Background
An estimated 249.4 million children under the age of five years, in low and middle-income countries (LMICs), are at risk of failing to achieve their full neurodevelopmental potential [
18,
23]. There is considerable evidence of higher incidence and prevalence of neurodevelopmental impairment in LMICs compared to high-income countries [
9,
18,
23]. The measurement of children at risk of neurodevelopmental impairment (NDI) is essential to providing interventions to children at highest risk, especially in LMIC settings where healthcare resources are often limited [
14]. Moreover, quantification of NDI burden in and across populations is the first important step in measuring its long-term impact and evaluating the effectiveness of intervention strategies. Nevertheless, epidemiological data on normative child development are sparse or non-existent in many parts of the world, particularly in those areas where children are most at risk [
3]. This creates challenges for early child development (ECD) surveillance in LMICs, where ECD outcomes in children from disparate geographical and/or cultural contexts are evaluated using instruments that have not been subjected to a rigorous standardization, adaptation and cultural-customization process for use in these settings. This is particularly significant in ECD assessments because social, cultural and language-related factors can adversely affect a child’s understanding of a test item and his/her subsequent performance on the ECD measure [
30]. The use of inadequately adapted ECD assessment tools, at a population-health level, could result in over- or under-estimation of the prevalence of NDIs and, at an individual level, result in the misclassification of children as being at risk (or not) of NDIs. The life-course and public health consequences of either outcome are significant.
It is important to have a reliable ECD assessment tool that can be utilized to screen and assess the neurodevelopment of children residing in LMICs, as it enables healthcare systems residing in LMICs to: (1) conduct primary screenings, such as those held at daycare centers or local schools, to identify children at risk for NDIs; (2) refer children identified to be at risk for NDIs to primary and secondary healthcare settings, such as their primary healthcare provider or local rehabilitation clinic, respectively; and (3) assess the changes in functioning associated with newly-developed interventions that have been implemented either as a part of community-based or institution-based programs that aim to reduce the prevalence of NDIs within the country.
The importance of cross-cultural adaptation of ECD instruments, their adequate translation in the context of colloquialisms and accents; and the robust evaluation of adapted instruments has been highlighted in the World Bank’s Toolkit for the assessment of ECD in children aged under five years in LMICs [
11,
12]. The Toolkit encourages cross-cultural adaptations because the data from the ECD assessments can provide tremendous developmental insight in areas where formal testing is unavailable, or rarely practiced. Measures that are not amended tend to demonstrate biases and poorer performance levels because some items do not function in the same manner across cultures [
11]. Moreover, the interpretation of unfamiliar phraseology and its linkage to familiar lexical items to express roughly the same concept, are executive functions that can only be expected to emerge during mid to late childhood [
10]. It is widely acknowledged that sociolinguistics is critically important when administering a standardized assessment within a culture, including within the same language, as pronunciation, tonality, sentence placement and the words themselves are very important in language processing [
15].
One region in which there is limited data on ECD is the 700-island Caribbean basin with a regional population of 44 million [
34]. Of the 33 countries and individual territories in the region, ten of these are classified as LMICs [
38]. Previously, tropical infections represented the greatest disease burden in the region, however in recent years the burden has shifted to chronic diseases (during adulthood) and ECD disorders, including NDIs (during childhood) [
27]. Importantly, the few coordinated efforts to assess individual or groups of young children in the Latin American and Caribbean region suggest that approximately 15% of children are at high risk of NDIs [
23]. Despite this, there are, to our knowledge, no ECD tools adapted to and standardized in the cultural and linguistic context of the English-speaking Caribbean. We hypothesize that there are no culturally-adapted ECD assessments specifically for children living in the Caribbean because the region’s residents speak dialects of English, French or Spanish. European and American instruments have been directly exported there without a consideration of how linguistic and cultural differences between these disparate geographical and social populations might influence the test delivery and children’s performance.
The current study addresses this methodological gap in ECD measurement in the Caribbean by, for the first time, (1) adapting (from British to Caribbean English) and culturally customizing a rapid, multi-dimensional, international, standardized ECD instrument, The INTERGROWTH-21st Neurodevelopment Assessment (INTER-NDA), for the measurement of cognitive, motor, language and behavioral outcomes in young Caribbean children, (2) comparing the psychometric properties of the adapted to the original tool and (3) evaluating the ability of non-specialist child development assessors to administer and score the adapted version of the INTER-NDA.
Discussion
This study presents, to our knowledge, the first adaptation of the INTER-NDA to the linguistic and cultural context of the Caribbean, making the INTER-NDA the first ECD measure specifically adapted for use in the English-speaking Caribbean. The adapted INTER-NDA is very similar to the original INTER-NDA, with linguistic and cultural adaptations in 8 of the 37 items while simultaneously maintaining conceptual equivalence between the original and adapted versions. The inter-rater and test-retest reliability of the adapted INTER-NDA, and its internal consistency, were satisfactory across domains and comparable to the original INTER-NDA. Importantly, especially in the context of LMIC settings, we also showed that the adapted INTER-NDA can be administered and scored by non-specialist child development assessors at high levels of protocol adherence and agreement with an expert specialist assessor.
The results of this study provide evidence to support the use of the adapted INTER-NDA to measure ECD prevalence in the English-speaking Caribbean and for non-specialist child development assessors, trained and standardized in the INTER-NDA, to use the tool to conduct ECD evaluations. Previous studies have shown that the INTER-NDA can be administered in school settings in Mexico, and in research settings in Brazil, India, Italy, Kenya and the UK, by non-specialist assessors at high levels of reliability and protocol adherence [
14,
31]. These studies have focused on the translation, back translation and cultural adaptation of the INTER-NDA, developed initially in British English, into non-English languages. This study extends this work by culturally and linguistically adapting the INTER-NDA for use in an English-speaking, yet culturally and geographically diverse, LMIC population, i.e. the English-speaking Caribbean. Moreover, we have presented a methodological template for this process, which we hope can be applied to the adaptation and subsequent pilot testing of other ECD tools for use in various Caribbean settings.
Strengths and limitations of the study
This study is important in a number of ways. First, consideration was given to different lexical items to express similar concepts in British and Creole English; for example, “throw” and “pelt”; “cubes” and “blocks”; and “teapot” and “kettle. This is particularly necessary in the Grenadian context where children are more familiar with Creole English, rather than British English, during early life. Second, care was taken to ensure that the components of the INTER-NDA’s kit were familiar to Caribbean children, and commonly encountered in Caribbean households. The use of items unfamiliar to the average child’s repertoire of household and play-related exposures, such as a puffed rice grain or a maize bean [
16,
19]; or items which the child may be forbidden from playing with, such as a matchbox [
16], are commonly overlooked factors that can skew ECD assessment results. Third, we assessed whether non-specialist child development assessors, in the Caribbean context, can measure ECD outcomes on the adapted INTER-NDA. In LMIC settings where reliance for ECD assessments on specialist professionals for their administration, scoring and interpretation is one of the key rate-limiting steps to the scalability of ECD surveillance and for the identification of children at risk who may benefit from interventions. Although non-specialists have been previously shown to administer and score the INTER-NDA reliably, this study is, to our knowledge, the first effort to assess this in a Caribbean setting and provides evidence to support ECD evaluations in Grenada by non-specialist child development assessors.
There are a number of limitations to consider. First, the sample size used to determine inter-rater reliability and test-retest reliability was small – 7 trainee-assessors performed 3 randomly-assigned child assessments each, for a total of 21 assessments. Thus, test-retest reliability and inter-rater reliability statistics should be interpreted with caution. Second, this study was carried out in Grenada, and while socio-cultural similarities exist among Caribbean nations, the adapted INTER-NDA should be piloted and assessed in other Caribbean settings to establish firmer reliability. Third, this adaptation of the INTER-NDA is limited to the English-speaking Caribbean, despite substantial proportions of the region’s population being native in Spanish and French. Fourth, our results are restricted to the INTER-NDA’s age range of assessment (22 to 30 months) and, as such, ECD tools for the assessment of a wider age range of children should be adapted and standardized in order to significantly impact ECD surveillance in the Caribbean across the early childhood. Fifth, we were not able to assess internal consistency of the gross motor domain within the INTER-NDA as a result of measurement challenges (i.e., ceiling effects) with one of the 3 items that make up this domain score.
Context of the study
An essential condition for achieving the United Nations Sustainable Development Goal (UN SDG) 4.2 (‘ensure that all girls and boys have access to quality early child development, care and preprimary education so that they are ready for primary education’) [
33] is the measurement of ECD outcomes at scale, in order to identify children at risk and to make cross-population comparisons [
14]. While a multitude of tools to measure ECD risk exist, none are culturally adapted and standardized for use in the Caribbean setting, especially for administration by non-specialists. By linguistically and culturally adapting the INTER-NDA to the context of the English-speaking Caribbean, without compromising its conceptual integrity, we have, for the first time, produced a practically applicable, culturally relevant ECD measure specifically adapted to this setting. The INTER-NDA is a unique clinical tool for use across all healthcare systems to measure neurodevelopmental milestones and associated behaviors in 2-year-olds uniformly and at scale, and to identify children at risk of NDIs who would benefit from specialist referral and further investigation [
14]. It is our hope that the adapted INTER-NDA will complement the INTERGROWTH-21st Project’s international INTER-NDA standards [
14] for the measurement of ECD outcomes in the Caribbean while simultaneously providing a methodological template for the adaptation of child developmental measures to cultural and linguistic contexts.
Conclusion
This study represents, to our knowledge, the first adaptation of an ECD instrument to the English-speaking Caribbean. This is important for standardized, robust ECD measurement in the region, at scale, and supports the inclusion of Caribbean infants in international efforts at ECD surveillance. Furthermore, this study highlights the importance of the linguistic and cultural adaptation of ECD measures, even between settings which appear to share the same primary language but in which social, cultural, geographic and economic contexts vary.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.