Serving the underserved: How can we reach autism families who systemically miss out on support?

Real progress has been made in autism research that tries to bring direct benefits for autistic individuals and their families. Screening and diagnostic instruments are available to aid the identification and diagnosis at increasingly young ages (Lord et al., 2022) and intervention research has started to establish an evidence base for autism interventions, especially those directed at young children (Sandbank et al., 2020). A recent special issue of this journal focused on how to apply these evidence-informed principles to community settings, using insights from the burgeoning field of implementation science (Boyd et al., 2022). In parallel, there have been tangible developments in global autism research, with increasing literature on autism in non-Western and low-resource contexts (Divan et al., 2021). This editorial aims to tie these exciting developments together to consider: How can we best leverage this research progress to make a real difference to the lives of individuals with autism and their families who are least likely to see immediate benefit from this research – those who are most vulnerable and underserved?

Key factors that may make a family vulnerable and at risk of missing out on support include poverty, limited access to education, poor literacy, being subject to structural racism and other forms of discrimination and stigma. Families caring for a child with autism are more likely to be affected by poverty because the child’s needs may prevent the caregiver from generating an income (Minhas et al., 2015; Tilahun et al., 2016). In countries where (mental) healthcare is paid for out-of-pocket, seeking help may result in catastrophic payments, meaning that the costs comprise a very high portion of household income (de Leeuw et al., 2020; Gona et al., 2016). Families from lower-income households report poorer access to care (Bishop-Fitzpatrick & Kind, 2017) and poorer quality (Magaña et al., 2015) of care for their child with autism. Even when children are diagnosed and families are offered intervention, barriers linked to poverty (e.g. transportation costs) may prevent them from attending the intervention (Koerting et al., 2013) or having the time and capacity to actively engage with the intervention (Szlamka et al., 2021; Tekola, Girma, et al., 2020).

Low literacy is common in low-income countries but also affects vulnerable groups in high-income countries. Caregiver-reported screening instruments developed and tested in primarily highly educated families may not work as well in groups with limited literacy (Khowaja et al., 2015). Similarly, interventions developed and tested with highly educated caregivers may require adaptation when applied to families with low literacy and limited prior education.

Families with a cultural minority status in the country of residence may be less likely to access services and report lower satisfaction when accessing such services (Magaña et al., 2015). This is likely due to systemic factors. For example, clinicians may be less likely to recognise autism in ethnic minority children than in children from the cultural majority (Begeer et al., 2009) and in some contexts autism identification rates in ethnic or racial minority children (Begeer et al., 2009; Maenner et al., 2021) or children from Indigenous backgrounds (Bailey & Arciuli, 2020) lag behind identification in White children. Access barriers linked to racism may also be intertwined with language diversity. Latino parents living in the United States with low English language proficiency reported receiving poorer primary healthcare services than those with good English language proficiency (Pippins et al., 2007).

Stigma towards autistic people and their families occurs globally and across different cultures (de Leeuw et al., 2020; Han et al., 2022). In low-resource contexts where awareness of autism is low, stigma may be particularly strong (Tekola, Kinfe, et al., 2020). Stigma may affect how families interpret the signs and symptoms of their child, when and where they seek help (de Leeuw et al., 2020) and how they engage with intervention services (Guler et al., 2018; Koerting et al., 2013).

How to address these barriers? For a better understanding of how to reach and support vulnerable families, it is worth looking at research from the global South, where ethnic and language diversity is common and where research and support systems by necessity need to be developed around resource constraints. Take South Africa as example, a country with stark economic inequalities, with 11 official languages spoken in a population of 60 million. Smith et al. (2017) adapted and translated the ADOS-2 (Autism Diagnostic Observation Schedule, Second Edition) to Afrikaans and evaluated the use of this diagnostic instrument in Afrikaans-speaking families with low/middle socioeconomic background. The authors, noting frequent code switching between Afrikaans and English by participants and differences in dialect spoken by the test administrator and the participants, provided practical recommendations on how to administer the tool in a context of language diversity. Smith et al. (2017) also investigated the appropriateness of ADOS activities using sociodemographic and ethnographic questionnaires and made suggestions for adaptations of components not familiar or unavailable to participants.

While it is best practice to pre-test tools during instrument development or adaptation, this pre-testing is not always employed using a wide range of participants. Pre-testing can reveal important misunderstandings (Borissov et al., 2021; Ceccarelli et al., 2021) or contextual mismatches (Smith et al., 2017) and this may be especially relevant for respondents with limited education or when implementing an intervention in a context of economic disparity or cultural or racial diversity. Identifying and addressing these issues prior to release of a tool or intervention can avoid systemic barriers when the instrument is implemented in practice.

In Ethiopia, an ethnically diverse low-income country in which over 80 languages are spoken, Tekola, Girma, et al. (2020) adapted the World Health Organization’s Caregiver Skills Training programme to make the programme suitable for use in caregivers with limited education and low literacy in a context with high stigma. The adapted programme focused on oral discussions rather than written text and allowed for more extensive explanations and discussions, including addressing parental feelings of guilt and stigma experience. Following a pre-test of the programme, some complex concepts and programme components were simplified or adapted further to suit the context and culture.

Like South Africa, Chile is a country with pronounced income disparities among its population (‘GINI Index Data’, 2020). Roman-Urrestarazu et al. (2021) evaluated a model to increase autism detection by embedding an autism screening tool into an existing Chilean health check programme. Families receive a monthly child benefit payment if their children’s health check-ups are up to date; this conditional cash transfer model is linked with high uptake of the health check programme. By including an autism screener in the health check programme, an autism diagnostic pathway was established that is more easily accessible to the lowest-income families in the country.

Research and practice from the global South can also provide useful pointers when considering who an intervention should be for, and who should facilitate the intervention. While research in the global North tends to focus on autism-specific early interventions delivered by specialists, interventions developed for use in low-resource contexts often consider low-intensity interventions for a wider group of participants and delivered using a ‘task-sharing’ approach, through a skilled but non-specialist facilitator (Divan et al., 2021). An example is the World Health Organization’s Caregiver Skills Training programme (Salomone et al., 2019), an intervention targeting caregivers of children with developmental delays and disabilities. Focusing on a wider range of disabilities and delays allows for children to be referred to the programme in absence of a specialist diagnosis – a prerequisite often unavailable in low-resource contexts. The programme is delivered by non-specialists with ongoing specialist supervision. While such task sharing is crucial in contexts where specialists are rare, a similar approach may also suit higher-income countries trying to grapple with sustainable scale-up of services in a context of growing demand.

The task-sharing model does not take away the need for more intensive and specialist-led support for the more complex and highest need families (de Vries, 2016). A range of intervention studies (e.g. Brookman-Frazee et al., 2022; Green et al., 2022) have explored how to move evidence-based autism interventions from the clinic to community health or education settings. Can we learn from the field of global mental health to consider additional avenues to reach the underserved? In low-resource contexts, it is common for parents to first seek help from traditional healers or religious institutions (Hussein et al., 2012; Tilahun et al., 2016), especially among low economic groups within these contexts (Gureje et al., 2015). Initiatives in HIV (Berkley-Patton et al., 2019) and mental health (Puffer et al., 2016) research, for example, have started to actively collaborate with church leaders to develop and evaluate church-based interventions. A research question worth exploring is whether similar initiatives may work to reach underserved autism families.

Whether transferring an intervention from lab to classroom, from a high-income to a low-income context, or from clinic to a religion-based location, careful adaptations are needed to ensure the intervention is suitable for the new context. Frameworks are available to guide the cultural adaptation of intervention content (Bernal et al., 1995; Chu & Leino, 2017; Resnicow et al., 2000) as well as the adaptation process (Moore et al., 2021; Wingood & DiClemente, 2008). A key component of intervention adaptation is to work directly with the community to ensure the intervention suits their needs and is culturally acceptable (Hoekstra et al., 2018; Wingood & DiClemente, 2008).

While the vast majority of autism research is still conducted in high-income contexts (Durkin et al., 2015), and skewed towards highly educated, White and ethnic majority families (Davenport et al., 2018; West et al., 2016), real progress is being made in improving global representation in autism research. By taking on board the lessons learnt in research from the fields of implementation science and global mental health, we will slowly be able to move the dial and make a concrete difference in the lives of the world’s most vulnerable and underserved autistic people.