Validation of the Caregiver Priorities and Child Health Index of Life with Disabilities (CPCHILD) in a sample of Turkish non-ambulatory children with cerebral palsy.

AIM: The aim of this study was to translate and transculturally adapt the Caregiver Priorities and Child Health Index of Life with Disabilities questionnaire into the Turkish language and test the reliability and validity.
MATERIAL AND METHODS: Eighty-two children with cerebral palsy and their parents were included in the study. The majority of children had spastic cerebral palsy. According to the Gross Motor Function Classification System, 26 children were level III, 30 children were level IV, and 26 children were level V. International accepted guidelines were used in the transcultural adaptation and validation process. Reliability was assessed through statistical analysis of the test results for test-retest and internal consistency. To assess construct validity, Caregiver Priorities and Child Health Index of Life with Disabilities was compared with the Child Health Questionnaire Parent Form. Concurrent validity was assessed by examining how Caregiver Priorities and Child Health Index of Life with Disabilities scores changed according to Gross Motor Function Classification System levels.
RESULTS: The mean total score of Caregiver Priorities and Child Health Index of Life with Disabilities was 58.34±26.39. The intraclass correlation coefficient for the total questionnaire score was 0.75, ranging from 0.43 to 0.89 for six domains. Cronbach's alpha was above 0.80 in all domains of Caregiver Priorities and Child Health Index of Life with Disabilities, except the health domain. The construct validity was good because there was a positive correlation between total Child Health Questionnaire Parent Form and Caregiver Priorities and Child Health Index of Life with Disabilities scores (r=0.58, p<0.01) according to the Pearson correlation analysis. Caregiver Priorities and Child Health Index of Life with Disabilities scores were found to be different between Gross Motor Function Classification System levels (p<0.05).
CONCLUSION: This study showed that the Caregiver Priorities and Child Health Index of Life with Disabilities appears to be easy to administer, seems to have significant validity and reliability, and may be useful in the evaluation of health-related quality of life of children with cerebral palsy.

Introduction

Cerebral palsy (CP) is a disorder of movement, muscle tone, and posture development, which are attributed to non-progressive disturbances occurring in the developing brain (1). It is the most common developmental disability that restricts children’s activity level and participation in daily life (1, 2). Although CP is a non-progressive disorder, the problems due to the symptoms can be progressive and this may increase the severity of the disability. Between 25–35% of children are severely affected and they exhibit various health problems. Children with severe CP also have difficulties in daily life activities (feeding, toilet, dressing), communication, and mobility (3). The children’s independence and quality of life decrease when their performance in activities of daily living are reduced (4).

Quality of life (QoL) is defined as an individual’s emotional, physical, and social well-being. It is the degree to which an individual’s life satisfies that individual’s goals, wants, expectations, and needs (5). In children with CP, the evaluation of functional status and quality of life are quite important for planning treatment and to evaluate its effectiveness. There are several generic and condition-specific QoL instruments available. Despite the availability of numerous pediatric QoL questionnaires, not all are suited to assess QoL in children with CP (6–8). The generic scales, which were used to evaluate the QoL of children with CP are not specific for neuromuscular diseases. These scales mostly focus on functioning rather than psychosocial well-being. On the other hand, the scales that are used for the evaluation of QoL in children are suitable for those with mild and moderate CP, but they are inappropriate for non-ambulatory children and severely affected children with CP because many of the items in these scales do not apply to children with a severe disability (6). These include the Child Health Questionnaire (CHQ) (9), the Pediatric Quality of Life InventoryTM (PedsQLTM) (10) and the KIDSCREEN (11), and do not accurately measure the health-related QoL (HRQoL) of individuals with severe CP.

To address the lack of evaluation scales for severely affected children with CP, the condition- specific Caregiver Priorities and Child Health Index of Life with Disabilities (CPCHILD) questionnaire was developed to evaluate the HRQoL and well-being of these children (7). Unlike other questionnaires, CPCHILD was specifically developed by taking into account the experiences of caregivers of children with severe CP as well as healthcare professionals who were responsible for the treatment of these children. The questionnaire was developed for the Canadian population in English. In addition to this, Brazilian-Portuguese, Dutch, German and Korean translations have been developed and validated (3, 12–14).

The aim of this study was to translate and transculturally adapt the CPCHILD questionnaire into Turkish, and to test the reliability, validity, and internal consistency of the Turkish version of the CPCHILD. We planned this study because of the absence of a suitable scale in Turkish, resulting in a lack of evaluation of the HRQoL of children with severe CP in Turkey.

Material and Methods

Design

This study was conducted at five different special education and rehabilitation institutions enrolling children who were involved with rehabilitation programs. All procedures performed were in accordance with the ethical standards of the institutional research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. This study was approved by the Abant İzzet Baysal University Clinical Research Ethics Committee. (13.04.2011/13). The parents/caregivers of the children included in the study signed a consent form in which the method and aim of the study was explained.

Participants

Sample size recommendations for such a study is between 50–100 participants (13, 15). The two original validation studies of the CPCHILD were conducted with 52 and 67 participants (7, 16). Other translation, cross-cultural adaptation, and validation studies were conducted with 30, 66, and 68 subjects (3, 13, 14). Eighty-two children with CP and their parents-caregivers were included in this study. Parents who were the primary caregivers of children with CP aged 5 to 18 years were included. The gross motor function of the children was classified using the Gross Motor Function Classification System (GMFCS) (17). The children who were at level III-V according to the GMFCS were included in the study. The clinical characteristics of the children and socio-demographic data of the caregivers were recorded.

Procedure

Scales

The CPCHILD, a condition-specific measure of health status and well-being of children with severe CP was developed specifically for this population. The CPCHILD consists of 37 items distributed over six sections representing the following domains: (1) personal care, (2) positioning, transfer and mobility, (3) communication and social interaction, (4) comfort, emotions and behavior, (5) health, and (6) overall QoL. Each item is rated on a 6- or 7-point scale, and for some items, in addition, a 4-point level of assistance modifier. Transformed scores from 0 (worst) to 100 (best) are calculated for a total CPCHILD score, as well as for each of the six sub-scales or domains (7, 16) (Apendix 1, Turkish version of scale).

The Child Health Questionnaire Parent Form (CHQ PF-50) has 13 domains or subscale scores: physical functioning (PF), role limitations-emotional/behavioral (REB), role limitations-physical (RP), bodily pain (BP), behavior (BE), mental health (MH), self-esteem (SE), general health perceptions (GH), parental impact-emotional (PE), parental impact-time (PT), family activities (FA), family cohesion (FC) and change in health (GGH). The CHQ-PF50 also includes three item measures of global functioning. The CHQ-PF50 has been validated in several countries and languages and results consistently indicate that the measure is psychometrically sound (18). CHQ scores range from 0 to 100, with higher scores indicating better HRQoL. In this study, the Turkish version of CHQ-PF50 was used (19). We chose the CHQ-PF50 because it evaluates the QoL and well-being of children with disabilities including CP. It was culturally adapted for Turkish people and the measurement properties were tested.

Translation and cross-cultural adaptation

Permission for the adaptation of the questionnaire was obtained from the authors of the questionnaire before the study. We used the cross-cultural adaptation designs proposed by Guillemin et al. (20) and Ruperto et al. (21) during the translation process. Two forward translations were performed from English to Turkish by translators whose native language was Turkish. One of the translators was blinded to the purpose of the study and the concepts being examined in the questionnaire. The other translator, who was a professional experienced in treating children with disabilities, was given information about the purpose of the study and the concepts being quantified. The two translations provided two preliminary Turkish versions from both clinical and literal perspectives.

A meeting was held with the two translators and two other physiotherapists experienced in pediatric rehabilitation to compare the two versions. A consensus was reached, which was then back translated into English by two native English speakers who did not know the purpose of the study, but had acquired the necessary reading and writing skills in Turkish. Each of the two translations was then compared with the original version. A bilingual team consisting of the 4 translators and 2 physiotherapists involved in the first meeting reviewed the Turkish version of the questionnaire in order to make the cross-cultural equivalence and to achieve semantic, idiomatic, experimental, and conceptual equivalence. The adapted translation was then tested by means of one-to-one interviews with 20 parents of children with CP. None of the statements were changed or altered.

Statistical Analysis

Assessment of reliability

The test-retest reliability of CPCHILD was determined by using the intraclass correlation coefficient (ICC) including 95% confidence intervals (CI). The CPCHILD was administered to the parents/caregivers twice, separated by a 2-week interval. For the ICC, values were accepted as follows: between 0 and 0.20 was poor, 0.21 and 0.40 was fair, 0.41 and 0.60 was moderate, 0.61 and 0.80 was good, and between 0.81 and 1.00 was excellent reliability (22, 23). For the internal consistency of CPCHILD, Cronbach alpha (α) coefficients were calculated. Alpha coefficients between 0.70 and 0.95 for all domains were considered adequate; the value of alpha should be above 0.80 for acceptance as high internal consistency (23).

Assessment of validity

In order to assess construct validity, CPCHILD scores were compared with the Child Health Questionnaire Parent Form (CHQ PF-50). The relation between CPCHILD and CHQ PF-50 was investigated using Pearson correlation analysis. Construct validity coefficients were accepted as follows: r ≥0.81 to 1.0 as excellent, 0.61 to 0.80 as very good, 0.41 to 0.60 as good, 0.21 to 0.40 as fair, and 0 to 0.20 as poor validity (24). Concurrent validity was assessed by examining how CPCHILD scores changed according to GMFCS levels. Differences in the CPCHILD scores according to GMFCS levels were analyzed using ANOVA and the post-hoc Tukey test. We used parametric tests for the CPCHILD because the scores had near-normal distribution.

All statistical analyses were performed using the SPSS 20.0 for Windows software package. A probability value of p<0.05 was considered to indicate a significant effect.

Results

The mean age of children included in the study was 11.04±5.26 years. All of the parents participating in the study were mothers. The clinical characteristics of the children and demographic characteristics of both the children and caregivers are given in Table 1. The average time to complete the questionnaire was 26.53±4.78 (range, 20–40) minutes; individuals with lower educational levels required more time to complete the questionnaire. The mean (±SD) total score of the CPCHILD was 58.34±26.39.

Table 1

Sociodemographic data of the participants

	n	%
Data of children
Age (X±SD, year)	11.04±5.26
Height (X±SD, cm)	124.42±30.82
Weight (X±SD, kg)	33.45±19.69
Girl	39	47.6
Boy	43	52.4
Communication
Yes	48	58.5
No	34	41.5
GMFCS
Level 3	26	31.7
Level 4	30	36.6
Level 5	26	31.7
Spastic	62	75.6
Hypotonic	6	7.3
Dyskinetic	6	7.3
Ataxic	2	2.4
Mixed	3	3.7
Caregiver
Data of mothers	82	100
Age (Mean±SD, year)	40.47±9.02
Educational level
Primary school	29	35.4
Secondary school	15	18.3
High school	24	29.3
University	8	9.8
None	6	7.3
Profession
Housewife	75	91.4
Government employee	3	3.7
Worker	4	4.9

SD: Standard deviation; GMFCS: Gross Motor Function Classification System

Test-Retest Reliability & Internal Consistency

The ICC for the total questionnaire score was 0.75 and ranged from 0.43 to 0.89 for the six domain scores. There was no difference between the test-retest scores (p>0.05). Cronbach’s alpha for the total CPCHILD was 0.97 and was above 0.80 in all domains of the CPCHILD, except the health domain (Table 2). These results support good test-retest reliability and high internal consistency of the questionnaire.

Table 2

Test-retest reliability of the CPCHILD domain scores and total score

Domains CPCHILD (items)	ICC (95% CI)	Cronbach α
Personal care and ADLss (9)	0.63 (0.55–0.63)	0.96
Positioning, transfer and mobility (8)	0.61 (0.45–0.68)	0.94
Communication and social interaction (7)	0.62 (0.55–0.70)	0.95
Comfort, emotions and behavior (9)	0.89 (0.65–0.98)	0.89
Health (3)	0.43 (0.02–0.50)	0.32
Overall quality of life (1)	0.64 (-0.46–0.79)	NA
Total score (36)	0.75 (0.60–0.84)	0.97

CPCHILD: Caregiver Priorities and Child Health Index of Life with disabilities; CI: Confidence intervals; ICC: Intraclass correlation; NA: Not applicable (Cronbach α since only one item)

Validity

According to the Pearson correlation analysis, there was a positive correlation between the total CHQ and CPCHILD scores (r=0.58, p<0.01). The relation between CHQ and CPCHILD domains is demonstrated in Table 3. This results showed that the construct validity of CPCHILD ranged between good and very good, except health domain, which had fair correlation (r=0.38, p=0.01).

Table 3

Pearson correlation coefficients between CPCHILD scores and CHQPF-50 total score

CPCHILD scores	CHQPF-50
	r	p
Personal care (8)	0.42	<0.001[a]
Positioning, transfer and mobility (8)	0.52	<0.001
Communication and social interaction (7)	0.42	<0.001[a]
Comfort, emotions and behavior (9)	0.62	<0.001[a]
Health (3)	0.38	0.01[a]
Overall quality of life (1)	0.45	<0.001[a]
Total score	0.58	<0.001[a]

CPCHILD: Caregiver Priorities and Child Health Index of Life with disabilities; Child Health Questionnaire- Parent Form;

p<0.05, Pearson correlation coefficients

In terms of concurrent validity, both total and domain scores of the CPCHILD, except the health domain (p>0.05), were found to be significantly different between each GMFCS level (p<0.05). Post-hoc test showed significant differences in CPCHILD scores between level III and IV, level IV and V, and level III and V of GMFCS (p<0.05, Table 4).

Table 4

Concurrent validity of the scale by using GMFCS levels (ANOVA analysis)

CPCHILD	GMFCS levels			p
	III (n=26) (Mean±SD)	IV (n=30) (Mean±SD)	V (n=26) (Mean±SD)
Personal care & ADLs	45.7±22.1	33.5±18.05	28.07±20.0	<0.001[a]
Positioning, transfer and mobility	78.40±25.71	58.71±32.56	41.62±22.70	<0.001[a]
Communication and social interaction	63.81±31.03	59.24±30.98	32±18.05	<0.001[a]
Comfort, emotions and behavior	37.48±22.78	33.98±24.09	28.87±23.27	<0.001[a]
Health	92.22±27.90	88±05.28.76	42.77±31.71	0.11
Overall quality of life	60.66±15.29	60±16.97	43±20±13.75	<0.001[a]
Total score	63.04±20.23	55.57±20.23	36.05±7.15	<0.001[a]
Statistics of the total score	p1
	p2
	p3

GMFCS: Gross Motor Function Classification System; CPCHILD: Caregiver Priorities and Child Health Index of Life with disabilities.

p<0.05; p1: Significant differences between levels III and IV. p2: Significant differences between levels IV and V. p3: Significant differences between levels III and V

Discussion

The results of our study showed that the Turkish version of the CPCHILD had good reliability and validity for Turkish-speaking caregivers and could be used for the evaluation of the HRQoL of Turkish children with severe CP for clinical and research purposes.

The ICC score of the original validation study was 0.85, ranging between 0.59–0.87 for the domains (16). In the Dutch validation study, the ICC score was found as 0.73 for the total questionnaire and ranged between 0.55–0.80 for domains (13). The Turkish version of CPCHILD showed good test–retest reliability in the total ICC (0.75) and domain scores (ICC, 0.43–0.89) as similar to previous studies. In terms of internal consistency, Cronbach’s alpha of all domains was above 0.89, except the health domain, which shows high internal consistency. These results were consistent with the German, Korean, and Dutch validation studies (12–14).

According to our study, the construct validity of the CPCHILD ranged between good and very good, except the health domain, which has 3 items. The health domain assesses the child’s health status and frequency of hospital admissions/visits, and these questions are not always the most important determinants of QoL for all children with CP.

For concurrent validity, domain scores and the total score of the CPCHILD were significantly different according to patients’ GMFCS levels. The CPCHILD was able to discriminate between the groups. The only exception for this was the health domain. This result is different from the original validation study (16), but similar to the German and Dutch validation studies (12, 14).

There are some limitations of this study. First, although the CHQ was designed for the evaluation of QoL of children with disabilities, it is not specific for children with CP. It can be inadequate for severely affected children with CP. In previous studies, families reported that some of the questions involved in CHQ PF-50 were not suitable for severely affected children with CP (25, 26).

Second, the CPCHILD was completed in accordance with the information received from the family. The educational and emotional status of parents may affect the results of the questionnaire. In our study, 50% of the caregivers had a low level of education (none, primary-secondary school). This might have an effect on the results of our study. As stated in the original study, it would be more effective if data from HRQoL evaluations were obtained directly from children instead of families. However, many of the children with severe CP have significant cognitive impairments, which makes this impossible. Nevertheless, for those children who are able, a child version of the CPCHILD will be useful and its validity should be assessed in further studies.

In children with non-walking and severely affected CP, children and caregiver needs can be significantly affected in many different ways. It also requires a versatile and multidisciplinary approach. For this reason, determination of the parameters affecting the QoL and problems experienced by severely affected children and caregivers are of great importance in determining rehabilitation goals. We aimed to study the validity and reliability of the Turkish version of the CPCHILD scale that we developed to close this gap in the literature. It is an important reference for future studies and rehabilitation program targets to be implemented. CPCHILD is able to help the effectiveness of rehabilitation programs and follow-up on the development of children, record changes in caregivers, observe how QoL is affected, and set new goals. In our country, there is a need to develop new scales that can assess HRQoL and the caregiving burden in children who are severely handicapped with chronic disabilities, and more work needs to be done to adapt the existing international scales to the Turkish language.

Conclusion

The Turkish version of CPCHILD was found to be a reliable and valid questionnaire for the evaluation of HRQoL of children with moderate and severe CP. This study also showed that CPCHILD could be easily used both for research and clinical settings. There is a need for further studies on the development of new scales that can effectively assess caregiver burden for children who cannot walk and are severely affected, HRQoL, and/or the adaptation of scales developed for use in international studies to the Turkish language.