A district level population-based study of structural inequalities in autism pathways

Studies show that early identification of autism improves developmental outcomes, but current pathways to diagnosis and support are dysfunctional, with systemic problems and inequalities at each ‘stage’ of the pathway. This project aimed to understand inequalities relating to ethnicity by combining data and lived experience information about autism pathways within the Bradford district. Analysis 1 explored the prevalence of autism diagnoses across different demographic groups and showed a lower diagnosis rate in South Asian children in comparison to white children. It also showed boys being diagnosed earlier than girls, however this finding only applied to boys from white and other ethnic groups. In Asian ethnic groups, girls were found to be diagnosed earlier than males. Analysis 2 investigated the power of the Early Years Foundation Stage Profile (EYFSP) to predict autism diagnosis by ethnicity, sex, and socioeconomic position. The analysis replicated Wright et al. (2019) but further showed children of Pakistani heritage with a low EYFSP score had two times decreased odds of being diagnosed with autism relative to children from white ethnic groups. Finally, qualitative analyses combined these data insights with the lived experiences of South Asian heritage parents and the views of Bradford autism service providers (including practitioners) and policymakers. Thematic analysis identified three overarching themes: (i) identification and diagnosis barriers and facilitators; (ii) barriers and facilitators of support and provision; (iii) suggestions for improving identification, diagnosis, and support. Together, these findings show population level inequalities for South Asian children within Bradford, increase our understanding of these inequalities, and indicate possible solutions. Recommendations include optimising community practices, healthcare, education, and the wider system. Further research on the inequalities operating in different areas of England is needed, and would benefit from similar use of connected data and qualitative techniques to explore the impact of multiple coexisting difficulties on likelihood of timely diagnosis across different ethnicities.