Improving access to early intervention for autism - findings from a proof-of-principle cascaded task-sharing naturalistic developmental behavioural intervention in South Africa

Abstract

Background Despite the high number of children living with neurodevelopmental disabilities in sub-Saharan Africa, access to early intervention is almost non-existent. It is therefore important to develop feasible, scalable early autism intervention that can be integrated into systems of care. While Naturalistic Developmental Behavioural Intervention (NDBI) has emerged as an evidence-based intervention approach, implementation gaps exist globally, and task-sharing approaches may address access gaps. In this South African proof-of-principle pilot study, we set out to answer two questions about a 12-session cascaded task-sharing NDBI - whether the approach could be delivered with fidelity, and whether we could identify signals of change in child and caregiver outcomes. Methods We utilized a single-arm pre-post design. Fidelity (non-specialists, caregivers), caregiver outcomes (stress, sense of competence), and child outcomes (developmental, adaptive) were measured at baseline (T1) and follow-up (T2). Ten caregiver-child dyads and four non-specialists participated. Pre-to-post summary statistics were presented alongside individual trajectories. Non-parametric Wilcoxon signed rank test for paired samples was used to compare group medians between T1 and T2. Results Caregiver implementation fidelity increased in 10/10 participants. Non-specialists demonstrated a significant increase in coaching fidelity (increases in 7/10 dyads). Significant gains were seen on two Griffiths-III subscales (Language/Communication - 9/10 improved, Foundations of Learning - 10/10 improved) and on the General Developmental Quotient (9/10 improved). Significant gains were also seen on two Vineland Adaptive Behaviour Scales (Third Edition) subscales (Communication - 9/10 improved, Socialization - 6/10 improved) and in the Adaptive Behaviour Standard Score (9/10 improved). Caregiver sense of competence improved in 7/10 caregivers and caregiver stress in 6/10 caregivers. Conclusions This proof-of-principle pilot study of the first cascaded task-sharing NDBI in Sub-Saharan Africa provided fidelity and intervention outcome data which supported the potential of such approaches in low-resource contexts. Larger studies are needed to expand on the evidence-base and answer questions on intervention effectiveness and implementation outcomes.

Figure 1

Formative activities that informed proof-of-principle cascaded task-sharing NDBI.

Figure 2

Caregiver coaching session structure of the proof-of-principle cascaded task-sharing NDBI.

Figure 3

Consort Flow Diagram

Figure 4

Box plots of Pre- and Post-ESDM Caregiver and Coach fidelity: median, IQR, and Pre- and Post-individual Caregiver and Coach fidelity trajectories.

Figure 5

Box plots of Pre- and Post-Griffiths Scales of Mental Development-III (Griffiths-III): A: Gross Motor; B: Personal/Social Communication; C: Foundations of Learning; D: Language & Communication; E: Eye-Hand Coordination; F: General Development DQs with median and IQR; and Pre- and Post-individual Griffiths-III DQ trajectories.

Figure 6

Box plots of Pre- and Post-Vineland Adaptive Behaviour Scales, Third Edition (VABS-3): A: Communication; B: Socialization; C: Daily Living Skills; D: Adaptive Behaviour Subscale Standard Scores with median and IQR; and Pre- and Post-individual VABS-3 Subscale Standard Score trajectories.

Figure 7

Box plots of Pre- and Post- Caregiver Sense of Competence Total Scores: median, IQR, and Pre- and Post-individual Caregiver Sense of Competence Total Scoretrajectories.

Figure 8

Box plots of Pre- and Post- Parenting Stress Index Total Score: median, IQR; and Pre- and Post-individual Parenting Stress Index Total Score trajectories.