Developmental screening using caregiver report : an evaluation of screening tools and childhood developmental delays in South Africa

Abstract

Introduction: Early childhood intervention (ECI) services are recommended to be integrated in primary healthcare (PHC) in underserved communities. Currently, developmental screenings are not routinely carried out at PHC level and there are shortages of appropriately skilled staff and resources to conduct screening, assessment and provide follow-up care. As a result, there is growing interest in parent-based screening with parental concerns often predictive of developmental delay. This research project aimed to explore the use of developmental screening tools and caregiver report from diverse low-income settings. Three study objectives were identified: the first aimed to establish a developmental profile for children under the age of three years from a low-income South African community; the second objective was to provide an overview of studies that used Parents’ Evaluation of Developmental Status (PEDS), the PEDS: Developmental Milestones, and the PEDS tools in combination, to identify developmental delays in children by means of a scoping review; and lastly to compare the outcomes of two caregiver report tools – namely the Bayley Scales of Infant and Toddler Development III (BSID-III) and PEDS tools – in an at-risk infant population from a low-income South African community. Method: Study I (Developmental characteristics of young children in a low-income South African community) utilized convenience sampling to select 353 caregivers and their infants, aged between 3-36 months from a primary health care clinic. The BSID-III was used for diagnostic developmental assessments by a speech-language pathologist (SLP) and final year SLP students. Study II was a scoping review on the use of the PEDS and PEDS:DM, as well as the PEDS tools in combination, which was conducted to provide an overview of these tools and their use to detect developmental delays globally. Five electronic databases were searched. Narrative synthesis was used during data analysis. In Study III, a cross-sectional, within-subject, comparative design was employed to determine the overall and domain-specific performance of the PEDS tools smartphone application and the BSID-III to detect developmental delays in 174 young children aged 3-18 months. Data was collected at a PHC in Mamelodi, an underserved high-risk community, in South Africa. Results and conclusion: Developmental characteristics of young children in a low-income South African community (Study I) identified signs of developmental delay in 51.8% (n=183) of 353 children. Prevalence of developmental delay increased with age from 33.1% for children younger than 12 months to 61.7% and 66.3% for children between 13 to 24 months and 25 to 36 months of age respectively. Females were 1.82 times more likely to have no signs of a developmental delay; they were 2.301 times more likely to have no signs of delay in the motor and 2.601 times more likely to have no signs in the adaptive behaviour domains. One-third (33%) of children presented with low levels of functioning in the adaptive behaviour domain. One hundred and one (28.6%) of participants across all age groups displayed superior social-emotional ability. In the scoping review of the use of the PEDS, PEDS:DM and PEDS tools (Study II), the search strategy identified 1468 records. Thirty articles, ranging from 2003 to 2020, conducted in both high-income countries (HIC) and low- and middle- income countries (LMIC) , qualified for final inclusion. Studies conducted in HICs primarily focus on the screening of special population groups, as well as the comparison of validated tools. Studies conducted in LMICs focused more on translations, combination of the PEDS tools, validation of the tools, and the use of an app-based tool (mHealth). High referral rates are typically obtained with the PEDS and PEDS:DM when administered in low-income settings, where at-risk populations are more prevalent, and where cultural differences are a concern. Detecting developmental delays in infants from a low-income South African community: comparing the BSID-III and PEDS tools (Study III), the PEDS tools identified 56% (n=97), and the BSID-III 35% (n=61) of the 158 children with possible developmental delays, with an overall agreement of 65% between tests. The PEDS tools referral rate was significantly higher (p=0.004) than that of the BSID-III, as domain specific outcomes (language, motor, and social emotional) of the PEDS tools and BSID-III was compared and showed that twice the number of participants were identified as having developmental delay on the PEDS tools in relation to the BSID-III in all domains. Study III contributed to information on the developmental characteristics of young children, and evaluated tools for use in developmental screening for these children, which can inform intervention and public health policy in South Africa. Despite poor developmental screening outcomes one third of children, in the low-income South African community, presented with superior social-emotional skills. This could possibly be attributed to familial structures and relationships. One third presented with low levels of functioning in the adaptive behaviour domain which could be attributed to cultural differences. The agreement between developmental assessment outcomes across the BSID-III and the PEDS tools was poorer than expected. The high-risk nature and young age cohort (<18 months) may have contributed to these outcomes. Findings raised concerns about the outcomes of the BSID-III or PEDS tools in isolation for screening and assessment of developmental delay in infants from LMICs like South Africa. Additionally, more research from diverse settings and LMICs at large with regards to the PEDS was identified as an outstanding need. Future research should evaluate performance of the PEDS tools mHealth version in older preschool children (between 2 and 5 years) to ascertain the influence of age. This research project identified future research needs into the validity of the PEDS tools and BSID-III for young infants and children in LMICs, that needs to be prioritised prior to large scale implementation.