Traditional and religious treatments administered to children with autism spectrum disorder in South Africa: a qualitative study of practices, composition and risk

Introduction The appraisal of treatments for autism spectrum disorder (ASD) has largely focused on interventions delivered within formal healthcare systems, particularly in high-income settings. In South Africa, children with ASD may receive concurrent care from hospitals, traditional health practitioners and religious leaders, yet the practices administered outside formal healthcare remain poorly documented. This study aimed to identify these practices and interpret their potential risks in relation to published evidence of harm. Methods A qualitative study was conducted at a public tertiary hospital in Soweto, Johannesburg. Semi-structured interviews were undertaken with 24 parents and caregivers and 10 hospital clinicians, alongside two focus groups involving traditional health practitioners and religious leaders, respectively. As practitioner participants did not recognise the term autism, they were briefed on its DSM-5 features before discussing practices previously administered to children presenting with similar characteristics. Their accounts therefore did not necessarily concern children with confirmed ASD diagnoses. Data were analysed using reflexive thematic analysis, followed by deductive recoding of reported practices into a route-of-administration taxonomy. Results Twenty-five distinct practices were identified. All children in the sample had received at least one intervention outside formal healthcare, with most receiving several alongside hospital treatment. Ingested and rectally administered preparations presented particular safety concerns because their constituents were undisclosed, dosing relied on clinical impression and paediatric formulations were unavailable. Neither caregivers nor clinicians could identify the substances administered, while practitioners declined to disclose their composition. Practitioners categorised ASD-related presentations within a broader conception of insanity, and one non-clinician proposed cutting tissue beneath a child's tongue to facilitate speech. Reported benefits, including explanatory frameworks, family connectedness and sustained hope, related primarily to caregivers' experiences rather than documented therapeutic improvements in children. Discussion The findings identify potentially harmful and insufficiently documented interventions administered alongside formal ASD care. Their safety cannot be adequately assessed without information about their constituents and administration. Route-based risk triage, non-punitive disclosure during clinical consultations and protected constituent disclosure are recommended. As the study was conducted at a single urban tertiary hospital, its findings are context-specific and should not be generalised beyond the setting investigated.