South African cultural adaptation and Northern Sotho translation of the modified checklist for autism in toddlers, Revised, with follow-up

Abstract

Rationale, aim and objectives There are currently no culturally sensitive, South African autism-specific screening instruments available. The global increase in prevalence, lack of local data, limited knowledge and late identification of autism in South Africa prompted the researcher to translate a universally known autism-specific screening instrument. The aim was to culturally adapt and translate the Modified Checklist for Autism in Toddlers, Revised with Follow-up™ and to determine the reliability and concurrent validity of the newly adapted and translated checklists. Method and results A mixed-method research design was employed in the study. This design allowed for the qualitative analysis of expert opinions in Study 1, and feedback from caregivers in Study 2 and 3. Quantitative data were generated in Study 2 and 3 when the two checklists were compared with one another and with the results of the VABS-3. For the three studies, experts and caregivers of children aged between 18 and 48 months were utilised as participants. In Study 1, Northern Sotho speech-language therapists identified culturally biased items in the M-CHAT-RFTM which were adapted. The adapted version of the checklist was scrutinised by an expert panel, translated into Northern Sotho and back-translated in English by professional translators. A comprehensive panel discussion and independent review of the linguistic, construct and technical equivalence of the checklists followed. The rigorous process of cultural adaption and translation resulted in an adapted English M-CHAT-R/F and a Northern Sotho translation of the source checklist. The two versions of the M-CHAT-R/F were used in a pilot study with 21 Northern Sotho caregivers. Caregivers were selected as the M-CHAT-R/F is a caregiver checklist and the comprehensibility of the new versions had to be determined. Participants completed both M-CHAT-R/F versions in randomised order along with a socio-demographic questionnaire. The Wilcoxon Signed Ranked test showed near-perfect agreement between the two checklists. There was no indication for any changes to the checklists. A larger-scale study (n=158) was conducted, comparing the checklists with each other and the Vineland-3. The results showed a low Cronbach alpha value but confirmed near-perfect agreement (p<0.001) between the two checklists. Concurrent validity was established between the Vineland-3 sub-domains and the Northern Sotho M-CHAT-R/F with a significant association at the 5% level. A total of 33.5% developmental delay was found in the reference sample (n=158). In line with international studies, it is recommended that autism-specific screening be conducted along with developmental screening or assessment. A marginal majority of participants (55.1%) preferred the Northern Sotho M-CHAT-R/F while a smaller group (44.9%) chose the adapted English checklist. The Northern Sotho majority preference aligns well with the accuracy of the results. The adapted English version may be used more often as English is one of the prominent languages of learning and teaching in South Africa. The larger-scale study identified some checklist items that required refinement. A phrase to clarify understanding was inserted. Gender-biased terminology was changed. Conclusion This study adapted and translated the M-CHAT-R/FTM resulting in two new versions. With the reliability outcomes, the checklists show linguistic, technical and construct equivalence. Predictive value must still be established. A functional framework for the cultural adaptation and translation of screening instruments is presented. Two feasible, reliable and concurrently valid autism-specific screening instruments are now available for use by caregivers, health care workers, and early childhood development practitioners in South Africa. Further investigation of developmental delay among preschool children in South Africa is required.