problem, [26]. the chronic health condition sample, missing responses
The feasibility of the Iranian version of the PedsQL™ ranged from 0.0 to 4.5 percent.
4.0 was determined based on the percentage of missing The internal consistency of the scale as measured by
values for each item. Ceiling and floor effects were eval- Cronbach’s alpha coefficients showed that all child self-
uated based on percentage of scores at the extremes of and parent proxy-report subscales of the PedsQL™ 4.0
the scaling range [27]. Floor or ceiling effects are con- exceeded the minimum reliability standard of 0.70,
sidered to be present if more than 15% of respondents except for emotional functioning for children and school
achieve the lowest or highest possible score, respectively functioning for both respondents. The total scale scores
[28]. Internal consistency (to test reliability) was internal consistency alphas were 0.84 and 0.88 for child
assessed by calculating Cronbach’s alpha (a) coefficient self-report and parent proxy-report, respectively. No
[29]. Alpha coefficients equal to or greater than .70 were floor effects were observed while ceiling effects detected
considered satisfactory. We computed the intraclass cor- ranged from 2.9%, for self-report total score, to 37.7%
relation coefficient (ICC) to evaluate child self-report for self-report social functioning (Table 1).
and parent proxy-report agreement on the PedsQL™ All monotrait-multimethod correlations demonstrated
4.0 subscales. ICCs ≤ 0.4 were considered poor to fair a moderate relation between child self-report subscales
agreement; 0.41-0.60 moderate agreement; 0.61-0.80 and parent proxy-report subscales, ranging from 0.37 to
good agreement and > 0.80 excellent agreement [30]. 0.43 (Table 2). All multitrait-multimethod correlations
The multitrait-multimethod was used to compute par- were lower than the monotrait-multimethod correla-
ent-child Pearson intercorrelations between and among tions. The average convergent correlation was 0.40 and
PedsQL™ 4.0 subscales. Correlations are designated as the average off-diagonal correlation was 0.30. The intra-
small (0.10-0.29), medium (0.30-0.49), and large (≥ 0.50) class correlation coefficients (ICC) were moderate to
[31]. Factor structure of the PedsQL™ 4.0 was extracted high for all scales, indicating good agreement between
Amirietal.HealthandQualityofLifeOutcomes2012,10:3 Page4of9
http://www.hqlo.com/content/10/1/3
Table 1Means, standard deviations, percent floor andceiling effects,andCronbach’sa forthe Iranian version of
PedsQL™4.0generic corescales (n=525)
Mean SD Skewness Kurtosis Percent Percent Cronbach’sa
floor ceiling
Childself-report
Totalscore 83.99 11.90 -1.26 1.76 0 2.9 0.84
Physicalfunctioning 85.29 13.95 -1.44 4.32 0 5.5 0.70
Emotionalfunctioning 80.55 16.75 -0.92 0.54 0 5.9 0.64
Socialfunctioning 86.88 16.01 -1.44 3.18 0 37.7 0.70
Schoolfunctioning 83.26 14.88 -1.36 3.35 0.2 19.6 0.60
Parentproxy-report
Totalscore 76.75 13.45 -0.31 -0.15 0 3.0 0.88
Physicalfunctioning 77.52 18.48 -0.74 0.47 0 17.7 0.83
Emotionalfunctioning 70.95 17.37 -0.22 -0.74 0 5.7 0.70
Socialfunctioning 80.77 17.89 -0.83 0.10 0 25.3 0.74
Schoolfunctioning 77.76 15.77 -0.47 0.19 0.2 13.9 0.63
Floororceilingeffectsareconsideredtobepresentifmorethan15%ofrespondentsachievethelowestorhighestpossiblescore,respectively.
child and parent reports except for the total scale scores, factor consisted of items 5 to 8. Similarly, the school
which showed a reasonable agreement (ICC = 0.67, 95% functioning scale was split into two factors; in the par-
CI: 0.61-0.72). Lowest agreement was detected for the ent and child reports, the first factor consisted of items
social functioning scale (ICC = 0.55, 95% CI: 0.46-0.62). 1 to 3 and second factor consisted of items 4 and 5
Exploratory factor analysis with varimax rotation (Table 3). The results of the CFA for 5- and 6-factor
extracted six factors from the PedsQL™ 4.0 for both models for self- and parent proxy-reports indicated a
self- and proxy- reports, accounting for 47.9 and 54.8% more acceptable fit for the 6-factor model. In addition
of total variance, respectively. Kaser-Meier-Olkin (KMO) all parameters relating to the factors and indicators were
values of 0.71 and 0.74 respectively for children and par- statistically significant (All P < 0.05) (Table 4).
ents EFA and P < 0.001 of Bartelet test of sphericity for There were statistically significant differences between
both children and parents confirmed the adequacy of healthy and chronically ill children for all subscales,
the factor model as well. The two physical functioning where children with chronic health conditions reported
items measuring pain and fatigue were split into a dif- lower scores than did healthy children (Figure 1). Also
ferent factor for both the child self-reports and the par- among healthy participants, a gender difference was
ent proxy-reports. In the parent-report, the first factor found in all subscales and total scores for child self-
consisted of items 1 to 6 and the second factor con- report. Compared to boys, girls scored higher on the
sisted of items 7 and 8; however in the children report, total scale score (86.7 ± 9.9 vs. 82.02 ± 11.2, P < 0.01),
the first factor consisted of items 1 to 4 and the second physical functioning (88.1 ± 13.1 vs. 84.2 ± 11.7, P <
Table 2Intercorrelations ofchildren self-report andparentproxy-report for the Iranianversion ofPedsQL™4.0
Generic Core Scalesby the Multitrait-Multimethod Matrix (n =525)
Childself-report Parentproxy-report
Physical Emotional Social School Physical Emotional Social
Childself-report
Physicalfunctioning
Emotionalfunctioning 0.42
Socialfunctioning 0.52 0.39
Schoolfunctioning 0.50 0.43 0.51
Parentproxy-report
Physicalfunctioning 0.40 0.25 0.31 0.29
Emotionalfunctioning 0.29 0.39 0.28 0.33 0.47
Socialfunctioning 0.28 0.23 0.37 0.28 0.53 0.42
Schoolfunctioning 0.20 0.25 0.23 0.43 0.46 0.45 0.42
AllcorrelationsweresignificantatP<0.001
Multitrait-monomethodcorrelationsareinbold;Monotrait-multimethodcorrelationsareunderlined;Multitrait-multimethodcorrelationsareitalicized.
Amirietal.HealthandQualityofLifeOutcomes2012,10:3 Page5of9
http://www.hqlo.com/content/10/1/3
Table 3Factor analysisresults forchild self-report (left) andthe parent proxy-report (right) forthe Iranianversion of
the PedsQL ™4.0Generic Core Scales
Factor1 Factor2 Factor3 Factor4 Factor5 Factor6
Physicalfunctioning
1.Itishardformetowalkmorethanoneblock 0.78 0.71 0.01 0.30
2.Itishardformetorun 0.73 0.73 0.22 0.38
3.Itishardformetodosportsactivitiesorexercise 0.56 0.78 0.39 0.26
4.Itishardformetoliftsomethingheavy 0.53 0.39 0.05 0.52
5.Itishardformetotakeabathorshowerbymyself -0.07 0.77 0.77 0.02
6.Itishardformetodochoresaroundthehouse 0.14 0.71 0.57 0.14
7.Ihurtorache 0.16 0.15 0.49 0.78
8.Ihavelowenergy 0.46 0.12 0.51 0.81
Emotionalfunctioning
1.Ifeelafraidorscared 0.68 0.70
2.Ifeelsadorblue 0.61 0.76
3.Ifeelangry 0.66 0.68
4.Ihavetroublesleeping 0.61 0.59
5.Iworryaboutwhatwillhappentome 0.64 0.61
Socialfunctioning
1.Ihavetroublegettingalongwithotherkids 0.69 0.76
2.Otherkidsdonotwanttobemyfriend 0.73 0.80
3.Otherkidsteaseme 0.59 0.66
4.Icannotdothingsotherkidsmyagecando 0.68 0.72
5.Itishardtokeepupwithmypeers 0.63 0.55
Schoolfunctioning
1.Itishardtopayattentioninclass 0.80 0.84 0.00 0.00
2.Iforgetthings 0.71 0.72 0.04 0.12
3.Ihavetroublekeepingupwithmyschoolwork 0.62 0.75 0.28 0.14
4.Imissschoolbecauseofnotfeelingwell 0.01 0.10 0.87 0.84
5.Imissschooltogotothedoctororhospital 0.18 0.09 0.81 0.84
0.001), emotional functioning (79.2 ± 16.1 vs. 82.8 ± Our study presents, the feasibility of PedsQL™ 4.0, as
15.9, P = 0.01), social functioning (90.4 ± 14.0 vs. 82.8 ± measured by a low percent of missing values, particu-
15.9, P < 0.001) and school functioning (85.5 ± 12.7 vs. larly in healthy children. Similar to previous studies, no
81.7 ± 15.0, P = 0.003) (Figure 2). floor effects were found in either the self-report or par-
ent proxy versions [8,17,22,36]. Most subscales in this
Discussion study indicated some ceiling effects, which support
This study investigated the psychometric properties of results of previous studies [18,22,36].
the Iranian version of the PedsQL™ 4.0 Generic Core Cronbach’s a coefficient to test reliability were accep-
Scales in children aged 8-12 years. Our results indicate table (exceeded .70) for all measures and showed strong
the preliminary reliability and validity of the Iranian ver- internal consistency reliability for the total scale, and
sion of PedsQL™ 4.0 as a child self-report and parent most subscales including physical, emotional and social
proxy-report measure of generic HRQOL in Iran. functioning in both children and parents. This
Table 4Goodnessoffit indices for 5-and6-factor CFA models ofparent andchild reports (n=525)
Model x2 df x2/df RMSR RMSEA(90%CI) CFI GFI AGFI
Parents-6factormodel 685.81* 211 3.25 0.058 0.066(0.060;0.071) 0.95 0.90 0.87
Parents-5factormodel 1261.72* 220 5.74 0.065 0.075(0.071;0.079) 0.94 0.89 0.86
Children-6factormodel 609.74* 214 2.85 0.052 0.059(0.054;0.065) 0.94 0.91 0.88
Children-5factormodel 1242.72* 220 5.65 0.061 0.074(0.070;0.078) 0.94 0.89 0.86
CFA,confirmatoryfactoranalysis;x2,chi-square;df,degreesoffreedom;x2/df,normedchi-square;RMSR,rootmeansquareresidual;RMSEA,rootmeansquare
errorofapproximation;CFI,comparativefitindex;GFI,goodnessoffitindex;AGFI,adjustedgoodnessoffitindex
*P<0.001
Amirietal.HealthandQualityofLifeOutcomes2012,10:3 Page6of9
http://www.hqlo.com/content/10/1/3
Healthy(n=496)
Chronically Ill(n=22)
100 a 100 b
90 * * * 90
80 * * 80 * * * *
*
70 70
60 60
50 50
40 40
30 30
20 20
10 10
0 0
Figure1ComparisonofPedsQL™4.0scoresbetweenhealthyandchronicallyillchildren.a)ChildSelf-report,b)Parent-proxyreport,
*P<0.01.
satisfactory level of internal consistency is almost similar multimethod correlations, providing evidence for the
to the original version and other translated versions validity of the instrument’s dimensions. In general, there
[8,19,37]. was a good agreement between children and parent
The multitrait-monomethod correlations were med- reports except for the total scale scores which had also
ium for child self-report and for parent proxy-report. been observed in our previous study [22] and could
Our results indicated that the multitrait-multimethod have been due to strong parental support in Iranian
correlations were smaller than the monotrait- families. These findings are inconsistent with the
Boys(n=180)
Girls(n=316)
100 a 100 b
* *
90 * * * 90 *
80 80 * * * *
70 70
60 60
50 50
40 40
30 30
20 20
10 10
0 0
Figure2ComparisonofPedsQL™4.0scoresbetweenboysandgirls.a)ChildSelf-report,b)Parent-proxyreport,*P<0.01.
Amirietal.HealthandQualityofLifeOutcomes2012,10:3 Page7of9
http://www.hqlo.com/content/10/1/3
original version and most of the other translated ver- Iranian version of the PedsQL™ 4.0 in children aged 8-
sions [17,18,38]. 12 years.
Factor validity of the scales for parent proxy-reports Previous researches have presented gender differences
and child self-reports was determined through factor in the pediatric HRQOL [34,40]. Our results demon-
analysis. According to our results, the 6-factor model strated that girls had significantly higher HRQOL in
showed much a better fit than the 5-factor model based total scale score, physical functioning, emotional func-
on the good of fitness (GOF) indices from CFA. In addi- tioning, social functioning and school functioning than
tion based on the criteria used (Scree test) and theoreti- boys, by self-report. The difference between girls and
cal consideration, the results of EFA on this set of data, boys reflected in our data support the results of Chen et
were best fitted with a 6-factor solution. However our al, who used the PedsQL™ 15-item short form for ado-
current EFA/CFA findings are not in agreement with lescent girls and their parents in Japan [17]. However,
the results of our previous study on the Iranian version our results were not consistent with our previous study
of PedsQL Generic Core Scales in adolescents [22], or and current literature [19,22,36,40]. The difference
with those from another study conducted to assess the between adolescents in our previous study and children
PedsQL™ Oral Health Scale in Iranian children [39] in this study may be due to the different perception of
findings which both support a priori 5-factor model; the social and environmental pressure and also to puberty
EFA/CFA findings of yet another study of a much smal- changes in adolescents.
ler sample of Iranian children and adolescents with The Iranian version of PedsQL™ 4.0 was able to
attention deficit/hyperactivity disorder ADHD, showed detect the hypothesized differences between healthy and
an acceptable fit of a 4-factor model [23]. chronically ill children supporting the initial discrimina-
In this study, for child self-report, the items related to tion validity of the Iranian version of the PedsQL™ 4.0
physical functioning were loaded in two separate factors for children. Consistent with other studies [8,22,41],
each containing 4 items. Item 5 to 8 were loaded to an children with chronic health conditions had lower
independent factor (factor 2). All items related to emo- HRQOL scores than healthy children.
tional functioning and social functioning had a clear fac- Our study has several limitations. The unavailability of
tor loading. For parent proxy-report, two items related information on nonrespondents and investigation of
to physical functioning (items 7 and 8) were loaded to only a sample from Tehran, the capital city of Iran, may
other factors. One possible explanation for the discre- limit the generalizability of the findings to other parts of
pancy between physical functioning factor of children Iran. Further Iranian PedsQL™ studies in the other
and their parents is their different perception of the regions in Iran are necessary. Retest reliability and
construction of the mentioned items. Hence, the loading responsiveness was not examined. However, it has been
of the first four items including, walking more than one debated that test-retest reliability may be less useful
block, running, doing sports activities or exercise and than internal consistency reliability in HRQL instrument
lifting something heavy, either in children or in their development [42]. Finally, the ceiling effects that been
parents, were similar and were included in the same fac- observed on some scales may limit the ability of the Ira-
tor. All the items mentioned above are related to physi- nian version of PedsQL™ 4.0 to detect HRQOL
cal activity and were interpreted by both children and improvement in some scales for healthy children.
their parents in a similar manner. On the other hand, Whether these ceiling effects are evident for chronically
children and their parents did not have the same inter- ill populations in Iran will require further evaluation
pretation for the last four items including, taking a bath with larger populations of ill children.
or shower by myself, doing chores around the house,
hurting or aching and have low energy. This difference Conclusion
may be due to the individual perception and vision of Our study demonstrates the initial reliability and validity
children regarding these items, which include responsi- of the Iranian version of the PedsQL™ 4.0 Generic
bility, independency as well as tolerability, and may Cores Scales as an outcome measure of generic HRQOL
explain the last two items about parents. in Iranian children aged 8-12 years old. Future research
All items related to emotional functioning and social is needed with larger samples of chronically ill children
functioning had a clear factor loading in both self- and aged 8-12 years, and samples of children aged 2-7 years.
parent proxy-reports. The last two items of school func-
tioning for both self- and parent proxy-report were
Abbreviations
loaded on a separate factor that was very similar to the
HRQOL:Health-RelatedQualityofLife;PedsQL™4.0:PediatricQualityofLife
EFA findings for the original US English version. To Inventory™4.0;EFA:ExploratoryFactorAnalysis;CFA:ConfirmatoryFactor
confirm the results of our EFA, we utilized CFA. Our Analysis
results support the initial construct validity of the
Amirietal.HealthandQualityofLifeOutcomes2012,10:3 Page8of9
http://www.hqlo.com/content/10/1/3
Acknowledgements andsensitivityoftheGenericCoreScalesandCerebralPalsyModule.
Weacknowledgethecontributionofallfieldworkerswhoconductedthe Developmentalmedicineandchildneurology2006,48(6):442-449.
datacollection. 14. VarniJW,LimbersCA,BryantWP,WilsonDP:ThePedsQL
MultidimensionalFatigueScaleintype1diabetes:feasibility,reliability,
Authordetails andvalidity.Pediatricdiabetes2009,10(5):321-328.
1ObesityResearchCenter,ResearchInstituteforEndocrineSciences,Shahid 15. VarniJW,SeidM,SmithKnightT,BurwinkleT,BrownJ,SzerIS:ThePedsQL
BeheshtiUniversityofMedicalSciences,Tehran,Iran.2StudentResearch inpediatricrheumatology:reliability,validity,andresponsivenessofthe
Committee,ShahidBeheshtiUniversityofMedicalSciences,Tehran,Iran. PediatricQualityofLifeInventoryGenericCoreScalesand
3EndocrineResearchCenter,ResearchInstituteforEndocrineSciences, RheumatologyModule.Arthritisandrheumatism2002,46(3):714-725.
ShahidBeheshtiUniversityofMedicalSciences,Tehran,Iran.4NationalPublic 16. GoldsteinSL,RosburgNM,WaradyBA,SeikalyM,McDonaldR,LimbersC,
HealthManagementCenter(NPMC)andDepartmentofStatisticsand VarniJW:Pediatricendstagerenaldiseasehealth-relatedqualityoflife
Epidemiology,FacultyofHealthandNutrition,TabrizUniversityofMedical differsbymodality:aPedsQLESRDanalysis.PediatrNephrol2009,
Sciences,Tabriz,Iran. 24(8):1553-1560.
17. ChenX,OrigasaH,IchidaF,KamibeppuK,VarniJW:Reliabilityandvalidity
Authors’contributions ofthePediatricQualityofLifeInventory(PedsQL)ShortForm15
PAdesignedthestudy,acquired,interpretedthedata,draftedthe GenericCoreScalesinJapan.Qualityofliferesearch:aninternational
manuscript,andwrotethepaper.GEassistedwithdraftingthemanuscript journalofqualityoflifeaspectsoftreatment,careandrehabilitation2007,
anddealtwiththeanalysisprocess.PMassistedindatainterpretationand 16(7):1239-1249.
reviewedthepaper.ForvalidationoftheIranianversionofthePedsQL™ 18. GkoltsiouK,DimitrakakiC,TzavaraC,PapaevangelouV,VarniJW,TountasY:
questionnairebacktranslationwasdonebyNSwhoalsocriticallyreviewed Measuringhealth-relatedqualityoflifeinGreekchildren:psychometric
thepaper,andeditedallrevisionsofthemanuscript.MAJcontributedto propertiesoftheGreekversionofthePediatricQualityofLifeInventory
thedataanalysisandinterpretation.FAsupervisedandadvisedthroughout (TM)4.0GenericCoreScales.Qualityofliferesearch:aninternational
thestudy.Allauthorsreadandapprovedthefinalmanuscript. journalofqualityoflifeaspectsoftreatment,careandrehabilitation2008,
17(2):299-305.
Competinginterests 19. ReinfjellT,DisethTH,VeenstraM,VikanA:Measuringhealth-related
Theauthorsdeclarethattheyhavenocompetinginterests. qualityoflifeinyoungadolescents:reliabilityandvalidityinthe
NorwegianversionofthePediatricQualityofLifeInventory4.0(PedsQL)
Received:30October2011 Accepted:5January2012 genericcorescales.Healthandqualityoflifeoutcomes2006,4:61.
Published:5January2012 20. BerkesA,PatakiI,KissM,KemenyC,KardosL,VarniJW,MogyorosyG:
Measuringhealth-relatedqualityoflifeinHungarianchildrenwithheart
disease:psychometricpropertiesoftheHungarianversionofthe
References
PediatricQualityofLifeInventory4.0GenericCoreScalesandthe
1. SpilkerB:Qualityoflifeandpharmacoeconomicsinclinicaltrials.
CardiacModule.Healthandqualityoflifeoutcomes2010,8:14.
Philadelphia:Lippincott-Raven;,21996.
21. HetheringtonEM,LernerRM,PerlmutterM,SocialScienceResearchCouncil
2. ZahranHS,KobauR,MoriartyDG,ZackMM,HoltJ,DonehooR:Health-
relatedqualityoflifesurveillance–UnitedStates,1993-2002.MMWR (U.S.):Childdevelopmentinlife-spanperspective.Hillsdale,N.J:Lawrence
ErlbaumAssociates;1988.
Surveillancesummaries:MorbidityandmortalityweeklyreportSurveillance
22. AmiriP,MAE,Jalali-FarahaniS,HosseinpanahF,VarniJW,GhofranipourF,
summaries/CDC2005,54(4):1-35.
MontazeriA,AziziF:ReliabilityandvalidityoftheIranianversionofthe
3. GanzPA:Impactofqualityoflifeoutcomesonclinicalpractice.Oncology
PediatricQualityofLifeInventory4.0GenericCoreScalesin
(WillistonPark)1995,9(11Suppl):61-65.
adolescents.Qualityofliferesearch:aninternationaljournalofqualityoflife
4. FayersPM,MachinD:Qualityoflife:theassessment,analysis,and
aspectsoftreatment,careandrehabilitation2010,19(10):1501-1508.
interpretationofpatient-reportedoutcomes.Chichester,Hoboken,NJ:J.
23. JafariP,GhanizadehA,AkhondzadehS,MohammadiMR:Health-related
Wiley;,22007.
qualityoflifeofIranianchildrenwithattentiondeficit/hyperactivity
5. PalDK:Qualityoflifeassessmentinchildren:areviewofconceptualand
disorder.Qualityofliferesearch:aninternationaljournalofqualityoflife
methodologicalissuesinmultidimensionalhealthstatusmeasures.
aspectsoftreatment,careandrehabilitation2011,20(1):31-36.
Journalofepidemiologyandcommunityhealth1996,50(4):391-396.
24. LinguisticvalidationofthePedsQL-aQualityofLifeQuestionnaire.
6. WorldHealthOrganization:ConstitutionoftheWorldHealth
[http://www.pedsql.org/translations.html].
Organization:Basicdocument.Geneva,Switzerland.1948.
25. ChouC-P,BentlerPM:Estimatesandtestsinstructuralequation
7. VarniJW,BurwinkleTM,LimbersCA,SzerIS:ThePedsQLasapatient-
modeling.InStructuralequationmodeling:Concepts,issues,andapplications.
reportedoutcomeinchildrenandadolescentswithfibromyalgia:an
Editedby:HoyleRH.ThousandOaks,CA,US:SagePublications,Inc;
analysisofOMERACTdomains.Healthandqualityoflifeoutcomes2007,
1995:37-55.
5:9.
26. KlineRB:Principlesandpracticeofstructuralequationmodeling.New
8. VarniJW,SeidM,KurtinPS:PedsQL4.0:reliabilityandvalidityofthe
York:GuilfordPress;,22005.
PediatricQualityofLifeInventoryversion4.0genericcorescalesin
27. McHorneyCA,WareJEJr,LuJF,SherbourneCD:TheMOS36-itemShort-
healthyandpatientpopulations.Medicalcare2001,39(8):800-812.
FormHealthSurvey(SF-36):III.Testsofdataquality,scaling
9. Felder-PuigR,BaumgartnerM,TopfR,GadnerH,FormannAK:Health-
assumptions,andreliabilityacrossdiversepatientgroups.Medicalcare
relatedqualityoflifeinAustrianelementaryschoolchildren.Medical
1994,32(1):40-66.
care2008,46(4):432-439.
28. McHorneyCA,TarlovAR:Individual-patientmonitoringinclinicalpractice:
10. VarniJW,BurwinkleTM,SeidM,SkarrD:ThePedsQL4.0asapediatric
areavailablehealthstatussurveysadequate?Qualityofliferesearch:an
populationhealthmeasure:feasibility,reliability,andvalidity.Ambulatory
internationaljournalofqualityoflifeaspectsoftreatment,careand
pediatrics:theofficialjournaloftheAmbulatoryPediatricAssociation2003,
rehabilitation1995,4(4):293-307.
3(6):329-341.
29. CronbachL:Coefficientalphaandtheinternalstructureoftests.
11. MistryRD,StevensMW,GorelickMH:Health-relatedqualityoflifefor
Psychometrika1951,16(3):297-334.
pediatricemergencydepartmentfebrileillnesses:anevaluationofthe
30. BartkoJJ:Theintraclasscorrelationcoefficientasameasureofreliability.
PediatricQualityofLifeInventory4.0GenericCoreScales.Healthand
Psychologicalreports1966,19(1):3-11.
qualityoflifeoutcomes2009,7:5.
31. CohenJ:Statisticalpoweranalysisforthebehavioralsciences.Hillsdale,
12. VarniJW,BurwinkleTM,KatzER,MeeskeK,DickinsonP:ThePedsQLin
N.J:L.ErlbaumAssociates;,21988.
pediatriccancer:reliabilityandvalidityofthePediatricQualityofLife
32. BentlerPM,BonettDG:Significancetestsandgoodnessoffitinthe
InventoryGenericCoreScales,MultidimensionalFatigueScale,and
analysisofcovariancestructures.PsychologicalBulletin1980,88(3):588-606.
CancerModule.Cancer2002,94(7):2090-2106.
33. LimbersCA,NewmanDA,VarniJW:Factorialinvarianceofchildself-
13. VarniJW,BurwinkleTM,BerrinSJ,ShermanSA,ArtaviaK,MalcarneVL,
reportacrosshealthyandchronichealthconditiongroups:a
ChambersHG:ThePedsQLinpediatriccerebralpalsy:reliability,validity,
Amirietal.HealthandQualityofLifeOutcomes2012,10:3 Page9of9
http://www.hqlo.com/content/10/1/3
confirmatoryfactoranalysisutilizingthePedsQLTM4.0GenericCore
Scales.Journalofpediatricpsychology2008,33(6):630-639.
34. VarniJW,LimbersC,NewmanD:FactorialInvarianceofthePedsQL™4.0
GenericCoreScalesChildSelf-ReportAcrossGender:AMultigroup
ConfirmatoryFactorAnalysiswith11,356ChildrenAges5to18.Applied
ResearchinQualityofLife2008,3(2):137-148.
35. LimbersCA,NewmanDA,VarniJW:Factorialinvarianceofchildself-
reportacrossagesubgroups:aconfirmatoryfactoranalysisofages5to
16yearsutilizingthePedsQL4.0GenericCoreScales.Valueinhealth:the
journaloftheInternationalSocietyforPharmacoeconomicsandOutcomes
Research2008,11(4):659-668.
36. UptonP,EiserC,CheungI,HutchingsHA,JenneyM,MaddocksA,
RussellIT,WilliamsJG:MeasurementpropertiesoftheUK-Englishversion
ofthePediatricQualityofLifeInventory4.0(PedsQL)genericcore
scales.Healthandqualityoflifeoutcomes2005,3:22.
37. ScarpelliAC,PaivaSM,PordeusIA,Ramos-JorgeML,VarniJW,AllisonPJ:
MeasurementpropertiesoftheBrazilianversionofthePediatricQuality
ofLifeInventory(PedsQL)cancermodulescale.Healthandqualityoflife
outcomes2008,6:7.
38. CremeensJ,EiserC,BladesM:Factorsinfluencingagreementbetween
childself-reportandparentproxy-reportsonthePediatricQualityof
LifeInventory4.0(PedsQL)genericcorescales.Healthandqualityoflife
outcomes2006,4:58.
39. PakpourAH,YekaninejadMS,ZareiF,HashemiF,SteeleMM,VarniJW:The
PedsQLOralHealthScaleinIranianchildren:reliabilityandvalidity.
Internationaljournalofpaediatricdentistry/theBritishPaedodonticSociety
[and]theInternationalAssociationofDentistryforChildren21(5):342-352.
40. BiseggerC,CloettaB,vonRuedenU,AbelT,Ravens-SiebererU:Health-
relatedqualityoflife:genderdifferencesinchildhoodandadolescence.
Sozial-undPraventivmedizin2005,50(5):281-291.
41. EiserC,MorseR:Areviewofmeasuresofqualityoflifeforchildrenwith
chronicillness.Archivesofdiseaseinchildhood2001,84(3):205-211.
42. VarniJW,BurwinkleTM,JacobsJR,GottschalkM,KaufmanF,JonesKL:The
PedsQL™inType1andType2Diabetes.Diabetescare2003,
26(3):631-637.
doi:10.1186/1477-7525-10-3
Citethisarticleas:Amirietal.:ValidityandreliabilityoftheIranian
versionofthePediatricQualityofLifeInventory™™4.0(PedsQL™™)
GenericCoreScalesinchildren.HealthandQualityofLifeOutcomes2012
10:3.
Submit your next manuscript to BioMed Central
and take full advantage of:
• Convenient online submission
• Thorough peer review
• No space constraints or color figure charges
• Immediate publication on acceptance
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution
Submit your manuscript at
www.biomedcentral.com/submit