Services for Adults With Autism Spectrum Disorder: a Systems Perspective

Purpose of Review We review original research about services for adults on the autism spectrum published from January 2013 through December 2018. The main aim is to characterize the topical and methodological aspects of research about services. We review research on services related to employment, living in the community, and social participation. We compare our results with those from a similar review published in 2012 to assess progress and identify where new directions in research about services for adults with autism are needed. Recent Findings We found the evidence base about services for adults on the autism spectrum remains very small and highly variable in aims and methods. There is wide variability in methods used to define sampling frames and recruit participants. Most studies focus on employment. Almost no studies examine the overall ecosystem of services serving autistic adults. Few studies use a conceptual framework for understanding access to, or improvement of, services. Summary The small size of the extant research coupled with inconsistent quality prevents the accumulation of new knowledge in ways that would significantly inform the improvement of systems of care for the growing population of adults on the autism spectrum.


Introduction
Most of a typical life is spent in adulthood, yet only 2% of all autism research funding in the USA focuses on adult issues [1]. An ecological life course perspective highlights the need to improve systems of services and care as a key component of efforts to achieve better health and social outcomes for this population [2, 3••]. A sole research focus on developing individual-level treatments targeting things like social skills and behaviors while ignoring the social determinants of meaningful community participation will not move the needle on population-level outcomes. Evidence-based behavioral and skills interventions do not self-disseminate or self-organize into services and programs that are widely accessible across demographic segments and locales. Echoing the national research agenda on transition for youth with autism [3••], we recommend two overarching priorities for future research on services for adults: (1) identifying how community-and systems-level determinants influence outcomes that are observable at a population-level and (2) increasing meaningful involvement of a broad array of community stakeholders, including autistic advocates, in the improvement of service systems.
Autism spectrum disorder (ASD) is characterized by impaired social communication and interaction coupled with unusual or repetitive interests or behaviors to a degree that significantly impacts daily functioning [4]. The most recent Centers for Disease Control and Prevention estimate of the prevalence of autism among 8-year-old children in the USA is 16.8 per 1000 [5]. A recent survey of US households with children yielded an estimated prevalence of 25.0 per 1000 [6]. Multiplying prevalence by population estimates of the number This article is part of the Topical Collection on Autism Spectrum Disorders of 17-year-olds in the USA, we estimate that between 70,700 and 111,600 youth on the autism spectrum will turn 18 years old this year-equating to roughly 707,000 to 1,116,000 over the next decade.
Autism prevalence estimates have steadily increased in recent decades with growing awareness and screening related to autism. Autism is conceptualized as a heterogeneous spectrum because people vary widely in terms of support needs, strengths, impairment levels in several domains, and the presence of co-occurring health and mental health issues. Many people on the autism spectrum will need special health care and supportive community services throughout adulthood. Services are embedded in institutional, financial, policy, and community contexts-an ecosystem of service provision.
This scoping review characterizes the topical and methodological aspects of research about services for adults (age 18 and older) on the autism spectrum published from 2013 through 2018. We focus on services related to employment, living in the community, and social participation. We compare our results with those from a similar review published in 2012 to assess progress and identify where new directions in research about services for adults with autism are needed [7••].

Literature Review Methods
We sought studies about the provision of services to support outcomes related to employment, education, social engagement, and independence for adults with ASD. We limited the review to studies conducted in English-speaking countries (the USA, Canada, the UK, and Australia) and published between January 1, 2013 and December 31, 2018. We found articles by searching online databases and conducting forward and backward citation searches with relevant articles. Additional articles were nominated for consideration by experts.
We screened articles based on the following inclusion and exclusion criteria. The study had to be original research involving, or pertaining to, individuals on the autism spectrum over the age of 18. If a study also included participants under 18, then the reporting of results had to break out findings for those over 18 or explicitly state the primary focus was on adult services. We excluded reviews, research agenda recommendation studies, and workgroup proceedings. If a study included multiple disability populations, then there had to be findings broken out by type of disability so the autism-specific findings could be clearly discerned. We excluded articles that focused primarily on issues related to caregivers or family members of individuals with ASD. Many articles ambiguously used the term "transition-age youth" that included individuals 18 years of age or older. We excluded articles that focused mainly on services aimed at preparing those under age 18 for adult life. To be included, a primary purpose of a study on transition-age youth had to be describing or evaluating support systems, services or programs for young adults with ASD or evaluating the impact of services on outcomes related to employment, education, and inclusion in the community. We also included articles that focused on topics related to service delivery for adults with ASD such as service provider perspectives, financing of services, and needs assessments. Articles describing adult outcomes that did not include a primary focus on services were excluded. Studies that focused mainly on characterizing impairments or genetic, biological, physiological, and neurological characteristics associated with autism were also excluded from the review. We excluded clinical, behavioral, and medication interventions aimed at mental health, physical symptoms, cognitive functioning, or individual behaviors unless the study was focused on how to scale or implement the intervention at a population level.
This review focuses on research about services for adults age 18 or older. However, we note that it is difficult to define a clear boundary demarcating "adult" versus "transition" research literature. Increasing recognition of a period of "emerging adulthood" has shifted our understanding of the upper age boundary of the transition to adulthood period. Starting in the 1990s, a growing body of research found that the transition into adulthood takes longer in contemporary society and today's youth oftentimes do not achieve independence until their late twenties [8]. The emerging adulthood stage of development is thought to occur between 18 and 29 years [9][10][11]. This reconceptualization is evident within recent federal and state regulations such as the Affordable Care Act that extended coverage to youth under their parents' insurance until age 26. The Annie E Casey Foundation now reports statistics on employment for "young adults" ages 18-29. The Pew Research Center's research on living arrangements of young adults investigates ages 18-34. This shift poses unique challenges in terms of producing a follow up review of adult services as we were faced with the challenge of discerning a clear boundary between the service architecture that is emerging to accommodate the transition period from what is historically referred to as "adult services."

Literature Review Findings
Fifty-two studies met the inclusion criteria (Table 1). To contextualize this number, we searched the PubMed database during the same time period using the keyword "autis*" and found 23,464 peer-reviewed studies. We experimented with other PubMed search parameters and consistently came up with roughly 23,000 published studies. Thus, the studies reviewed here represent less than 1% of scholarly output on autism from 2013 through 2018.
Of the 52 studies, 42 took place in the USA, four in Canada, three in the UK, two in Australia, and one used data from both the USA and the UK. About one-fourth came from four teams. The Nicolaidis and Raymaker group did two

Distribution of Topics and Aims
Over half (n = 27) the studies were about employment. Other focal topics included social participation (n = 15) and postsecondary education support programs for adults with autism (n = 11). Eight studies focused on healthcare services for adults on the autism spectrum. Seven studies included a focus on the cost of services. Five studies were related to behavioral services. Relatively few studies examined financial needs (n = 4) or independent living (n = 2). Two studies were about both employment and postsecondary education. One study focused on describing the characteristics of people with autism who use residential services. Nine studies assessed the impact of interventions that included assistive technology (AT) [13,17,18,25,26,29,35,38,50]. Among the AT studies, six studied the use of AT to improve employment outcomes [13,17,18,25,26,29], one involved the use of a computer-based intervention for college students [35], one involved using AT to support the development of social skills [38], and one involved using AT to teach community shopping skills [50].
Sixteen studies used large secondary datasets to examine questions at a population-or systems-level. Seven were based on data from the US Rehabilitation Services Administration system. Two studies included analysis of data from the National Core Indicators Adult Consumer Survey, and one study utilized data from the National Longitudinal Transition Study-2.

Diverse Approaches to Study Design, Wide Variability in Sample Characteristics
In studies using Rehabilitation Services Administration data, sample sizes ranged from 1696 to 34,501. The number of participants with ASD ranged from one to 87,683 in the remaining studies. Among the studies that did not use large administrative or national survey datasets, sample sizes ranged from one to 225 (mean 41).
Four studies did not involve data about participants on the autism spectrum-two examined physicians' perspectives on providing healthcare services to adults with autism [47,49] and two used secondary data to estimate costs or cost effectiveness of services [61•, 62]. Seven studies solely used qualitative methods [36,44,49,53,55,58,59] and two used mixed methods [23,35]. Three of the 48 studies did not report the distribution of males and females [21,30,63]. The mean percentage of males was 78.5% in studies using large administrative or survey datasets and 65.8% in the remaining studies.
The majority (42 of 48) of studies that included participants with autism characterized the age of the ASD sample in some way. However, age information was incomplete in several instances. Three studies lacked specific details about the ages of ASD participants, other than noting participants were adults or that the services described were intended for adults [15,51,63]. One study used three age categories but the age range or mean age of participants was not reported [15]. In four studies, the upper age limit of the sample was not specified [43,52,55,57].The lower boundary of the sample's age range was not reported in one study [14]. In studies that reported the mean age of ASD samples, the means ranged from 19 to 33 [14, 16••, 24-26, 29, 32, 35, 40, 53, 60, 64].

Conclusion and Considerations for Future Research
Similar to findings presented in a 2012 research review of adult services [7••], we found the evidence base about services for adults on the autism spectrum remains very small and highly variable in aims and methods. Most people with autism have co-occurring health, mental health, and social challenges that change as people age and require accessing services and supports from a wide range of providers. A life course systems perspective emphasizes the complex and evolving dynamics of interaction among culture, history, institutions, organizations, policies, funding, and families that impact individual development, service accessibility, delivery, coordination, and effectiveness [8,[69][70][71][72][73][74][75][76][77][78][79][80][81][82][83][84][85]. None of the reviewed studies adopted a systems perspective or produced new findings with immediate salience for improving complex systems of care and related outcomes. Systems thinking involves a focus of four core domains including the generation and dissemination of new information, network-based approach to facilitate collaboration within and across disciplines and organizations, the use of modeling strategies to guide decision-making processes strategically, and creating systemic change to promote better functioning and internal organization [86]. Systems perspectives have been adopted in public health practice and research as systems thinking facilitates greater understanding of how a complex range of components within health systems are structured, interacting, and functioning [86]. The systems perspective is essential to address complex public health challenges, such as improving services and outcomes for adults on the spectrum, which require coordinated interaction of multiple complex systems [86].
The wide variety of service-related topic areas identified in this review reflects the need for transdisciplinary systems research which could yield a more integrated understanding of how a diverse range of autism service systems function, coordinate, and can be strengthened or improved to promote better outcomes for adults on the autism spectrum. We found no studies rooted explicitly in improvement science-the study of identifying, implementing, evaluating, and disseminating strategies to bring about incremental improvements in system performance [87]. These approaches have a record of improving systems of care in other vulnerable populations and complex care ecosystems [74, 87-89, 90•, 91•, 92]. Most reviewed studies focused on a single intervention, program, or service system.
Future research needs to explicitly adopt frameworks for understanding, and interventions for improving, systems of care. Researchers have applied a systems perspective to understand how factors related to service systems, as well as individual and family-level factors, impact the transition to adult healthcare services for transition-age youth with disabilities. This systems-based approach was useful in identifying strategic systems changes which could promote better outcomes, and similar approaches could be used to improve autism service systems [93]. Future studies need to trace how individuals and families interact with multiple care team members simultaneously, whether care team members coordinate efforts and how to improve alignment of efforts and resource use across organizations.
Approaches to characterizing impairments and severity, sample demographics and socioeconomic position distributions, and co-occurring conditions were highly variable across studies and sometimes missing altogether. This emerging field would benefit from the development of consensus guidelines on study design and reporting standards. In the absence of reporting standards, it is very difficult to understand which subgroups and settings services-related research might generalize to.
The reviewed literature generally did not incorporate a contemporary life course perspective that considers the accumulation of disadvantage, or the impacts of opportunities, across the life span of autistic individuals. The life course health perspective frames health and functioning as interconnected and resulting from the complex interaction of multiple levels of determinants including biological, social, system, and economic contexts [82]. The life course consists of evolving social roles and shifts in the quality and availability of services that a person experiences over the course of their lifetime [82]. Future research could build on these findings by exploring disparities in pathways and trajectories across disability subgroups. Research that examines how gender, race, and social position intersect with disability is also needed. Prior research suggests that children from low-income households [94] and those with ASD [95] are separately at risk for poor outcomes, yet few studies have described the characteristics of children who meet both conditions. Although sociodemographic factors are generally considered to have low mutability [96], scholars have argued that the inclusion of such factors can help researchers identify and control for the effects of systems of social stratification [7••]. Specifically, sociodemographic variables can help to identify inequities in the distributions of services and resources across social groups, as well as help to identify the underlying mechanisms that give rise to such inequities [66].
AT was a focus in nine articles and was primarily used to develop employment-related skills. The emerging body of research evaluating the effectiveness of incorporating AT into services for adults on the autism spectrum suggests that technology can often facilitate the development of skills and promote increased independent functioning. That said, AT research with adults on the autism spectrum is still in an emergent state. States are now required to describe in their vocational rehabilitation state employment plans how a broad range of AT services and devices will be provided to improve vocational outcomes for people with disabilities. Further research is needed to identify the full range of potential applications of AT for adults on the spectrum and their association with positive outcomes for this population.
Healthcare for adults with ASD emerged in this review as a relatively new area of focus. Researchers utilized quantitative and qualitative data from a variety of sources, including physicians [47,49] and adults with autism [43••, 44] to demonstrate that there is a significant need to improve physician training to serve this population. Further research is needed to determine how these findings can be translated into physician training and other efforts that will lead to better health outcomes for adults with ASD. Similar research approaches should be applied to better understand how adults with autism access and experience mental health services.
Most studies included information about sex, but description of gender and non-binary/transgender gender identities was not presented. Future work should be consistent with evolving understanding of gender and research on gender identity in adults with ASD [97]. As research related to gender identity in adults with ASD is emerging, researchers should use best practices for ascertaining gender identity [98] and develop strategies to recruit and include non-binary and transgender people with ASD in research.
Participants' verbal ability was not characterized in the majority of reviewed studies. IQ was reported frequently but may have less of an impact on adult outcomes than other factors like communication ability and adaptive functioning, which were also often not reported. The persistent absence of a consistent and rigorous approach to measuring individual factors that influence outcomes inhibits the ability to determine which subgroups any given research findings may be relevant to.
The articles included in the review involved samples that were predominantly male and white. Greater efforts are needed to include racially and ethnically diverse samples in studies related to adults with ASD. Without this research, disparities cannot be identified or addressed.
The field of research on services for adults on the autism spectrum remains small and generally lacks unifying conceptual frameworks or consistent methodological approaches. These lacks undermine the potential for knowledge to accumulate and be applied to help specific subgroups of people. We recommend two overarching priorities for future research on services for adults: (1) identifying how community-and systems-level determinants influence outcomes and then measuring outcomes at a population-level and (2) increasing meaningful involvement of a broad array of community stakeholders, including autistic advocates, in the improvement of service systems. A study comparing employment-related service providers' evaluations of their work to those of autistic adults and their families revealed that providers tended to have much more positive perceptions regarding the effectiveness of their services than people on the spectrum [23]. This study illustrates how different stakeholder groups can hold vastly different perspectives on the problems and relationships between components within a system. Focusing on a diverse range of stakeholder perspectives, including service providers, service users, and family members, in research on adult autism services could provide unique insight into system changes which could improve outcomes for this population. For example, a study included in the review that focused on healthcare experiences reveals that many autistic adults experience significant barriers to healthcare, highlighting a phenomenon that was previously not well-known in this population [44].

Compliance with Ethics Guidelines
Conflict of Interest The authors declare that they have no conflicts of interest.
Human and Animal Rights and Informed Consent This article does not contain any studies with human or animal subjects performed by any of the authors.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.