PedsQL TM 4.0 Generic Core Scales for adolescents in the Yoruba language: Translation and general psychometric properties

Abstract

Background:

Quality of life (QOL) is a universally accepted concept for measuring the impact of different aspects of life on general well-being. Adaptation of existing QOL instruments to local cultures has been identified as a better strategy than development of new ones.

Aims:

To translate and adapt the Paediatric Quality of Life Inventory™ Version 4.0 Generic Core Scales (PedsQL™) to the Yoruba language and culture and to test the psychometric properties of the adapted instrument among adolescents.

Methods:

Psychometric properties including internal consistency reliability, construct and factorial validity of the Yoruba version of PedsQL™ were evaluated using standard procedures.

Results:

The self report and proxy scales of the Yoruba PedsQL™ were developed with good cultural relevance and semantic/conceptual equivalence. Results from 527 adolescents revealed a Cronbach’s coefficient which exceeded 0.7 for internal consistency reliability for all scores. The healthy subjects reported higher PedsQL™ scores than those with mental health and physical problems, which confirmed construct validity. Confirmatory factor analysis revealed a good model fit for the Psychosocial Health score, but not for the other measures.

Conclusions:

The Yoruba PedsQL™ is culturally appropriate and with good internal consistency, reliability and construct validity. More work is needed regarding its factorial validity.

Keywords

Adolescents Nigeria psychometrics quality of life Yoruba

Introduction

According to information available at the official website of the Paediatric Quality of Life Inventory^TM (PedsQL^TM) (www.pedsql.org.), the Paediatric Quality of Life instrument has been translated and adapted to over 70 languages and cultures, and has been used in more than 400 scientific publications. Despite its popularity and wide usage, the PedsQL^TM is yet to be translated to or validated in any language or culture in sub-Sahara Africa. This has created a situation of dearth of paediatric Health Related Quality of Life (HRQOL) research in the region. This is a critical omission because up to one-quarter of the world’s child and adolescent population are currently domiciled in sub-Sahara Africa (UNICEF, 2011). In addition, the facilities to optimally nurture this burgeoning population of children are lacking due to festering poverty and social inequalities in most parts of sub-Sahara Africa (Earth Trends, 2003; Shaw & Tschiwula, 2002). This scenario is expected to create a situation of high prevalence of child mental health problems. Studies have actually established a link between communal poverty and social inequalities and a high prevalence of childhood mental health problems (Costello, Compton, Keeler, & Angold, 2003; Friedman & Chase-Lansdale, 2002; Wachs, 2000). Furthermore, the mere fact that almost 50% of global burden of mental disorders are first seen in childhood (Kessler et al., 2007) makes child and adolescent mental health (CAMH) a priority public health issue in a predominantly youthful population like that of the sub-Sahara Africa region.

Facilities and human resource for child mental health are grossly deficient in many sub-Sahara African countries (Robertson, Omigbodun, & Gaddour, 2010), and so is epidemiological research in child and adolescent psychiatry. A recent meta-analysis of community based epidemiological studies of the prevalence of child mental health problems in sub-Sahara Africa yielded only 11 studies spread over six of the 48 countries in the region (Cortina, Sodha, Fazel, & Ramchandani, 2012). The review also highlighted the dearth of culturally adapted or validated psychometric tools for assessing psychopathology among children in the region. Other than being the global direction in service planning; epidemiological data can be particularly invaluable in resource-constrained regions like sub-Sahara Africa for judicious allocation of scarce resources. Unfortunately, the non-availability of culturally appropriate CAMH psychometric instruments still remain a major issue that a CAMH epidemiology researcher in a sub-Sahara African country will have to contend with. Cultural and linguistic adaptations of commonly used psychometric instruments are increasingly needed (Perneger, Leplège, & Etter, 1999) in the region and adaptation of existing instruments to local cultures has been identified as a better strategy than the development of new ones (Reichenheim & Moraes, 2007).

Among several psychometric instruments used in epidemiological research, self-reports of HRQOL remain an important health outcome measure in paediatric populations (Varni et al., 1999). The HRQOL concept is very useful in monitoring health status and tracking treatment outcomes among children and adolescents (Varni et al., 1999). For this reason, psychometrically sound and culturally appropriate adaptations of quality of life measures are urgently needed in cross-cultural and multinational child health surveys (Chassany, Sagnier, Marquis, Fullerton, & Aaronson, 2002; Schmidt & Bullinger, 2003). The most widely used measure of HRQOL among paediatric populations is the Paediatric Quality of Life Inventory^TM (PedsQL^TM) Version 4.0 Generic Core Scales (Varni, Seid, & Rode, 2001). The popularity of PedsQL^TM is not unconnected with the fact that it is a problem-oriented instrument that incorporates aspects of physical, emotional, social and school functioning of the child or adolescent into the measure of quality of life. In addition, it has been found to have appropriate psychometric properties (Varni, Seid, & Kurtin, 2001).

There is therefore, an urgent need to adapt the PedsQL^TM into dominant languages and cultures in sub-Sahara Africa. This will boost paediatric epidemiology research in the region. It will stimulate discourse on paediatric HRQOL measurement as a veritable tool for assessment of severity of psychopathology and for effective monitoring and evaluation of child-health interventions in the region. Also, comparing quality of life reports for adolescents in this region with global data will be more accurate if done with a well adapted and validated instrument. The overall effect will be a better capacity for the few child mental-health service providers in this region to impact more on children’s mental health.

Yoruba is a major language and culture in sub-Sahara Africa. The traditional abode of the Yorubas is in the south-western part of Nigeria. The use and influence of the Yoruba language and culture, however, extends beyond Nigeria to include some regions in Togo, Ghana, Cuba, Brazil, Haiti, Puerto Rico and Trinidad (Ojo, 2006). This study is a description of the process of translation and cultural adaptation of the PedsQL^TM 4.0 Generic Core Scales to Yoruba. The psychometric properties of the resultant document were also tested among school-going adolescents.

Objectives

This study aimed at translating and culturally adapting the PedsQL^TM 4.0 Generic Core Scales to the Yoruba language and culture. The psychometric properties, in terms of internal consistency reliability, construct validity and convergent validity of the Yoruba version, were also assessed.

Methods

Setting

The study was carried out in Ibadan, Nigeria. Ibadan is the third largest city in Nigeria, with an estimated population of about 2.6 million spread over 11 local government districts. The principal language spoken among community dwellers is Yoruba.

Study design and sample selection

Participants for this study were randomly sampled from two randomly chosen high schools in the Ibadan region. The two schools were randomly chosen from a list of 17 public high schools – as obtained from local authorities – within the locality of the research institute where the study was conducted. School psychologists contacted 587 adolescents (aged 13–18 years). They informed all the adolescents about the study. The only exclusion criterion was demonstrable inability to read and write the Yoruba language. Of all contacted, 527 (89.77 % response rate) agreed to participate and gave consent. The adolescents completed the questionnaire at schools in order to prevent a low response rate. The questionnaires were administered to the adolescents while they were seated in the school multipurpose halls. They had enough space for comfort and privacy.

Instruments

PedsQL^TM

The items on the PedsQL^TM include: (1) Physical Functioning – PF (eight items), (2) Emotional Functioning – EF (five items), (3) Social Functioning – SoF (five items), and (4) School Functioning – ScF (five items) Scales (Varni et al., 1999). Within the scales, all items were presented in a five-point response scale (0 = never a problem; 1 = almost never a problem; 2 = sometimes a problem; 3 = often a problem; 4 = almost always a problem), all were reverse-scored and linearly transformed to a 0–100 scale. Higher scores indicate better quality of life (QOL). Three scale scores were computed. The Physical Health Summary Score was computed as the sum of the items divided by the number of items answered in the PF Scale. To create the Psychosocial Health Summary Score (15 items), the mean was computed as the sum of the items divided by the number of items answered in the EF, SoF and ScF Scales. Finally, the total PedsQL^TM score was computed as the sum of all items divided by the number of items answered. If more than 50% of the items in the scale were missing, a score was not computed. The instrument was provided by MAPI Research Trust, who, being the copyright owner, granted permission for its translation and cross-cultural validation according to their standardized procedures.

Strengths and Difficulties Questionnaire (SDQ)

The Strengths and Difficulties Questionnaire (SDQ) was administered to assess general mental health problems (Goodman, 1997). The SDQ possesses 25 items comprising the following five-item scales: emotional problems, conduct problems, hyperactivity/inattention, peer relationship problems, and pro-social behaviour. Each item was scored on a three-point scale and the sum of all answered items in a scale creates its total score (possible range 0–10), while the sum of all answered items in the first four scales create the total score (possible range 0–40). Higher total scores indicate greater difficulties. Additionally, the SDQ total scores were classified as ‘normal’ (range 0–13), ‘borderline’ (range 14–16), and ‘abnormal’ (range 17–40), indicating the degree of presence of general psychopathology (Goodman, 1997). The Yoruba version was obtained from www.sdqinfo.org .

Procedures and data analysis

Translation and cultural adaptation

The translation and cultural adaptation of the PedsQL^TM self- and proxy-report were performed according to the protocol for translating the PedsQL^TM measures provided on the official website. The translation and cultural adaptation included preparation, forward translation and reconciliation, back translation, harmonization, cognitive debriefing, and finalization. The aim was to develop a version equivalent to the original (item, semantic and operational equivalence), but with significant cultural sensitivity too.

The translation team comprised two researchers with experience in child and adolescent mental health research, who also have a good grasp of psychological constructs and are fluent in English and Yoruba languages. They performed two independent forward translations (English–Yoruba). From these versions, a single form was developed (reconciliation I). This form was back-translated into English by two independent English–Yoruba translators. One back translation was developed (reconciliation II) that was compared with the originals (harmonization). The purpose was to highlight all differences that arose during the translation process and overcome the discrepancies in the concepts between the originals and the translations.

The entire process resulted in the pre-final versions that were pretested in semi-structured interviews on a group of 10 Yoruba-speaking adolescents (aged 13–17 years) together with two Yoruba-speaking parents. Semi-structured interviews were organized separately for adolescents and parents. The aim of this step was to explore comprehensibility and the response process of every item in the questionnaires, as well as clarity and appropriateness (cognitive debriefing). The adolescents/parents were asked to ‘think aloud’ about items and how they formed answers, to try to explain the meaning of each item and to give an opinion on whether all items were clear and appropriate (unambiguous and free of wording that might adversely affect a child). Afterwards, an expert panel meeting was organized to evaluate the content and face validity of the versions, the pre-test results and the equivalence with the originals. Finally, the Yoruba versions were developed.

Statistical procedures for psychometric testing

Descriptive analysis

The mean, standard deviation, skewness and kurtosis as well as the floor and ceiling effects of the variables were calculated for all scales and total score. Missing data was also described.

Internal consistency reliability and inter-scale correlations

The internal consistency of the scales was assessed by computing Cronbach’s coefficient-α and the recommended value of ≥ 0.7 was set as minimum reliability criterion in determining the appropriateness of the Yoruba version (Terwee et al., 2007). In order to determine whether the scales assess separate concepts within a more general one, QOL, correlations between subscales were computed (using Pearson’s correlation coefficient). A scale is considered redundant if Pearson’s coefficient exceeds 0.7 (Nunnally & Bernstein, 1994).

Construct validity

The factor structure as a form of construct validity of the Yoruba version of the PedsQL^TM was tested using confirmative factor analysis (CFA). This was done using the Analysis of Moment Structures Version 7 (AMOS-7). Four models of the factor structure of the Yoruba version of the PedsQL^TM were tested. The first is a model representing all 23 items on the instrument and the corresponding five factors (Limbers, Newman, & Varni, 2008). The second model is a second order four-factor model which represents the four original scales (Petersen, Hägglöf, Stenlund, & Bergström, 2009). The third and fourth models include the PedsQL^TM Physical Health model (one factor represented by eight Physical Functioning items) and PedsQL^TM Psychosocial Health model (a second-order model derived from the EF, SoF and ScF scales) (Stevanović et al., 2011). The following statistics assessed the adequacy of the models, as the degree of fit between estimated and observed variance: Tucker Lewis index (TLI) (> 0.90 acceptable, > 0.95 excellent), the comparative fit index (CFI) (> 0.90 acceptable, > 0.95 excellent) and root mean square error of approximation (RMSEA) (< 0.08 acceptable, < 0.05 excellent) (Meuleners, Lee, Binns, & Lower, 2003; Nevitt & Hancock, 2001).

Construct validity was also determined utilizing the known-groups method (Terwee et al., 2007). The known-groups method compares scale scores across groups known to differ in the health construct being investigated. In the study, groups differing in known mental status, based on the SDQ normal/borderline/abnormal score, and physical health, based on the presence/absence of certain chronic conditions, were computed using ANCOVA corrected for the FAS score considering that the rates of those with high affluence were lower than the other two groups. Based on previous PedsQL^TM findings (Limbers, Newman, & Varni, 2008; Petersen et al., 2009), we anticipated that healthy children would report significantly higher PedsQL^TM 4.0 scores than children with mental or physical health problems.

Convergent validity

Spearman’s rho correlations between the scores on the PedsQL^TM and SDQ scales were taken as a measure of convergent validity of the former. The hypothesis was that the PedsQL^TM Psychosocial Health scales would show negative correlations with the emotional and conduct problem, hyperactivity/inattention, and peer relationship problems scales of the SDQ on one hand, and positive correlations with the pro-social scale on the other (Stevanović, Lakić, & Damnjanovic, 2010). However, it was expected that the PedsQL^TM Physical Health scale would show a comparably weaker correlation with all SDQ scales than the Psychosocial Health.

Ethical consideration

Permission to interview students was obtained from the Ministry of Education of Oyo state, Nigeria. The school principals, acting on behalf of the State Commissioner for Education, gave written approval of the study. Adolescents who were over 16 years of age signed individual consent forms in accordance with relevant laws in Nigeria (Laws of Ogun State of Nigeria, 2006). Those below the age of 16 signed individual assent forms and also submitted signed parental consent forms to the investigators.

Results

The Yoruba version of the PedsQL^TM

The main results of the translation and adaptation process were the Yoruba version of PedsQL^TM for adolescents’ self- and proxy-report versions. The original Yoruba versions as developed and the back-translated English versions can be obtained with permission from the Mapi Institute. The most important translational strategies applied were semantic re-arrangements, minor supplementations to items and substitutions of certain words with culturally appropriate synonyms. No item was omitted in whole or part; only some slight changes were made. The most outstanding of these was the use of the statement: ‘walking more than the distance from one pole to the other’ to replace the original ‘walking more than one block’, as used in the original PedsQL^TM 4.0 generic version. Also, the phrase ‘missing school to go to the doctor or hospital’ was expanded to include ‘…or prayer houses and traditional healers’.

Psychometric properties of the Yoruba version of PedsQL^TM

Sample characteristics

The final sample included 527 adolescents; 238 (45.2%) were males and 289 (54.8%) females. The mean age of the sample was 14.98 (1.26) years. Of all, 75 (14.2%) had some chronic disorder or disability.

Descriptive statistics

The overall amount of missing data was 0.8%. The items’ means (SD) ranged from 75.5 (17.4) to 84.6 (21.9). All scales’ scores were negatively skewed and leptokurtic. All scales except Psychosocial Health scale had ceiling effects above 20%. Floor effects were below 1%.

Internal consistency, reliability and inter-scale correlation

Scale internal consistency reliability determined by Cronbach’s coefficient was appropriate with the value above 0.7 for all scales’ scores (Table 1). Inter-scale correlations ranged 0.51–0.57.

Table 1.

Descriptive statistics and Cronbach’s coefficients (α) of the Yoruba PedsQL^TM scores (N = 527).

PedsQL^TM Score	M (SD)	Floor (%)	Ceiling (%)	Skewness (SE)	Kurtosis (SE)	α
Emotional Functioning	75.51 (17.44)	0.4	21.4	–0.92 (0.11)	0.47 (0.12)	0.71
Social Functioning	84.60 (21.87)	0.8	40.2	–1.91 (0.11)	3.61 (0.12)	0.79
School Functioning	83.67 (17.84)	0.4	34.7	–1.68 (0.11)	3.27 (0.12)	0.73
Physical Health	82.70 (22.03)	0	32	–1.55 (0.11)	1.79 (0.12)	0.86
Psychosocial Health	81.26 (21.51)	0	13.1	–1.59 (0.11)	3.21 (0.12)	0.87
Total	81.76 (19.40)	0	10.2	–1.55 (0.11)	2.86 (0.12)	0.91

Construct validity

Allowing some error correlations between two items in the same scales as indicated by modification indexes, the CFA statistics assessing the adequacy of the models revealed acceptable model fit for the PedsQL^TM Psychosocial Health model (Table 2). For the hypothesized models of four and five factors for the PedsQL^TM self-report, the CFI and TLI values were slightly below acceptable ranges, while for the PedsQL^TM Physical Health model the RMSEA value was slightly above.

Table 2.

Confirmatory factor analyses of the Yoruba PedsQL^TM (N = 527).

Statistic	PedsQL^TM model
	Five-factor^a	Four-factor^b	Physical Health^c	Psychosocial Health^d
Root mean square error of approximation (RMSEA)	0.06	0.06	0.09	0.06
Comparative fit index (CFI)	0.89	0.89	0.94	0.93
Tucker Lewis index (TLI)	0.88	0.88	0.91	0.91

Error correlations items 1–2.

Error correlations items 1–2 and 22–23.

Error correlations items 1–2 and 7–8.

Error correlations items 22–23.

Further assessment of construct validity using the known-groups method revealed that the healthy subjects had reported significantly higher PedsQL^TM 4.0 scores than those with mental health problems as indicated by the SDQ scores (Tables 3 and 4). Except for the SoF Scale, they also reported significantly higher PedsQL^TM 4.0 scores than children with borderline mental health problems as indicated by the SDQ. Furthermore, the healthy subjects reported higher PedsQL^TM 4.0 scores than children with health problems in all PedsQL scales, but in the Physical Health and SoF Scale score the difference did not reach statistical significance.

Table 3.

Distribution of the Yoruba PedsQL^TM scores according to the SDQ scores (N = 527).

PedsQL^TM score	SDQ score range	N	M	SD	Fancova (p value)
Total	Normal	346	83.90	16.69	6.08 (0.000)^a
	Borderline	104	77.99	19.22
	Abnormal	77	76.30	16.55
Physical Health	Normal	346	84.93	20.93	3.66 (0.02)^a
	Borderline	104	79.06	23.22
	Abnormal	77	77.60	22.84
Psychosocial Health	Normal	346	83.53	17.16	5.93 (0.03)^a
	Borderline	104	77.88	19.60
	Abnormal	77	75.61	16.49
Emotional Functioning	Normal	346	77.47	20.93	2.84 (0.02)^a
	Borderline	104	71.54	25.16
	Abnormal	77	72.08	21.45
Social Functioning	Normal	346	87.08	20.08	5.70 (0.001)^b
	Borderline	104	82.31	22.58
	Abnormal	77	76.56	24.05
School Functioning	Normal	346	86.05	18.65	5.31 (0.001)^a
	Borderline	104	79.81	21.26
	Abnormal	77	78.18	18.28

Normal versus borderline p < 0.05, normal versus abnormal p < 0.02, borderline versus abnormal p > 0.0.

Normal versus borderline p = 0.14, normal versus abnormal p < 0.001, borderline versus abnormal p = 0.21.

Table 4.

Yoruba PedsQL^TM scores among healthy adolescents with chronic conditions (N = 527).

PedsQL^TM score	Chronic condition	N	M	SD	Fancova (p value)
Total	Yes	75	76.39	19.05	4.36 (0.01)
	No	452	82.65	17.01	4.36 (0.01)
Physical Health	Yes	75	77.46	21.26	2.62 (0.07)
	No	452	83.57	21.86	2.62 (0.07)
Psychosocial Health score	Yes	75	75.82	20.04	2.28 (0.01)
	No	452	82.16	17.30	2.28 (0.01)
Emotional Functioning	Yes	75	69.47	21.28	3.63 (0.03)
	No	452	76.51	22.01	3.63 (0.03)
Social Functioning	Yes	75	81.07	23.92	1.2 (0.29)
	No	452	85.19	21.05	1.2 (0.29)
School Functioning	Yes	75	76.93	24.41	5.41 (0.005)
	No	452	84.79	18.22

Convergent validity

As hypothesized, the Yoruba PedsQL^TM psychosocial health scales showed convincing negative correlations with the SDQ scales measuring emotional and conduct problems, hyperactivity/inattention and peer relationship problems and positive correlations with the SDQ pro-social scale. The Yoruba PedsQL^TM Physical Health scale, however, showed weaker correlations (Table 5).

Table 5.

Correlations between the Yoruba PedsQL^TM and the SDQ scale scores (N = 527).

SDQ scale	PedsQLTM
	Emotional Functioning	Social Functioning	School Functioning	Psychosocial Health	Physical Health
Emotional symptoms	–0.22^a	–0.19^a	–0.24^a	–0.27^a	–0.22^a
Hyperactivity/inattention problems	–0.05	–0.07	–0.08	–0.09^a	–0.08
Conduct problems	–0.7	–0.15^a	–0.10^a	–0.12^a	–0.07
Peers problems	–0.07	–0.10^a	0.01	–0.08	–0.06
Pro-social behaviour	0.04	0.02	0.05	0.05	0.08

Correlations are significant at the p < 0.05 level.

Discussion

The Yoruba PedsQL^TM as developed has features that make it culturally appropriate and semantically accurate. The results of the psychometric testing also suggest good internal consistency, reliability and construct validity.

The Yoruba version of the PedsQL^TM

The replacement of the word ‘block’ with ‘pole’ in the Yoruba versions was a strategic step in ensuring contextual appropriateness. Urban and regional planning is still a major environmental and developmental issue in Nigeria as many towns and cities in Nigeria are constructed in a very haphazard way (Daramola & Ibem, 2010; Mabogunje, 2002). As such, most people are not familiar with the ‘block’ system of estimating distance within the local geography as used in well planned towns and cities in developed countries. The ubiquitous electric poles that line the streets of most towns and cities in Nigeria, which are usually set about 100 metres apart, was the closest equivalent of the distance portrayed by the word ‘block’ in other countries. The phrase ‘walking more than the distance from one pole to the other’ was found to be well understood among respondents and they confirmed the popularity of such usage in estimating distance within locality among their peers.

In the same vein, unlike in developed countries, health services in Nigeria are still limited and under-utilized due to widespread alternative beliefs in the causation of illness, especially mental illness, in the world-view of the Yoruba people (Abdulmalik & Saleh, 2012; Atilola & Olayiwola, 2010; Morakinyo & Akiwowo, 1981; Odejide, Sanda, & Olatawura, 1978). In fact, studies have shown that traditional healers and prayers houses were the first points of call for the majority of parents seeking care for physical, emotional and behavioural problems for their children in Nigeria (Abdulmalik & Saleh, 2012; Onyeama, 2004). It was therefore pertinent to include alternative pathways to health seeking in the Yoruba PedsQL^TM. This informed the inclusion of ‘prayer houses or traditional healers’ in the item which enquired whether children had missed school to go to the ‘doctor or hospital’.

Psychometric properties of the PedsQL^TM

In general, the Yoruba PedsQL^TM has sufficient basic measurement characteristics. However, the normality parameters showed negative skewness of all scales, with higher scores, and the distribution was more peaked than expected. Additionally, the scales showed significant ceiling effects. These findings indicate that adolescents tended to rate their HRQOL with more positive values. This distribution and the tendency of adolescents to rate HRQOL more positively was observed in other studies and for other PedsQL^TM versions as well (e.g. Amiri et al., 2010; Reinfjell, Diseth, Veenstra, & Vikan, 2006; Stevanović et al., 2011; Varni et al., 2001).

The overall internal consistency reliability of the Yoruba PedsQL^TM is adequate, with the alpha value above 0.7 for all scores. In terms of internal consistency measured by Cronbach’s coefficients in adolescent samples as in this study, the Yoruba version had similar coefficients to the Iranian and Norwegian versions (Amiri et al., 2010; Reinfjell et al., 2006). Exploring the internal structure further, none of the separate constructs of QOL in the Yoruba version of PedsQL^TM was considered redundant or unrelated to QOL, considering that the inter-scale correlations were all below 0.7, but still correlated substantially.

Based on the CFA results, construct (factorial) validity could be supported for the Psychosocial Health scale of the Yoruba PedsQL^TM, a common finding among earlier versions of the instrument (e.g. Petersen et al., 2009; Stevanović et al., 2011). The fit indexes of the other three tested models also either showed acceptable model fit or were very close to acceptable levels (Meuleners et al., 2003). The originally hypothesized model of five factors (Limbers et al., 2008) and the four-factor model confirmed for the Swedish version (Petersen et al., 2009) had similar fit index values. In general, these results indicate that when considered as independent scores, the Yoruba PedsQL^TM Psychosocial Health score is appropriately represented by the items allocated in its scales, while Physical Health might not be appropriately represented by the items in one latent factor (scale). Further analyses of the construct validity demonstrated that the hypothesis stating that children with some mental health problems would have lower HRQOL score across Yoruba PedsQL^TM scales was confirmed in the study. Additionally, the healthy subjects reported significantly higher HRQOL than those with physical health problems. These findings agree with the results of construct validity for the original and other translated version found in a number of studies (e.g. Amiri et al., 2010; Limbers et al., 2008; Reinfjell et al., 2006; Stevanović et al., 2011)

However, this is only a preliminary evidence for construct validity, as we evaluated only the basic structure, without considering multiple groups CFA to examine factorial invariance like configural invariance, metric invariance and scalar invariance across different groups (Beckstead, Yang, & Lengacher, 2008). With a larger number of adolescents recruited from different backgrounds, it would be possible to include item response theory techniques to further test its construct validity.

The Yoruba PedsQL^TM Psychosocial Health scales showed stronger correlation with the SDQ compared with the physical health scales in this study. This is in support of the findings of other studies which have found strong correlations between other versions of PedsQL^TM and other measures of mental health of children and adolescents (Bastiaansen, Koot, Bongers, Varni, & Verhulst, 2004; Petersen et al., 2009). Furthermore, evidence was provided in this study as hypothesized that the Yoruba PedsQL^TM Physical Health scale measured a non-psychosocial construct, but still, additional work is required to fully confirm the convergent validity, considering that this scale strongly correlated with emotional problems reported by the SDQ. It is possible that some of the items in the Yoruba PedsQL^TM Physical Functioning scale measured some emotional aspects and not purely physical functioning. This could be an additional explanation for questionable factorial validity of this scale, a possibility that has been mooted in the report for the Serbian version of PedsQL^TM (Stevanović et al., 2011).

Clinical implications

Clinicians in the Yoruba-speaking part of Nigeria and other parts of the world where the language and culture is shared should take advantage of the availability and psychometric soundness of the Yoruba version of PedsQL^TM to ensure that they incorporate routine evaluation of HRQL in their assessment and monitoring of children and adolescents in their clinics. Studies – including the present one – have found a strong correlation between mental health morbidity and poorer HRQL (Bastiaansen et al., 2004; Petersen et al., 2009). Quality of life measures are also known to provide reliable information about the impact of specific treatments on well-being and functioning (Acquadro et al., 2003). Therefore, the Yoruba version of PedsQL^TM could be a reliable and objective way of monitoring progress among children with mental health problems in the Yoruba-speaking regions of Nigeria and beyond. To ensure further objectivity of assessments, HRQL measures can complement subjective assessments that are based on parental reports and clinical evaluation. Wide variations in HRQL measures and subjective clinical assumption of improvement can be used as an indicator for a need for further assessment. These assertions, however, depend on whether HRQL improves with improving mental health. Future studies can clarify this using structural equation modelling to assess the extent and direction of the relationship between mental health interventions and HRQL.

Limitations

Though this is the first translation and validation of the PedsQL^TM in a sub-Sahara African language or culture, the findings should be interpreted with consideration for some limitations. Firstly, the sample was sourced from school-going adolescents and from just two urban-based schools. A wider-catching sampling method which factored rural and urban adolescents as well as out-of-school adolescents would have been more representative. Secondly, no comparable QOL or well-being measure was used to evaluate convergent validity. Thirdly, test–retest was not done and, as such, the temporal stability and responsiveness of the Yoruba version was not reported. Lastly, the psychometric properties of the parent version were not assessed.

Conclusions

The Yoruba version of the PedsQL^TM as developed is a feasible, short and easily scored questionnaire for QOL assessment in adolescents. It contains items that are comprehensive, precise and relevant for HRQOL assessment in Yoruba-speaking regions, and it is very close to the original in terms of item, semantic and conceptual equivalence. The scales have appropriate internal consistency reliability sufficient for group evaluations, and good convergent validity for psychological constructs. Construct validity was also demonstrated using known-groups methods, but the factor structure is confirmed only for the Psychosocial Health score, and not for the entire measure. More work is needed regarding its true construct validity.

Footnotes

Acknowledgements

The authors wish to acknowledge Mr Abimbola A. Atanda of the Department of Social Work, University College Hospital Ibadan, Nigeria for graciously agreeing to serve as a member of the Yoruba translation team.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Author biographies

Olayinka Atilola MD, FWACP, FMCPsych., was born in Nigeria. He holds a Bachelors Degree in Medicine and he is a Fellow of the Nigerian and the West-African Colleges of Psychiatrists, being the highest medical certification achievable in Psychiatry in the region. He is also a Diploma-degree holder in Children and Youth Studies of the Institute of Social Studies of Erasmus University, Netherlands. He currently works as a researcher, lecturer and consultant Psychiatrist at Lagos State University College of Medicine/Teaching Hospital in Nigeria. His key area of interest is socio-cultural aspects of child and general adult psychiatry. He has authored at least 20 publications in reputable local and international peer-reviewed journals.

Dejan Stevanović, MD, was born in Serbia. He obtained his medical education at the Medical School University of Belgrade, where he was trained in Child and adolescent psychiatry and Neuropsychology. His main research interests are child neuropsychiatry, cognitive neurology, and pharmaco-economics. Specifically, Dejan is working and researching children with epilepsy and neurobiological syndromes, but the patient-reported outcome concept as well. Currently, he is affiliated with General Hospital Sombor in Serbia and has authored more than 25 articles in peer-review journals.

References

Abdulmalik

J. O.

Saleh

(2012). Pathways to psychiatric care for children and adolescents at a tertiary facility in northern Nigeria. The Journal of Public Health in Africa, 3, 15–17.

Acquadro

Berzon

Dubois

Leidy

N. K.

Marquis

Revicki

Rothman

, & PRO Harmonization Group (2003). Incorporating the patient’s perspective into drug development and communication: An ad hoc task force report of the Patient-Reported Outcomes (PRO) Harmonization Group meeting at the Food and Drug Administration, February 16, 2001. Value in Health, 6, 522–531.

Amiri

Ardekani

E. M.

Jalali-Farahani

Hosseinpanah

Varni

J. W.

Ghofranipour

…Azizi

(2010). Reliability and validity of the Iranian version of the Pediatric Quality of Life Inventory™ 4.0 Generic Core Scales in adolescents. Quality of Life Research, 19, 1501–1508.

Atilola

Olayiwola

(2011). The Nigerian home video boom: Should Nigerian psychiatrists be worried? Lessons from content analysis and views of community dwellers. International Journal of Social Psychiatry, DOI: 10.1177/0020764011408544 (Published online on 3 August 2011).

Bastiaansen

Koot

H. M.

Bongers

I. L.

Varni

J. W.

Verhulst

F. C.

(2004). Measuring quality of life in children referred for psychiatric problems: Psychometric properties of the PedsQL 4.0 generic core scales. Quality of Life Research, 13, 489–495.

Beckstead

J. W.

Yang

C. Y.

Lengacher

C. A.

(2008). Assessing cross-cultural validity of scales: A methodological review and illustrative example. International Journal of Nursing Studies, 45, 110–119.

Chassany

Sagnier

Marquis

Fullerton

Aaronson

(2002). Patient reported outcomes: The example of health-related quality of life a European guideline document for the improved integration of health-related quality of life assessment in the drug regulatory process. Drug Information Journal, 36, 209–238.

Cortina

M. A.

Sodha

Fazel

Ramchandani

P. G.

(2012). Prevalence of child mental health problems in sub-Saharan Africa: A systematic review. Archives of Pediatrics & Adolescent Medicine, 166, 276–281.

Costello

E. J.

Compton

S. N.

Keeler

Angold

(2003). Relationships between poverty and psychopathology: A natural experiment. Journal of the American Medical Association, 290, 2023–2029.

10.

Daramola

Ibem

E. O.

(2010). Urban environmental problems in Nigeria: Implications for sustainable development. Journal of Sustainable Development in Africa, 12, 124–143.

11.

Earth Trends. (2003). Economic Indicators – Nigeria. Available at: http://www.earthtrends.wri.org.

12.

Friedman

R. J

Chase-Lansdale

P. L.

(2002). Chronic adversities. In Rutter

Taylor

(Eds.), Child and adolescent psychiatry. 4th ed. (pp. 261–276). Oxford, UK: Blackwell Scientific.

13.

Goodman

(1997). The strengths and difficulties questionnaire: A research note. Journal of Child Psychology and Psychiatry, 38, 581–586.

14.

Kessler

R. C.

Amminger

G. P.

Aguilar-Gaxiola

Alonso

Lee

Ustun

T. B.

(2007). Age of onset of mental disorders: A review of recent literature. Current Opinion in Psychiatry 2007, 20, 359–364.

15.

Laws of Ogun State of Nigeria (2006). Child Rights Law. Vol. 1, Ogun State Ministry of Justice. Abeokuta, Nigeria: LexisNexis Butterworths (PTY) Ltd.

16.

Limbers

C. A.

Newman

D. A.

Varni

J. W.

(2008). Factorial invariance of child self-report across healthy and chronic health condition groups: A confirmatory factor analysis utilizing the PedsQLTM 4.0 Generic Core Scales. Journal of Pediatric Psychology, 33, 630–639.

17.

Limbers

C. A.

Newman

D. A.

Varni

J. W.

(2008). Factorial invariance of child self-report across socioeconomic status groups: a multi-group confirmatory factor analysis utilizing the PedsQL 4.0 Generic Core Scales. Journal of Behavioral Medicine, 31, 401–411.

18.

Mabogunje

(2002). Re-constructing the Nigerian city: The new policy on urban development and housing. Paper presented at the National Conference on the City in Nigeria, 11 April, Ile Ife.

19.

Meuleners

L. B.

Lee

A. H.

Binns

C. W.

Lower

(2003). Quality of life for adolescents: Assessing measurement properties using structural equation modeling. Quality of Life Research, 12, 283–290.

20.

Morakinyo

Akiwowo

(1981). The Yoruba ontology of personality and motivation. A multidisciplinary study. Journal of Social Biology and Structure, 4, 19–38.

21.

Nevitt

Hancock

G. R.

(2001). Performance of bootstrapping approaches to model test statistics and parameter standard error estimation in structural equation modeling. Structure Equation Model, 8, 353–377.

22.

Nunnally

Bernstein

I. J.

(1994). Psychometric theory (3rd ed.). New York, NY: McGraw-Hill.

23.

Odejide

O. A.

Sanda

A. O.

Olatawura

M. O.

(1978). Some socio-psychiatric attributes of patients utilizing the facilities of traditional healers in the city of Ibadan. Tropical and Geographical Medicine, 30, 115–119.

24.

Ojo

(2006). A global evaluation of the teaching and learning of Yorùbá language as a second or foreign language. Selected Proceedings of the 36th Annual Conference on African Linguistics (pp. 116–120). Somerville, MA: Cascadilla Proceedings Project.

25.

Onyeama

(2004). Childhood psychiatric disorders: Problems and prospects in Enugu, Nigeria. Journal of College of Medicine, 9, 49–53

26.

Perneger

T. V.

Leplège

Etter

(1999). Cross-cultural adaptation of a psychometric instrument: Two methods compared. Journal of Clinical Epidemiology, 52, 1037–1046.

27.

Petersen

Hägglöf

Stenlund

Bergström

(2009). Psychometric properties of the Swedish PedsQL, pediatric quality of life inventory 4.0 Generic Core Scales. Acta Paediatrica, 98, 1504–1512.

28.

Reichenheim

M. E.

Moraes

C. L.

(2007). Operationalizing the cross-cultural adaptation of epidemiological measurement instruments. Revista de Saúde Pública, 41, 1–9.

29.

Reinfjell

Diseth

T. H.

Veenstra

Vikan

(2006). Measuring health-related quality of life in young adolescents: Reliability and validity in the Norwegian version of the Pediatric Quality of Life Inventory 4.0 (PedsQL™ 4.0) generic core scales. Health and Quality of Life Outcomes, 14, 61.

30.

Robertson

Omigbodun

Gaddour

(2010). Child and adolescent psychiatry in Africa: Luxury or necessity? African Journal of Psychiatry, 13, 329–331.

31.

Schmidt

Bullinger

(2003). Current issues in cross-cultural quality of life instrument development. Archives of Physical Medicine and Rehabilitation, 84, 29–34.

32.

Shaw

Tschiwula

(2002). Developing citizenship among urban youth in conflict with the law. Environment & Urbanization, 14, 59–69.

33.

Stevanović

Lakić

Damnjanovic

(2011). Some psychometric properties of the Pediatric Quality of Life Inventory^TM Version 4.0 Generic Core Scales (PedsQL^TM) in the general Serbian population. Quality of Life Research, 20, 945–949.

34.

Terwee

C. B.

Bot

S. D.

de Boer

M. R.

van der Windt

D. A.

Knol

D. L.

Dekker

…de Vet

H. C.

(2007). Quality criteria were proposed for measurement properties of health status questionnaires. Journal of Clinical Epidemiology, 60, 34–42.

35.

UNICEF (2011). Adolescence: An age of opportunity. State of world’s children, 2011. Available at: http://www.unicef.org/sowc2011/fullreport.php.

36.

Varni

J. W.

Seid

Kurtin

P. S.

(2001). PedsQL 4.0: Reliability and validity of the Pediatric Quality of Life Inventory version 4.0 Generic Core Scales in healthy and patient populations. Medical Care, 39, 800–812.

37.

Varni

J. W.

Seid

Rode

C. A.

(1999). The PedsQL: Measurement model for the Pediatric Quality of Life Inventory. Medical Care, 37, 126–139.

38.

Wachs

T. D.

(2000). Necessary but not sufficient: The respective roles of single and multiple influences on individual development. Washington, DC: American Psychological Association.

PedsQL TM 4.0 Generic Core Scales for adolescents in the Yoruba language: Translation and general psychometric properties

Abstract

Background:

Aims:

Methods:

Results:

Conclusions:

Keywords

Introduction

Objectives

Methods

Setting

Study design and sample selection

Instruments

PedsQLTM

Strengths and Difficulties Questionnaire (SDQ)

Procedures and data analysis

Translation and cultural adaptation

Statistical procedures for psychometric testing

Descriptive analysis

Internal consistency reliability and inter-scale correlations

Construct validity

Convergent validity

Ethical consideration

Results

The Yoruba version of the PedsQLTM

Psychometric properties of the Yoruba version of PedsQLTM

Sample characteristics

Descriptive statistics

Internal consistency, reliability and inter-scale correlation

Construct validity

Convergent validity

Discussion

The Yoruba version of the PedsQLTM

Psychometric properties of the PedsQLTM

Clinical implications

Limitations

Conclusions

Footnotes

Acknowledgements

Funding

Author biographies

References

PedsQL^TM

The Yoruba version of the PedsQL^TM

Psychometric properties of the Yoruba version of PedsQL^TM

The Yoruba version of the PedsQL^TM

Psychometric properties of the PedsQL^TM