A Systematic Review of What Barriers and Facilitators Prevent and Enable Physical Healthcare Services Access for Autistic Adults

We identified six papers investigating the barriers and facilitators to healthcare as reported by autistic adults. Despite the limited literature, there are similar barriers and facilitators reported across the six studies; efforts to devise interventions and measurement tools are underway, and initial steps towards evaluating the effectiveness of implementing change have been evidenced (Nicolaidis et al. 2016; Dern and Sappok 2016; Saqr et al. 2018; Vogan et al. 2017; Nicolaidis et al. 2015; Raymaker et al. 2017).

Nicolaidis et al. reported that verbal communication skills were an autism-related factor affecting healthcare access (Nicolaidis et al. 2015). Social or communicative atypicalities are part of the diagnostic criteria for autism spectrum conditions (McPartland et al. 2012), however, autistic people display highly heterogeneous profiles of social and communicative ability (Frazier et al. 2012). Thus, social communication accommodations required by autistic patients are likely to be highly heterogeneous and unique to each patient. General difficulties in verbal and non-verbal communication can act as a barrier to healthcare, or a lack of initiative in reporting medical conditions (Dern and Sappok 2016). By way of example, literal interpretation of language was reported to impair some autistic people’s ability to answer questions about, for instance, quantifying pain (Nicolaidis et al. 2015). These findings suggest that healthcare practitioners should be sensitive to the communication atypicalities associated with autism spectrum conditions. A ‘one size fits all’ approach to patient interaction is unlikely to work for autistic people; hence efforts to create individualised reports tailored to the autistic patient’s needs (Nicolaidis et al. 2016; Hislop et al. 2016). Given autistic patients are more likely to report barriers to healthcare compared to both adults with disabilities (vision/hearing, mobility, learning/remembering, activities of daily living, leaving home alone, working at a job) and non-autistic adults without disabilities (Raymaker et al. 2017), such attention to individual needs may reduce barriers to healthcare (Nicolaidis et al. 2016). For example, negative experiences with healthcare providers (Vogan et al. 2017) and ‘adverse events’ that might affect medical procedures (i.e. measuring vital signs) or increase anxiety about attending healthcare appointments (Saqr et al. 2018).

Importantly, these communication difficulties only represent one aspect of barriers to healthcare. Several studies reported different barriers related to stigma or stereotyping, healthcare providers’ lack of autism knowledge, healthcare providers’ openness to different modes of communication, and trust in professionals. Almost one quarter of autistic people in Vogan et al.’s study reported a fear of stigmatisation or labelling (Vogan et al. 2017); although the sample size was small (N = 40) this does echo the findings of other studies that used qualitative methods. For instance, some autistic people who use alternative communication devices reported that their healthcare professional doubted the autistic patient’s competence (Nicolaidis et al. 2015). Provider attitudes towards autistic patients was discussed in qualitative studies (Dern and Sappok 2016; Nicolaidis et al. 2015) and was a discrete factor in the ‘Barriers to Healthcare Checklist- Long Form’ (Raymaker et al. 2017). For instance, healthcare providers were reported to make assumptions about autistic patients from their initial presentation (Nicolaidis et al. 2015), not take into account rigid thinking (Dern and Sappok 2016), or assume behaviours expressing symptoms (i.e. pain) were behaviours more directly related to autism (Dern and Sappok 2016). These findings are in line with healthcare professionals’ self-reporting. For instance, 79% of adult medicine, 88% of obstetrics/gynaecology, and 70% of mental health professionals in the United States rated their ability to provide healthcare to autistic patients as poor or fair (Zerbo et al. 2015). Furthermore, in the United Kingdom 39.5% of general practitioners (n = 120 out of 304) reported having no formal training in autism, and limited confidence in their ability to offer healthcare to autistic patients (despite having a good knowledge of the key features of autism) (Unigwe et al. 2017). Again, these findings have important implications for healthcare professionals. It is important that healthcare professionals with autistic patients take the time to learn about autism spectrum conditions, and the individual needs of autistic patients. Nicolaidis et al. reported that the AHAT was rated positively by primary care providers (82% rated it as moderately or very useful) (Nicolaidis et al. 2016).

The final ‘global’ barrier reported was sensory sensitivities. Autistic people report elevated sensory processing sensitivity compared to the general population (Crane et al. 2009; Tavassoli et al. 2014). This over-sensitivity is across a range of sensory modalities (sight, taste, touch, sound, smell) and can therefore have a significant impact on daily living (Tavassoli et al. 2014). The studies in this review agree with these findings. Compared to both non-autistic adults without disability (vision/hearing, mobility, learning/remembering, activities of daily living, leaving home alone, working at a job) and adults with disability, autistic people endorsed significantly more sensory items (e.g., lights/smells/sounds make visits uncomfortable or communicate well in healthcare settings) (Raymaker et al. 2017). Lighting (too harsh or bright), crowded waiting rooms (noise, smell, proximity of others), and unpredictable waiting times were also reported to be sources of sensory overload (Nicolaidis et al. 2015). Saqr et al. describe a ‘feedback’ model in which sensory overload reinforces feelings of anxiety, exacerbating communication difficulties when accessing healthcare (Saqr et al. 2018). These findings are important for ensuring the healthcare visit goes well. Healthcare settings should seek to minimise sensory sensitivities where possible. This could include having quiet waiting areas for autistic patients or allowing autistic patients to wait outside the healthcare building and being brought directly into their healthcare appointment.

The intra-person cognitive factors of processing speed, organisation, and memory were not extensively mentioned but have important implications for healthcare settings. The Barriers to Healthcare Checklist includes items about processing speed affecting healthcare discussions and finding it difficult to follow spoken instructions (Raymaker et al. 2017). It is well documented that autistic children struggle to integrate multi-modal sensory information (Marco et al. 2011), or have difficulty processing rapidly presented information (Gepner and Mestre 2002). Evidence suggests that autistic adults are better than the general population at identifying parts of faces, but are poorer at identifying a face holistically (Lahaie et al. 2006). This is important as healthcare appointments with detailed conversations involve processing three streams of information: verbal (listening to the information), visual (looking at the healthcare provider, which some autistic people find aversive, Hurlbutt and Chalmers 2002), and processing (thinking about the information being presented and preparing a verbal response). Thus, an autistic adult may be overwhelmed by the demands of maintaining a social conversation and thinking through the implications of what they are being told. Furthermore, prospective memory (planning ahead for an event, and remembering to carry it out) may be impaired for autistic people. For instance, there are documented problems with timing-based prospective memory (i.e. do event x at time y) in both autistic children (Williams et al. 2013, 2014) and autistic adults (Altgassen et al. 2012). Whilst these are experimental data and may not be entirely valid for healthcare settings, the findings do agree with the anecdotal accounts (Nicolaidis et al. 2015) and measure development (Raymaker et al. 2017) that was identified by this review. Therefore, it is essential that healthcare providers account for these individual differences when interacting with autistic patients. Patients may need more support during healthcare appointments with how health information is disseminated, and more support with making and attending healthcare appointments. The AHAT for primary care providers has information about communication issues for autistic people, and allows the autistic patient to inform their care provider if they have processing speed issues (Nicolaidis et al. 2016). As the AHAT was shown to significantly decrease barriers to healthcare over time (Nicolaidis et al. 2016), accommodating these differences may improve the delivery of healthcare to, and consequently the health of, autistic adults.

Although the age range of participants in the included studies was good (18-64 years) some questions remain about differences in healthcare access needs across the lifespan for autistic people. Currently there is a dearth of research about the characteristics and needs of older autistic adults (Mukaetova-Ladinska et al. 2012; Herrema et al. 2017; Rodgers et al. 2018). Yet, autistic people have many questions about how they will experience the ageing process and the additional support that will be required. For instance, what is the future like for an older autistic person cared for by elderly relatives? (Michael 2016). There is a wealth of literature about the transition experiences of young autistic people (Adreon and Durocher 2007; Hendricks and Wehman 2009; Taylor and Seltzer 2011). Yet, little is known about transitions for older autistic people (Mukaetova-Ladinska et al. 2012; Rodgers et al. 2018). A recent study with autistic adults (mean age 36 years) identified several factors important to autistic people as they contemplate the future (Rodgers et al. 2018). The overarching theme was uncertainty about the future, and specifically, worry about support needs and relationships, living circumstances, and health (Rodgers et al. 2018). These findings agree with a recent survey of 45 individuals who were either autistic individuals or carers of autistic individuals. The survey identified long-term management and awareness of ageing in autism as key research priorities (Mukaetova-Ladinska and Stuart-Hamilton 2015). Taken together, this research highlights the dearth of evidence that could inform healthcare practice for older autistic people. Exploring other literatures could provide a starting point for investigation. For instance, dementia research has looked at promoting a positive hospital experience for patients with cognitive impairment (Prato et al. 2018). Prato and colleagues identified themes that determined the quality of the hospital experience: valuing the person, activities to promote empowerment, and creating a suitable environment to support well-being (Prato et al. 2018).

None of the included studies included samples of autistic people with intellectual disability (ID). It is feasible that some of the excluded studies did in fact have participants with ID who were also on the autism spectrum. However, this review took the position to only include papers that had reported details about samples (or subsamples) of autistic people. Hence, there could be some strategies to improve healthcare access for autistic people with ID that were not included here. For instance, toolkits designed to improve healthcare self-efficacy in ID populations (Lennox et al. 2012) could be tested with both autistic people, and autistic people with co-occurring ID.

A final observation is that overarching healthcare ‘systems’ may impact on healthcare. For instance, Nicolaidis et al. identified system level factors affecting access to healthcare (Nicolaidis et al. 2015). These factors included availability of formal or informal supports and the complexities of the healthcare system (for instance the number of ‘administrative hoops’ patients have to navigate). Indeed, some autistic people reported they felt that they could not navigate the healthcare system without help (Nicolaidis et al. 2015). This is supported by Vogan et al.’s findings whereby just over 50% of participants reported that the steps to seeking help and almost 70% reported that not knowing where to find help were barriers to accessing health (Vogan et al. 2017). These findings do suggest that healthcare professionals could alleviate some difficulties for their autistic patients by providing accurate and concise signposting to healthcare services. Signposting to appropriate services (i.e. towards services the autistic person needs) has been reported to improve access to a range of services (i.e. education and support) and overall well-being in a non-funded community group (Southby and Robinson 2018). This does suggest that directing autistic people to appropriate services does not entail extra costs, and the result could improve effective access to care for autistic people.

Overall, the quality of the six included studies was acceptable to good (60–80%, with one study scoring 20%); however, they addressed different questions with an array of methods and some observations can be made. First, the total sample size for each study tended to be small (i.e. number of participants between 40 and 259), and participants from diverse demographic groups were included (or instance, a range of educational attainment and ethnicity, Nicolaidis et al. 2015, 2016; Raymaker et al. 2017); one study did include participants who were from high-income areas, Vogan et al. 2017). Two studies included people with autism self-diagnosis (Nicolaidis et al. 2015; Raymaker et al. 2017); many adults may be autistic, yet lack a formal diagnosis because current adults were less likely to have received a diagnosis in childhood (Stuart-Hamilton and Morgan 2011; Lewis 2016). It is possible that those with a self-diagnosis face comparable barriers to those with a formal diagnosis, however further research is required to determine this. As such, clinicians should be sensitive, or receptive, to those who self-identify as being on the autism spectrum. Second, three studies used a co-designed approach whereby autistic people were involved in choosing research questions, designing the studies, recruitment, consent process, data collection and analysis (Nicolaidis et al. 2015, 2016; Raymaker et al. 2017). Two of these studies evaluated a new tool. One, to help healthcare providers by creating a statement of individualised healthcare adjustments that could help improve the healthcare appointment for the autistic patient (Nicolaidis et al. 2016) and the other a checklist to measure barriers to healthcare access (Raymaker et al. 2017). Both of these tools were reported to have high face validity and adequate measurement properties. It is reasonable to suggest that this co-design process was a key component in the success of developing these measures (also, these three studies obtained the highest ratings in the quality assessment). Future studies could emulate this approach to make research more accessible for autistic people. Third, there is now a validated checklist that healthcare providers and researchers can use to tackle future research questions about healthcare barriers with the autism population (Raymaker et al. 2017). Fourth, there is not yet a consensus on how to identify and group barriers to healthcare access. For instance, the Barriers to Healthcare Checklist has eight domains (i.e. emotional, executive functioning, support, transportation) (Raymaker et al. 2017). The model proposed by Nicolaidis et al. does encapsulate many, but not all, of these domains (Nicolaidis et al. 2015). This conceptualisation is important to guide future research coherently, to ensure that a full account of barriers to healthcare is given, and to specify hypotheses for studies. As the research into this area is very recent (i.e. all studies are within the past 3 years) more comprehensive studies are likely to follow.

This review had several strengths. First, the agreement between the two authors who screened papers for the 10% sample of included studies (over 99%) suggests that both the selection criteria and initial screening process were valid. Second, the two authors who independently conducted the quality assessment had a very high degree of agreement. These two points suggest the papers included in the final review are likely exhaustive of the available literature and that each paper has been evaluated reliably. The search strategy was developed in an iterative way, so that search terms were refined. Experts in this research area were also included in the development of the search strategy. This study use the QATSDD to assess the quality of the included studies. This was selected a priori as a range of study methodologies was anticipated. This measure allows for a comparison between qualitative, quantitative, and mixed-methodology studies. However, it should be noted that a range of tools for different study types are available and that future studies could use multiple measures as per study type. Considering the limitations of this review, first, our review only identified a small number of papers meeting inclusion criteria. Thus, the barriers and facilitators identified in this review should not be considered definitive. It remains to be seen whether a broader set of barriers will consistently emerge when further research is conducted. Second, the exclusion criteria set for this study mean that our review may not have not have identified some facilitators of healthcare access used in clinical services. Thus, studies about clinicians’ perspectives on healthcare access for autistic people were not included. In the future, studies should explore the perspectives of autistic people, relatives/carers of autistic people, and clinicians about barriers and facilitators to healthcare access.