QM

Obco(Can

StQu

Restareppu

Coattthe

Ttre

bymuarethekmdieinvcli

toimtiorecan

comlifeinf

ARICFCFCCFO

508UALITY-OF-LIFE IN CHILDREN AND ADOLESCENTS WITH CYSTIC FIBROSISANAGED IN BOTH REGIONAL OUTREACH AND CYSTIC FIBROSIS CENTER

SETTINGS IN QUEENSLAND

CLARE THOMAS, MBBS, PENNY MITCHELL, BN, PETER OROURKE, BSC, BA, PHD, GCED, AND CLAIRE WAINWRIGHT, MBBS, MD

jectives To assess the health-related quality-of-life (HRQOL) of children/adolescents with cystic fibrosis (CF) andmpare HRQOL in children managed by cystic fibrosis outreach service (CFOS) with those treated in a cystic fibrosis centerFC). To compare HRQOL of children with CF in Queensland with previously published HRQOL data from the United Statesd examine the relationship between HRQOL scores and pulmonary function.

udy design Participants were children/adolescents with CF and their parents managed by the Royal Childrens Hospitaleensland at a CFC or CFOS. Two HRQOL surveys were used: PedsQLTM and Cystic Fibrosis Questionnaire (CFQ).

sults There were 91 CFC and 71 CFOS participants with similar demographics. PedsQLTM total summary score wastistically higher in CFOS, P .05. There was no significant difference in CFQ scores between groups. Queensland parentsorted lower HRQOL for their children compared with US parents (P

HRQOL using The Child Health Questionnaire in childrenfrom lower socioeconomic contexts in Australia,6 and Cam-eron et al found reduced HRQOL using the same tool inchare

chcha geastioreglatsurcoUn

theWteadrediacomgoAutocensibRe

rectalgroFFnoyeahig12col

pityea

HR

HRHRval

Pe

all

self-report includes questionnaires for ages 5 to 7, 8 to 12, and13 to 18 years. Parent proxy-report includes questionnairesfor ages 2 to 4 (toddler), 5 to 7 (young child), 8 to 12 (child),ancepanfoutiores(agThavabetscothethecattheingsumoffunmaToscaforan

CF

sisyeaagePaChdifphfunCFdeThverrectheHRiteanthecat

Da

peanrou

Qu nagedIn eensildren with diabetes living in regional compared with urbanas in Victoria (Australia).7,8

We hypothesized that HRQOL may be reduced inildren with CF from regional areas as was found inildren with diabetes. This study was designed using botheneric HRQOL measure, the PedsQLTM,9 and a dis-e-specific HRQOL measure, the Cystic Fibrosis Ques-nnaire (CFQ),10,11 to examine HRQOL in urban andional children and adolescents with CF and the corre-ion between the HRQOL and disease severity as mea-ed by change in pulmonary function, and to enablemparison of HRQOL with published data from theited States10,12 and Australia.13

METHODSThe participants were children and adolescents, and

ir parents, living in Queensland and Northern New Southales who were treated by the Royal Childrens Hospital CFm in a tertiary (CFC) or outreach setting (CFOS). Chil-n were between 2 and 19 years of age and had a provengnosis of CF based on genetic and/or sweat testing andpatible clinical features. Geographical location was cate-

rized according to the Accessibility/Remoteness Index ofstralia (ARIA).14 ARIA classification uses road distancescalculate remoteness and accessibility to general serviceters, and it is grouped into five categories: Highly Acces-le, Moderately Accessible, Accessible, Remote, and Verymote.

Demographic details were collected from medicalords or available pathology databases including the neona-screening records. CF genotypes for all children wereuped as homozygous for the F508 mutation (F508/508), heterozygous F508 mutation (F508/other), no508 mutation (other/other), or not available. Socioeco-mic status (SES) was available for a cohort of 6- to 13-r-olds as collected on the CFQ. It was derived from thehest level of maternal education and categorized into gradeor less, completed grade 12/diploma/trade, and professional/lege degree.

Approval was obtained from the Royal Childrens Hos-al and District Ethics committee. Parents and children 8rs of age gave written consent for the study.

QOL QuestionnairesTwo HRQOL surveys were administered; a generic

QOL measure, PedsQLTM, and a disease-specificQOL measure, the CFQ. Both have been previouslyidated and tested for reliability.10,15

dsQLTM

The PedsQLTM is a generic measure comprising par-el child self-report and parent proxy-report formats. Child

ality-Of-Life In Children And Adolescents With Cystic Fibrosis MaBoth Regional Outreach And Cystic Fibrosis Center Settings In Qud 13 to 18 years (adolescent), and it assesses parents per-tions of the childs HRQOL.15 Both the child self-reportd the parent proxy-report comprise 23 items, measuringr multidimensional generic scales (Physical, Social, Emo-nal, and School), based on a 4-week recall. A 5-pointponse scale is used for all but the young child self-reportes 5 to 7 years), which only has a 3-point response scale.e PedsQLTM was scored according to the instructionsilable on the website,16 with higher scores representing ater HRQOL score. Any missing item was given the meanre of items completed for that domain if more than half ofanswers for that specific domain were available; otherwisey were left blank and recorded as missing.16 Results areegorized into four generic core scales (mean of the sum ofitems for each Physical functioning, Emotional function-, Social functioning, and School functioning) and threemary scores: Psychosocial Health Summary Score (meanthe sum of the items of Emotional functioning, Socialctioning, and School functioning), Physical Health Sum-ry Score (same as Physical functioning scale score) andtal Scale Score (mean of the sum of all the items in all theles). The Total Scale Score is suitable as a summary scorethe primary analysis of HRQOL outcome in clinical trialsd group comparisons.17

QThe CFQ is a disease-specific HRQOL measure con-

ting of four versions. A self-report CFQ-Teen/adult (14rs of age), CFQ-Child interview version (6-11 years of), CFQ-Child self-report (12-13 years of age) and a CFQ-rent (proxy) version used in conjunction with the CFQ-ild version (6-13 years of age). Each version of the CFQ isferent, but all are composed of broad domains includingysical symptoms, emotional functioning, vitality, socialctioning, and school functioning plus domains specific tosuch as body image, eating disturbances, treatment bur-

n, respiratory symptoms, weight, and digestive symptoms.e number of items for each domain varies according to thesion of the questionnaire, and all are based on a 2-weekall. The CFQ was scored according to the instructions inmanual,10 with higher scores representing a betterQOL. Any missing item was given the mean score of

ms completed for that domain if more than half of theswers for that specific domain were available, otherwisey were left blank and recorded as missing.10 Results wereegorized into each domain according to the CFQ version.

ta CollectionThe PedsQLTM was administered before the CFQ as

r the recommended administration guidelines.9 Patientsd parents completed surveys in the waiting room of atine clinic visit or had surveys mailed to them, thus simul-

land 509

taneous pulmonary function measurements were not alwaysavailable. Questionnaires were self-administered for parentsand for children 8 to 18 years of age for the PedsQLTM and12adPesurpetioPeviecom

aryfroexpCFHeCFAPthrexppa

Co

daUnsisriacoh

weUSCFhameor

Da

PaChDeusescowagrofunFE69FElinratpre

measurements available for participants from January 1, 2000to December 31, 2002 and correlated with CFQ andPedsQLTM HRQOL scores using Pearsons correlation co-effianFEfroPaCFmerat

papopo(53difdeforCFinde.54CF

Pe

thethi8 tothScdolowgro

CF

(70froCFcombytheteeBoWnoHRDime

510years of age to adult for the CFQ. Questionnaires wereministered by interview for children 5 to 7 years of age fordsQLTM and 6 to 11 years of age for CFQ. Interviewveys conducted in the waiting room were completed inde-ndently from the parents. For the mailed surveys, instruc-ns were included on how to complete the surveys. ThedsQLTM has been validated for the parent to act as inter-wer.17 The PedsQLTM takes approximately 5 minutes toplete, and the CFQ takes approximately 15 minutes.Clinical data were collected retrospectively, from Janu-

1, 2000 to December 31, 2002. PFT data were collectedm database records for children 8 years of age. Forcediratory volume in one second (FEV1) was measured in theC using a Vitalograph Compact II (Fisher & Paykelalthcare Pty.Ltd., Rongwood, Victoria, Australia) and atOS using a calibrated portable spirometer (Microlab 3300;S Diagnostics, Gladsville, NSW, Australia). The best ofee maximal forced expiratory maneuvers was recorded andressed as a percentage of reference values based on thetients height, age, and sex.18

mparison of HRQOL Scores With Published DataComparisons of the PedsQLTM scores were made with

ta collected from population samples in Australia13 and theited States.12 The Australian cross-sectional cohort con-ted of 9- to 12-year-old children from the state of Victo-,13 whereas the US data consisted of a cross-sectionalort of 2- to 16-year-olds from California.Comparisons of CFQ scores and pulmonary function

re made with published data from the United States. TheNational Validation Study,10 which was conducted at 18centers, included participants who completed surveys and

d spirometry data collected while attending clinic appoint-nts that were not associated with respiratory exacerbationshospital admissions.

ta AnalysisStatistical analysis was performed using the Statistical

ckage for the Social Sciences (version 11.5; SPSS Inc.,icago, Ill.), with the level of significance set at P .05.scriptive statistics (means and standard deviations) wered to characterize the demographic variables and scaleres for CFQ and PedsQLTM. One-way analysis of variances used to assess differences between CFC and CFOSups for PedsQLTM and CFQ HRQOL scores. Pulmonaryction was categorized into normal (90% of predictedV1), mild (70%-89% of predicted FEV1), moderate (40%-% of predicted FEV1), and severe (40% of predictedV1) for comparison between CFC and CFOS. Simpleear regression was used to calculate FEV1 percent predictede of change against time for each child (FEV1 percentdicted slope), using all historical FEV1 percent predicted

Thomas et alcient. PedsQLTM scores were compared with Australian13

d US12 published data and CFQ scores, and maximumV1 percent predicted were compared with published datam the US National Validation Study10 using the t test.ired correlations were used to determine convergence ofQ-Parent and Child, paired t test to detect differences inan scores between CFQ-Parent and Child and varianceios for differences in SD.

RESULTSThe overall survey completion rate was 75% (162 of 217

rticipants). There was a higher completion rate in the CFCpulation (88.4%, 91 of 103) compared with the CFOSpulation (62.28%, 71 of 114), P .001. More females.7%) than males responded, P .01, but there was noference between genotype of respondents and nonrespon-nts. Of the 46 teens, participation was significantly higherthe CFC (24 of 27, 89.0%) compared with those from theOS (10 of 19, 53.0%), P .006. There was no differencemean (SD) maximum FEV1 percent predicted for respon-nts, 89.6 (16.8) % and nonrespondents 87.4 (15.1) %, P . The demographics of the participants for the CFC andOS are shown (Table I; available at www.jpeds.com).

dsQLTM

One-hundred sixty-two children/adolescents completedPedsQLTM. Six parent proxy-reports were missing for

s group of children/adolescents. Scores were lower for theo 12 age group self-report and proxy, compared with aller age groups, on the Psychosocial, Social, Emotional, andhool domains and the 2 to 4 age group on the Physicalmain, P .05 (Table II). Females reported a significantlyer HRQOL score on the Emotional scale for all ageups, P .001.

QThere were 34 participants for the CFQ-Teen survey

.6% from CFC), 80 for the CFQ-Parent survey (56.3%m CFC), and 83 for the CFQ-Child survey (55.4% fromC). For the CFQ-Parent survey, 72 (90%) surveys werepleted by the mother, 7 (9%) by the father, and one (1%)

the grandmother. Females had a lower HRQOL score forEmotional scale for the CFQ-Child, P .02. Female

ns had a trend toward lower scores for Physical, Vitality,dy Image, and Emotional scales and a higher score foreight compared with male teens, but these differences weret significant. Teens reported a significantly higherQOL than the younger children for Social, Eating, andgestion scales. Parents reported the lowest score for Treat-nt Burden, P .001 (Figure).

The Journal of Pediatrics April 2006

Co

waapscothecal(THRsigchBomeCFcansca

Co

chAuStastuPeViblerep

moin

Table II. PedsQL scales scores expressed as mean (SD) scores for each age group and CFC/CFOS group

AgTotal scale Physical Psychosocial

ealthEmotional Social School

2-4C 1 (10C 9 (10

5-7C 8 (17C 1 (18

5-7C 2 (10C 0 (11

8-1C 4 (16C 3 (17

8-1C 6 (17C 9 (11

13-C 8 (13C 2 (13

13-C 3 (18C 4 (16

ToC 2 (16C 6 (15

PP,

QuInmparison of CFC and CFOS HRQOL ScoresFor the PedsQLTM (Table II) the Total Scale Score

s higher in CFOS group compared with the CFC, and thisproached significance, P .05. The individual domainres were higher in CFOS compared with CFC except forEmotional scale, although differences were not statisti-

ly significant. This trend was also seen in the CFQ surveysable III). Teens from the CFOS group had a higherQOL score for all domains, but this difference was onlynificant for Social and Vitality scales, (P .05). CFOSildren had better HRQOL scores for Physical, Social, anddy Image but worse HRQOL scores for Emotional, Treat-nt Burden, Respiratory, and Digestion compared withC children, although these differences were not signifi-t. There was no significant difference between any of thele scores for the CFQ-Parent (proxy).

mparison of HRQOL Scores With Published DataPedsQLTM scores from a large population sample of

ildren and their parent proxies appear to be similar in anstralian cohort from Victoria13 and from the Unitedtes.12 Patients with CF and their parent proxies in thisdy reported significantly reduced total HRQOLdsQLTM scores compared with scores reported from thectorian cohort and US population and parent proxies (Ta-IV). Children 8 to 12 years of age and their parent proxiesorted the lowest HRQOL scores (Table IV).

e (y) N score health h

PPFC 14 82.5 (13.6) 84.4 (21.2) 81.FOS 21 85.3 (8.1) 90.0 (9.8) 81.SRFC 20 74.6 (14.7) 82.2 (15.2) 70.FOS 8 79.6 (15.8) 81.7 (14.5) 77.PPFC 20 78.2 (11.1) 86.1 (12.4) 76.FOS 7 78.6 (13.3) 72.3 (26.2) 81.2 SRFC 26 69.8 (16.6) 75.4 (18.8) 67.FOS 23 69.9 (19.1) 74.9 (23.6) 65.2 PPFC 25 68.0 (14.8) 75.2 (16.6) 64.FOS 23 69.8 (13.4) 74.3 (19.3) 66.18 SRFC 31 76.4 (13.8) 77.2 (18.3) 76.FOS 17 83.7 (11.5) 89.9 (9.6) 80.18 PPFC 27 71.0 (19.4) 69.2 (23.6) 72.FOS 17 77.3 (13.7) 85.8 (10.8) 72.

talFC 163 73.7 (15.7) 77.6 (18.9) 72.FOS 116 77.0 (15.1) 81.6 (17.9) 73.

parent proxy; SR, self-report.

ality-Of-Life In Children And Adolescents With Cystic Fibrosis MaBoth Regional Outreach And Cystic Fibrosis Center Settings In QuIn both this study and the National Validation Study,10

stly mothers completed the surveys (Table IV). However,this study the female response rate was even higher (90% vs%, P .02). Lung function was similar in both groupsable V). CFQ parent proxy HRQOL scores were lower forscales in this study compared with the US National Val-tion Study except for Respiratory, and this was statisticallynificant for Physical, Emotion, Eating, Treatment Burden,hool (P .001), and Health (P .01). CFQ scores wereilar for teens and children in the US National Validationdy and Queensland, although compared with the UStional Validation Study, our teens scored significantlyer for Body Image and Eating (P .05), and 6- to 13-r-olds scored lower for Eating (P .03) but higher fordy Image (P .02).

mparison of Parent Proxy and Child HRQOLores

There was significant correlation between PedsQLTM

rent and child scores and for CFQ-Parent and Child scoresall scales (Table VI). For the CFQ the strongest agree-nt was for Physical, Body Image, Eating, and Respiratory.001) with a weaker correlation seen for Digestion andotion. Paired t tests showed that parents scored highern the child for Emotion, Respiratory, and Digestion, buts was not statistically significant. Parents scored signifi-tly lower for Physical, Body Image, Eating (P .01), and

functioning functioning functioning

.2) 74.3 (14.0) 88.2 (13.4) 80.8 (16.2)

.5) 77.6 (15.3) 93.3 (8.1) 75.1 (17.9)

.1) 72.0 (22.6) 75.0 (19.3) 65.5 (21.1)

.5) 76.3 (26.7) 77.5 (27.6) 75.0 (16.0)

.6) 76.3 (14.8) 85.0 (13.7) 68.3 (19.9)

.6) 68.6 (26.6) 91.4 (5.5) 82.9 (10.7)

.9) 69.4 (21.1) 73.7 (21.8) 58.5 (19.2)

.1) 64.1 (23.7) 74.7 (18.8) 57.6 (21.2)

.1) 59.6 (18.5) 73.1 (18.6) 60.2 (21.5)

.8) 63.3 (14.8) 73.9 (18.4) 63.2 (15.8)

.3) 72.5 (18.8) 84.5 (15.7) 69.2 (19.2)

.6) 74.1 (20.6) 87.1 (14.5) 77.2 (18.7)

.6) 65.7 (24.2) 79.1 (21.4) 68.2 (22.1)

.5) 63.2 (19.4) 82.4 (27.3) 73.2 (15.8)

.0) 69.5 (20.2) 79.3 (18.8) 66.0 (20.7)

.4) 68.8 (20.5) 82.0 (19.6) 69.3 (19.0)

land 51178(TallidasigScsimStuNalowyeaBo

CoSc

paforme(PEmthathican

nagedeens

TrVaan

agewirepcom

CoFu

corslocatmoHRIm8 t(6.CFpa(nanPu

for t

512eatment Burden (P .001) in both CFC and CFOS.riability of child and parent scores was the same for CFCd CFOS.

For the PedsQLTM (Table II) children 5 to 7 years ofhad a lower HRQOL score on the Social scale compared

th their parent proxy, P .03. The 13 to 18 age grouported a higher HRQOL score on the Emotional scalepared with the parent proxy for this age group, P .04.

rrelation of HRQOL Scores With Pulmonarynction (clinical validity)

Thirty-three teens had Pulmonary function available forrelation with CFQ scores. They had a mean (SD) FEV1pe for all variables of 1.43% (4.90) per year, which indi-es an overall decline in lung function. This declining Pul-nary function correlated significantly with a worseQOL score for CFQ scales Physical (r 0.50), Bodyage (r 0.44), and Health (r 0.36), P .05. Childreno 13 years of age had a mean (SD) FEV1 slope of 0.15%8) per year. There was no correlation of the CFQ-Parent orQ-Child scores with Pulmonary function for matchedrticipants (n 46). For the PedsQLTM 8 to 12 self-report 43), parent proxy (n 41), 13 to 18 self-report (n 46),d parent proxy (n 43) there was no correlation withlmonary function for any of the domains.

Figure. Comparison of CFQ

Thomas et alDISCUSSIONWe report the quality-of-life of children with CF

ated in both tertiary and outreach settings and provide anportant insight on quality-of-life for children and adoles-ts with CF in Queensland. Interestingly, 8- to 12-year-s tended to report a poorer HRQOL on the Psychosocial,cial, Emotional, and School domains compared withunger children and teens on the PedsQLTM. Both the childf-report total scores and parent proxy-report data for thisgroup were similar to values reported from children with

wly diagnosed cancer receiving chemotherapy.17 Children 63 years of age also had a lower HRQOL score on thecial scale for the CFQ. This may reflect children in lateildhood having reduced understanding surrounding theiress and their treatment requirements despite only mini-lly progressed disease, and resenting the imposition ofending health clinics and the management needed. Alter-tively, adolescents may have scored contrastingly higherause of the need for normalization and peer group accep-ce.19 These concerning findings of poorer HRQOL in lateildhood need to be further investigated and may perhaps bedressed through the development of appropriate counselingd educational programs designed to minimize the potentialrmful impact of chronic disease and its management onality-of-life in children.

eens, child, and parent.

The Journal of Pediatrics April 2006treimcenoldSoyoselageneto1Sochillnmaattnabectanchadanhaqu

Outreach versus Tertiary Quality-of-lifeDespite reduced access to multidisciplinary health ser-

vices compared with their city counterparts attending theCFarePeicascaimicasmsoc

anHRtiaphthefougrohenescoenspoproratthi

pochcomdeabiCF

no

provincial children with or without disease to have betterHRQOL. Further study is necessary to better understandthe characteristics of the population by comparing the

Co

lownotiosamencommawhpriQupodeGuthe

Ta pre

S

Ch

C46)

Phy 21.9)Ro AVit AEm 14.8)Soc 15.9)Bo 24.6)Eat 26.2)TB 25.2)He AW ARe 20.5)Dig 26.9)Sch A

TB,

Qu nagedIn eensC, CF patients attending CFOS and living in provincialas reported a statistically slightly better HRQOL. For thedsQLTM questionnaire these differences may not be clin-lly significant given that the difference between the totalle score was less than 4.5, which is the minimal clinicallyportant difference previously reported.12 The minimal clin-lly important difference is a value that is defined as theallest difference in scores required to be clinically andially meaningful to the patient.12

SES was similar between the respondents of the CFCd CFOS, and this may help to account for why theQOL was similar between the two groups with poten-

lly similar values placed on activities, relationships,ysical well-being, health, and independence that may ben reflected in perceived quality-of-life. Spurrier et alnd that children from a lower socioeconomic back-und have significantly more negative experiences withalth,6 and lower SES has been associated with remote-ss.5 The comparison between CFC and CFOS HRQOLres and SES may however have been significantly influ-ced by potential bias in response rates as 80% of nonre-nders were from the outreach group. This was mostnounced for the teens with a significant high refusale in the CFOS and thus a poorer HRQOL or SES ins group cannot be excluded.

Another potential source of bias included parental re-rting bias as more parents from CFOS interviewed theirild and thus could have influenced their childs responsespared with CFC families. We found no systematic evi-

nce of parental influence over the scores with equal vari-lity between parent and child scores for both CFC andOS.

The statistical discrepancy in HRQOL scores mayt be due to CF but rather to the trend for all outreach/

ble III. CFQ-Teen, Child and Parent: scale scores ex

cales

Teen (n 34)

CFC(n 24)

CFOS(n 10)

CF(n

sical 72.6 (23.7) 90.4 (13.1) 76.0 (le 76.2 (21.4) 86.6 (21.9) N/ality 56.0 (25.9) 74.2 (15.9) N/otional 77.2 (18.4) 82.7 (17.0) 76.1 (ial 76.4 (19.1) 94.0 (8.0) 70.2 (dy 72.2 (23.2) 76.7 (23.1) 78.3 (ing 80.6 (23.5) 94.4 (12.0) 76.1 (

56.0 (21.1) 65.6 (26.4) 68.4 (alth 57.4 (21.4) 72.2 (23.6) N/eight 59.7 (34.0) 66.7 (31.4) N/spiratory 68.3 (18.3) 72.8 (12.7) 70.8 (estion 84.3 (16.4) 92.2 (10.5) 76.1 (ool N/A N/A N/

Treatment Burden.

ality-Of-Life In Children And Adolescents With Cystic Fibrosis MaBoth Regional Outreach And Cystic Fibrosis Center Settings In QuQOL of healthy children and adolescents living inional/rural areas with those living in the city with thedsQLTM tool. Our findings are in contrast to a studymparing HRQOL of children with diabetes treated inan and outreach settings.7 Cameron et al7 showed thational diabetic youth have a markedly lower HRQOLmpared with urban diabetic and regional nondiabeticuth using a generic quality-of-life measure, the Childalth Questionnaire. The children had similar diabeticntrol between the two groups. The authors proposed thater-related socialization and a disease-related sense oflation might increase the burden of diabetes experiencedchildren with diabetes from regional areas.7,8 However,there is considerable variability in the construction ofality-of-life measures 20 comparisons of results usingferent HRQOL tools should be made with caution.

mparison of HRQOL Scores With Published DataAs expected, the PedsQLTM scores were significantly

er in our CF patients compared with Australian and USrms supporting their use in discriminating among popula-n groups. Indeed, the comparison between a populationple and our cohort may have underestimated the differ-

ce compared with using a healthy children sample forparison. For the CFQ there was strong correlation for the

jority of the self-report quality-of-life domains assessed,ich reinforces the existing research evidence base,10,21 Sur-singly, the CFQ-Parent (proxy) HRQOL scores foreensland families were generally far lower than those re-rted for the US National Validation Study for all domainsspite having similar pulmonary function in both groups.yatt et al have shown that the parents own feelings towardchilds illness and treatment resulted in lower quality-of-

ssed as mean (SD) scores for CFC and CFOS

ild (n 83) Parent (n 80)

CFOS(n 37)

CFC(n 45)

CFOS(n 35)

77.2 (23.1) 74.2 (20.5) 71.7 (22.1)N/A N/A N/AN/A 64.3 (15.4) 65.0 (16.9)

74.8 (14.9) 76.1 (16.8) 75.0 (21.4)71.9 (19.4) N/A N/A81.1 (25.4) 72.2 (28.7) 69.2 (26.4)76.6 (26.6) 63.0 (33.7) 70.5 (24.3)63.7 (23.8) 45.2 (19.8) 51.4 (18.9)

N/A 69.4 (22.9) 68.3 (21.2)N/A 56.3 (37.5) 57.1 (33.9)

66.9 (23.4) 74.3 (21.0) 74.8 (23.2)72.1 (29.9) 76.5 (16.3) 77.1 (20.0)

N/A 64.5 (27.7) 65.1 (22.6)

land 513HRregPecourbregcoyoHecopeisobyasqudif

lifeingsurVaratchHRwibyspe

Table IV. Comparison of PedsQL HRQOL scores for CF population with published data from the UnitedStates and Australia

pare 12

To

Phy

Psy

Em

Soc

Sch

Re

Par

Sel

To

Phy

Psy

Em

Soc

Sch

514appraisals22 and may adversely influence the proxy scor-. In our study more mothers filled out the CFQ-Parentveys, 90%, as opposed to only 78% in the US Nationallidation Study, and women in general experience increasedes of depression, a vulnerability heightened by caring for aild with a chronic illness and that may affect the parentsQOL report.23 This has been highlighted in our study

th very low HRQOL scores for treatment burden reportedparents, which in addition supports the use of a disease-cific tool and parent proxy-reports to determine the impact

Totaled overall ages com

Respondent N

tal scale score Parent 154Child 125

sical health summary Parent 154Child 125

chosocial health summary Parent 154Child 125

otional Parent 154Child 125

ial Parent 154Child 125

ool Parent 141Child 122

Total scores by ages and respondent c

spondentAge group

(y)

C

N

ent 24 3557 27812 48

1318 44f 57 28

812 491318 48

812-year-olds compared wit

Respondent N

tal scale score Parent 48Child 49

sical health summary Parent 48Child 49

chosocial health summary Parent 48Child 49

otional Parent 48Child 49

ial Parent 48Child 49

ool Parent 48Child 49

Thomas et aldisease has on the perception of the childs HRQOL byparents.The impact of neonatal screening for CF in Australia

o needs to be considered, as the National Validation Studys conducted at 18 CFCs in the United States, where fewters have neonatal screening. Neonatal screening has beenplace since 1985 in Queensland, and thus almost all par-ipants for this study are likely to have been diagnosedough neonatal screening. Neonatal screening has beenven to have benefits for patients with CF, with earlier

d with Varni et al

CF data US data

P valueMean SE Mean

75.3 1.22 81.34 .00174.7 1.42 82.87 .00179.3 1.52 83.26 .01179.2 1.64 86.86 .00173.3 1.23 80.22 .00172.3 1.47 80.73 .00168.0 1.55 80.28 .00170.7 1.92 78.21 .00181.7 1.53 82.15 .7878.8 1.73 84.04 .00368.8 1.65 76.91 .00165.6 1.86 79.92 .001

ared with Varni et al (Table 1)12

ort US data

P valueSE Mean

1.78 87.42 .0782.31 78.02 .912.03 78.86 .0012.64 79.45 .0282.82 81.86 .0492.52 83.31 .0011.94 83.65 .019

illiams et al (Table 2)13

cohort Australian data

P valueean SE Mean

68.8 2.03 82.1 .00169.9 2.52 79.9 .00174.8 2.57 86.2 .00175.2 3.00 84.9 .00265.7 2.12 77.1 .00166.4 2.41 77.3 .00161.4 2.42 75.8 .00166.9 3.18 72.9 .06873.3 2.64 84.2 .00174.2 2.89 81.8 .01161.6 2.72 79.6 .00158.1 2.86 77.2 .001

The Journal of Pediatrics April 2006thethe

alswaceninticthrpro

omp

F coh

Mean

84.278.368.873.476.069.979.0

h W

CF

M

diatiadeinflWdiathedissigcepperatposol

theof-

wipausecoppadisNocopterchStawo

forchCl

Table V. CFQ-Parent (proxy for age 613 y): Comparison with US National Validation Study

CFQ-Parent, National

ScaPVEBETHWRDS

PFTM

PFT

Taan

PedPESSPT

CFPEBETRD

QuIngnosis resulting in better nutritional status24 and, poten-lly, avoidance for families of the stress associated withlayed diagnosis,25 although no research exists evaluating itsuence on parental attitudes toward the illness or the child.hen children are diagnosed because of clinical symptoms,gnosis and institution of appropriate therapy may reinforcenotion that healthcare can impact and ameliorate the

ease; however, early diagnosis often before the onset ofnificant symptoms might prevent parents from ever con-tualizing having a healthy child and may lead to a moressimistic view of having a child whose health will deterio-e with time despite appropriate healthcare provision. Thessible impact of an early sense of loss and grief may con-idate into a longer-term negative appraisal of the child and

CFQ-Parent, this study(n 80) Mean (SD)

leshysical 73.1 (21.1)itality 64.6 (16.0)motional 75.7 (18.8)ody image 70.9 (27.6)ating 66.3 (30.0)reatment burden 47.9 (19.5)ealth 68.9 (22.0)eight 56.7 (35.7)espiratory 74.5 (21.9)igestion 76.8 (17.9)chool function 64.7 (25.4)age 613 yax FEV1 % pred. 88.9 (16.1)

, Pulmonary Function Test.

ble VI. Correlation of paired value for Parentd Child in each scale

Scale Correlation P value

sQLhysical 0.61 .001motional 0.52 .001ocial 0.53 .001chool 0.57 .001sychosocial summary 0.65 .001otal scale score 0.67 .001Qhysical 0.70 .001motion 0.24 .036ody image 0.43 .001ating 0.60 .001reatment burden 0.28 .013espiratory 0.63 .001igestion 0.29 .01

ality-Of-Life In Children And Adolescents With Cystic Fibrosis MaBoth Regional Outreach And Cystic Fibrosis Center Settings In Qufamily unit, and manifest in a perceived poorer quality-life of the CF child.

It is also possible that cultural differences, particularlyth regards to coping strategies between US and Australianrents, may explain the discrepancy in scores. Staab et ald a HRQOL questionnaire to assess how parents wereing with the disease of their child.26 They found that

rents HRQOL was not affected by the severity of theease of their child with CF but by the coping styles used.cultural differences between US and Australian parents ining style have been investigated or reported to date. In-estingly, there were few differences found between theildrens reported HRQOL in Australia and the Unitedtes, and thus if cultural differences are responsible theyuld have to affect parents more than their children.

Although patient self-report is considered the standardmeasuring HRQOL, it is the parents perception of their

ilds HRQOL that may influence healthcare utilization.17

early this issue requires further careful research to investi-e whether these findings are reproduced in other popula-ns to evaluate possible mechanisms responsible.

mparison of Parent Proxy and Child HRQOLores

Strong correlations were seen between parent proxy andild self-report highlighting the importance of both childd parent proxies, although there was weaker agreementween the more subjective domains such as emotion withs being a common occurrence with proxy reporting.27

ality-of-life and Pulmonary FunctionThe clinical validity of the CFQ tool was demonstrated

the correlation of worse HRQOL score with deteriorationpulmonary function for adolescents, which is supported by

Validation Study(n 183) Mean (SD) P value

84.0 (19.0) .00166.9 (15.7) .2783.9 (13.0) .00175.8 (25.6) .1684.1 (20.9) .00173.2 (18.8) .00176.0 (21.1) .01358.8 (38.2) .6772.3 (21.0) .2977.5 (19.0) .7879.0 (22.4) .001

86.8 (20.5) .78

land 515gattio

CoSc

chanbetthi

Qu

byin

nagedeens

published data.10 No such association was found for youngerchildren. This has also been observed in other studies 21 andis attributed to the less severe lung disease in this younger agegroup and the poor sensitivity of pulmonary function mea-sures in children with mild disease.28 There was no correla-tion between the generic PedsQLTM questionnaires and pul-monary function for any of the age groups, confirming thelack of sensitivity of generic HRQOL tools and healthcareoutcomes.

chHRpaLoasstimbetcliriofurHR

Weanancolrev

1.cysme2.etfibr3.fibrand4.sis.5.rem6.tialnat7.Regout8.Au2009.QuHeped

10. Quittner AL, Buu AB, Watrous M, Davis M. Cystic Fibrosis Question-naire (CFQ): A Health-Related Quality of Life Measure: Users Manual. Wash-ington, DC: Cystic Fibrosis Foundation; 2000.11. Quittner AL, Buu AB, Messer M, Modi A, Watrous M. Developmentand validation of The Cystic Fibrosis Questionnaire (CFQ) in the UnitedStates: a health related quality-of-life measure for cystic fibrosis. Chest2005;128:234754.12. Varni JW, Burwinkle T, Seid M, Skarr D. The PedsQLTM* 4.0 as apediatric population health measure: feasibility, reliability, and validity. Am-bulatory Pediatrics 2003;3:329-41.13. Williams J, Wake M, Hesketh K, Maher E, Waters E. Health-relatedqua14.Rem20015.valiCo20016.Qu20017.inInvMo18.frommo19.Lip20.wit21.hea20022.triaPhaPub23.andAv24.Nu19925.D,par26.et aimp27.Ped28.Ped29.conCli11

516This study raises many questions regarding HRQOL inildren and adolescents with CF and the perception of theirQOL by their parents. In general HRQOL was similar in

tients managed in regional centers compared with CFCs.ngitudinal studies of HRQOL would be of interest toist in evaluating the magnitude of change in HRQOL overe, using both generic and disease-specific tools to enableter assessment of change in HRQOL with changes innical status29 and during key developmental transition pe-ds such as late childhood to adolescence. In addition,ther studies are required to confirm and investigate the lowQOL reported by parents for their children in this study.

acknowledge Lyza Norton, Sharon Smith, and the medicald allied health staff at the Royal Childrens Hospital Brisbaned at the Regional Centers who helped with administering andlecting the questionnaires. We thank Dr Stephen Geoghegan foriewing the manuscript.

REFERENCESDakert-Roelse J, te Meerman G. Long term prognosis of patients with

tic fibrosis in relation to early detection by neonatal screening and treat-nt in a cystic fibrosis centre. Thorax 1995;50:712-8.

Mahadeva R, Webb K, Westerbeek R, Carrol N, Dodd M, Bilton D,al. Clinical outcome in relation to care in centres specialising in cysticosis: cross sectional study. BMJ 1998;316:1771-5.Walters S, Britton J, Hodson ME. Hospital care for adults with cystic

osis: an overview and comparison between special cystic fibrosis clinicsgeneral clinics using a patient questionnaire. Thorax 1994;49:300-6.Cystic Fibrosis Foundation. Clinical Practice Guidelines for Cystic Fibro-

Bethesda: Cystic Fibrosis Foundation; 1997.The Australian Institute of Health and Welfare. Health in rural and

ote Australia. AIHW Cat. No. PHE 6. Canberra: AIHW; 1998.Spurrier N, Sawyer M, Clark J, Baghurst P. Socio-economic differen-

s in the health-related quality of life of Australian children: results of aional study. Aust NZ J Public Health 2003;27:27-33.

Cameron F, Clarke C, Hesketh K, White E, Boyce D, Dalton V, et al.ional and urban Victorian diabetic youth: clinical and quality-of-lifecomes. J Paediatr Child Health 2002;38:593-6.Waters E, Salmon L, Wright M. The Child Health Questionnaire in

stralia: reliability, validity and population means. Aust NZ J Public Health0;24:207-10.Varni JW. The PedsQLTM 4.0 Measurement Model for the Pediatric

ality of Life InventoryTM Version 4.0. San Diego: The Center for Childalth Outcomes; 2000. Available online at: http:www.pedsql.org/about_sql.html.

Thomas et allity of life of overweight and obese children. JAMA 2005;293:70-6.Commonwealth Department of Health and Aged Care. Measuringoteness: Accessibility/Remoteness Index of Australia (ARIA). October1. rev. ed. Canberra: The Department of Health and Aged Care; 2001.Varni JW, Seid M, Kurtin PS. The PedsQLTM 4.0: Reliability and

dity of the Pediatric Quality of Life InventoryTM Version 4.0 Genericre Scales in healthy and patient populations. Medical Care1;39:800-12.Varni JW. The PedsQLTM 4.0 scoring algorithm. Scoring the Pediatric

ality of Life Inventory. San Diego: Center for Child Health Outcomes;0. Available online at: http://www.pedsql.org/score.html.Varni JW, Burwinkle T, Katz E, Meeske K, Dickinson P. The PedsQL

pediatric cancer: reliability and validity of the Pediatric Quality of Lifeentory Generic Core Scales, Multidimensional Fatigue Scale, and Cancerdule. Cancer 2002;94:2090-106.Hibbert M, Lannigan A, Landau L, Phelan P. Lung function valuesa longitudinal study of healthy children and adolescents. Pediatr Pul-

nol 1989;7:101-9.Kaplan H, Sadock B. Synopsis of Psychiatry. 8th ed. Philadelphia:

pincott Williams & Wilkins; 1998.Eiser C, Morse R. A review of measures of quality of life for children

h chronic illness. Arch Dis Child 2001;84:205-11.Modi A, Quittner AL. Validation of a disease-specific measure of

lth-related quality of life for children with cystic fibrosis. J Pediatr Psychol3;28:535-46.Guyatt G, Jaeschle R, Feeny D, Patrick D. Measurements in clinical

ls: choosing the right approach. In: Spiker B, ed. Quality of Life andrmacoeconomics in Clinical Trials. 2nd ed. Philadelphia: Lippincott-Ravenlishers; 1996, pp. 418.Australian Bureau of Statistics. 1997 National Survey of Mental HealthWell-being of Adults. Canberra: Australian Bureau of Statistics; 1997.

ailable online at: http://search.abs.gov.au.Farrell P, Kosorok M, Laxova A, Shen G, Koscik R, Bruns W, et al.

tritional benefits of neonatal screening for cystic fibrosis. N Engl J Med7;337:963-9.Merelle M, Huisman J, Alderden-van der Vecht A, Taat F, Bezemer

Griftioen R, et al. Early versus late diagnosis: psychological impact onents of children with cystic fibrosis. Pediatrics 2003;111:346-50.

Staab D, Wenninger K, Gebert N, Rupprath K, Bisson S, Trettin M,l. Quality of life in patients with cystic fibrosis and their parents: what isortant besides disease severity? Thorax 1998;53:727-31.Waters E. Assessing quality of life. In: Moyer V, ed. Evidence Based

iatrics and Child Health. London: BMJ Books; 2000, pp. 7990.Tiddens H. Detecting early structural lung damage in cystic fibrosis.

iatr Pulmonol 2002;34:228-331.Schipper H, Clinch J, Olweny C. Quality of life studies: definitions and

ceptual issues. In: Spiker B, ed. Quality of Life and Pharmacoeconomics innical Trials. 2nd ed. Philadelphia: Lippincott-Raven Publishers; 1996, pp.23.

The Journal of Pediatrics April 2006

Table I. Demographics of participants

Variable CFC CFOS Total (%) P value

Patient number 91 (56.2%) 71 (43.8%) 162 (100%)Patient gender

Male 45 (49.5%) 30 (42.3%) 75 (46.3%)Female 46 (50.5%) 41 (57.7%) 87 (53.7%) .36

Patient mean age (SD)Age group 613 y 9.7 (2.4) 10.2 (2.1) 9.9 (2.3) .36Age group 1418 y 15.9 (1.5) 14.9 (1.8) 15.7 (1.6) .12

Patient genotypeF508/F508 48 (52.7%) 32 (45.1%) 80 (49.4%)F508/other 35 (38.5%) 26 (36.6%) 61 (37.7%)Other/other 4 (4.4%) 4 (5.6%) 8 (4.9%)Missing 4 (4.4%) 9 (12.7%) 13 (8.0%) .26

ARIAHighly accessible 80 (87.9%) 28 (39.4%) 108 (66.7%)Accessible 8 (8.8%) 30 (42.3%) 38 (23.5%)Moderately accessible 3 (3.3%) 9 (12.7%) 12 (7.4%)Remote 0 3 (4.2%) 3 (1.9%)Very remote 0 1 (1.4%) 1 (0.6%) .001

SES (613 y) (maternal education)Some high school or less 17 (42.5%) 13 (39.4%) 30 (39.0%)Completed grade 12/diploma/trade 17 (42.5%) 10 (30.3%) 27 (35.0%)Professional/college degree 10 (22.7%) 10 (30.3%) 20 (26.0%)Total SES data available 44 (57.1%) 33 (42.9%) 77 (100%) .67

PulN 35 (49.3%) 91 (56.2%)M

F

L, L

QuInmonary functionumber available for analysis 56 (61.5%)ax FEV1 % pred (% of no. available)Normal (90% pred) 33 (58.9%)Mild (7089% pred) 16 (28.6%)Moderate (4069% pred) 6 (10.7%)Severe (40% pred) 1 (1.8%)

EV1 slope mean (SD) L/y - all ages 0.98 (6.38)

itres.

ality-Of-Life In Children And Adolescents With Cystic Fibrosis MaBoth Regional Outreach And Cystic Fibrosis Center Settings In Qu17 (48.6%) 50 (54.9%)14 (40.0%) 30 (33.0%)4 (11.4%) 10 (11.0%)0 1 (1.1%) .59

0.34 (17.80) 0.47 (12.00) .58

nagedeensland 516.e1

QUALITY-OF-LIFE IN CHILDREN AND ADOLESCENTS WITH CYSTIC FIBROSIS MANAGED IN BOTH REGIONAL OUTREACH AND CYSTIC FIBROSIS CENTER SETTINGS IN QUEENSLANDMETHODSHRQOL QuestionnairesPedsQLTMCFQData CollectionComparison of HRQOL Scores With Published DataData Analysis

RESULTSPedsQLTMCFQComparison of CFC and CFOS HRQOL ScoresComparison of HRQOL Scores With Published DataComparison of Parent Proxy and Child HRQOL ScoresCorrelation of HRQOL Scores With Pulmonary Function (clinical validity)

DISCUSSIONOutreach versus Tertiary Quality-of-lifeComparison of HRQOL Scores With Published DataComparison of Parent Proxy and Child HRQOL ScoresQuality-of-life and Pulmonary Function

AcknowledgmentREFERENCES