In this secondary analysis we examined the feasibility, acceptability and impact of a programme of early detection and intervention for children with developmental disability, in a rural sub-Saharan African setting. This contributes to the important area of early intervention for children with developmental disabilities, for which data is still lacking particularly in LMICs. Our mixed-methods evaluation found early detection and intervention to be feasible and acceptable in this setting, and showed positive impacts on HCW knowledge, referral rates of affected children, and family quality of life. Several important programmatic barriers were identified including stigma, poverty, gender equality and geographical spread of enrolled families.
Early child development training and mentorship for HCWs was found to be feasible and acceptable with high attendance at initial training, and significant improvements in knowledge and confidence. However, only two-thirds of HCWs were able to return for the refresher training six months later. Difficulties in delivering ongoing training and supervision to HCWs is a common issue in ECD programming, with high staff turnover and poor retention cited as a challenge [24]. Commitment by district health services to reduce staff rotation may improve retention and facilitate retention of knowledge and skills. Despite challenges in sustainability of training, a significant increase in the annual referral rate for children with developmental disabilities was seen. Whilst referrals were not always assessed using a validated neurodevelopmental tool due to time constraints, all children were clinically assessed by trained therapists and classified according to a recognised classification system [25]. Almost all referrals were found to be appropriate with the number of referred children without developmental disability reported as few by trained therapists. This is an important finding for the field, as there is no universally agreed referral threshold for early intervention programmes for children with developmental disabilities [26]. However, an international group is currently field testing recently developed population-level metrics (Global Scales for Early Development), and future plans include creating a global individual-level measure for screening [27].
Our community-based, participatory early intervention programme was found to be feasible and acceptable to facilitators (‘expert-parents’) and enrolled families, although notable barriers to access were identified including geographical spread of participants and poverty, making transport a substantial challenge for many families. Despite these challenges, over three-quarters of families had satisfactory attendance of 6 or more modules. Family engagement with the programme was high, with groups running at maximum capacity depending on the availability of facilitators and location of residence. Caregivers reported that they felt services were more accessible, an important benefit of the programme given the widespread paucity of services for children with disability in LMICs [28]. Whilst efforts were made to run groups as locally as possible to families, due to the large geographical spread they often still had to travel some distance. Creating a sustainable delivery platform with more geographically diverse facilitators, would increase access and attendance to the programme and facilitate a local network of peer support for families. However capacity for ongoing mentoring by trained therapists is an important consideration in scale-up to ensure maintenance of high fidelity programme delivery, as highlighted in previous literature [24].
Existing literature shows that caregivers of children with developmental disabilities report more negative experiences than caregivers of children without developmental disabilities, and that this is exacerbated in low-income settings [29]. This includes caregiver mental health problems such as depression and anxiety [30–33]. Previous work in Uganda found that mothers faced substantial social, emotional and financial difficulties, and stigma [2], which can lead to social exclusion [34]. A significant improvement in family quality of life was seen in the pre-post evaluation with the largest effects in social and emotional functioning. This was supported by caregivers reporting an increase in their knowledge, confidence and skills to care for their child, and a change in their attitudes and that of family members, reducing emotional stress and self-stigma. Improved community engagement has also been seen in earlier pilot work in the capital city of Kampala [19] and in the evaluation of a similar programme in Ghana [33]. Caregivers (the vast majority being mothers) reported that the programme empowered them to care for their child, access services, and tackle stigma. This empowerment has the potential to address gender inequality which can have wide-ranging positive impact on children and their families. In addition, the proportion of children with moderate-severe malnutrition reduced following completion of the intervention, which is important due to the widespread issue of poor nutrition in this population and the fact that development can be further impaired by this [34]. The reduction was not significant, although as nutritional status would commonly worsen over time in children with developmental disabilities, this still may represent positive impact. Caregivers also reported that feeding had improved after obtaining skills through the programme.
Strengths and limitations
This study adds to the limited evidence base on intervention programmes for children with disability in low-resource settings. We utilised a mixed-methods approach to offer a more comprehensive evaluation, and the quantitative and qualitative findings supported each other to show positive impact in both early detection and intervention. Important programmatic barriers were identified including regular rotation of HCWs being a barrier to sustainability, and wide geographic spread and poverty as barrier to access for families in this context.
Our evaluation has some limitations. Firstly, interpretation of findings is limited by study design and a range of challenges in conducting rigorous monitoring and evaluation in this resource-limited setting. The relatively small sample size, low number of participants with pre- and post-evaluation data and lack of a control group, may limit data interpretation. Evaluation data were on occasion collected by programme staff, which may have introduced bias and led to overestimation of impact. However, encouragingly the qualitative findings, led by an independent social scientist not involved in programme delivery, mirrored the positive quantitative findings. Finally, findings may not be generalisable to other settings due to the rural context with high levels of poverty, however the results do mirror those reported during programme piloting in an urban setting [19].