Abstract
For autistic individuals, the transition to adulthood is marked by unique challenges including unemployment/underemployment, social isolation, and difficulties living independently. This pilot trial tests feasibility, acceptability, and preliminary clinical impact of an existing transition support program, STEPS (Stepped Transition to Employment and Postsecondary Success program), as delivered in the community. STEPS was piloted with 12 autistic individuals aged 16–35 in a local community agency. Participants provided ratings on several domains including readiness to transition to adulthood, self-efficacy, mechanisms targeted in STEPS (i.e. self-knowledge, self-determination, and self-regulation), and satisfaction with STEPS. Results supported STEPS feasibility, evidenced by high treatment fidelity and moderately strong alliance. Results also suggest high acceptability, evidenced by low attrition, high session attendance, moderate homework completion, and high program satisfaction. Results partially support preliminary clinical impact of STEPS by showing improvement in the primary outcomes (transition readiness and self-efficacy) for some participants. Given the goal of this pilot was to provide an initial examination of feasibility and acceptability to prepare for a larger implementation trial, challenges and strategies for future implementation are discussed. Future research could utilize innovative implementation approaches to test strategies that might promote adoption and long-term sustainability of STEPS when delivered by community providers.
Lay abstract
Autistic adults are less likely to be employed, attend college, and live independently compared to nonautistic peers. Programs for autistic adults that address skills necessary for these domains exist, although they are often not research-tested or readily available in the community. Stepped Transition to Employment and Postsecondary Success program (STEPS) is a research-tested program that seems beneficial and well-liked by autistic adults, although it has not been tested since STEPS was expanded to focus on goals in employment and independent living, in addition to college. To prepare for a larger study looking at the clinical impact and implementation, we ran this pilot trial to provide preliminary data. We ran STEPS with 12 autistic teenagers and adults in a local community organization. We asked them to complete forms before and after on their preparedness to transition to adulthood, confidence in themselves, ability to set goals, problem-solve, and regulate themselves, and their satisfaction with STEPS. We found mixed results on improvement. Several adults reported feeling more prepared to meet their goals in adulthood and more confidence in dealing with stressful situations, whereas others reported no change or feeling less prepared and confident. Some adults also reported being able to set goals, problem-solve challenges, and speak up for themselves more often. Participants and their caregivers liked STEPS, found it beneficial for their goals, and would recommend it to others. The therapist was able to run STEPS sessions, as intended, within the community organization. These findings suggest that STEPS can be helpful for autistic adults. These findings also suggest that STEPS can be delivered in community organizations, although more research is needed to test whether STEPS is as effective and do-able when delivered by clinicians who work in other organizations and who may have less autism training. This pilot trial was to prepare for implementing STEPS in the community with a larger sample of autistic adults, so we also consider challenges and suggest ways to strengthen future implementation.
Introduction
Approximately one in 36 children have autism spectrum disorder (Maenner et al., 2020). For the roughly 70,000 autistic teens who enter adulthood every year (Shattuck et al., 2018), research suggests they are at a higher risk of outcomes like lower quality of life and challenges living independently, compared to age- and cognitive ability-matched non-autistic peers (Biggs & Carter, 2016; Lin & Huang, 2019). They also experience lower rates of employment and postsecondary school enrollment despite having the cognitive ability necessary to succeed in these settings (Elias & White, 2018; Newman et al., 2011; Taylor et al., 2015). In fact, autistic adults have the lowest employment rates compared to peers with other disabilities and autistic adults often work fewer hours and have lower weekly wages than their peers (Nord et al., 2016; Roux et al., 2015; Shattuck et al., 2012). Once autistic adults secure employment, an important secondary challenge becomes maintaining that employment (Taylor et al., 2015). Although research suggests that average-range IQ may be protective against the poorest outcome profile (i.e. unable to lead independent life and needing a high degree of support), higher IQ does not guarantee a good outcome (Mason et al., 2021; Pickles et al., 2020). Furthermore, racial and ethnic minoritization and poverty may exacerbate risk for poor adult outcomes like underemployment (Eilenberg et al., 2019; McCauley et al., 2020; Nord et al., 2016). Given the risk of these poorer outcomes in adulthood for autistic people, supportive programming during the transition to adulthood is needed. We present the results of a small pilot trial of one such program, the Stepped Transition to Employment and Postsecondary Success program (STEPS), to build skills necessary for the transition to adulthood, with the primary goals of understanding whether autistic adults liked and found the program helpful and whether the program could be delivered effectively in a community-based setting to inform future iterations of STEPS.
Transition-related services are generally split into two disconnected segments: services in high school and services after leaving the K-12 school system. These supports involve several entities who lack between-entity communication that families must navigate with little, or no, assistance. School-based planning can be too focused on academic goals and skills, often overlooking skills more pertinent to long-term employment and independent living (Snell-Rood et al., 2020), and these services are less likely to be tailored to meet the specific needs of autistic youth without co-occurring intellectual disability (ID; Taylor et al., 2015; Taylor & Henninger, 2015). Once autistic individuals leave school, they experience a “service cliff” or loss in support that is still needed to facilitate a successful transition to adulthood (Kirby et al., 2020a; Laxman et al., 2019). Caregivers indicated that their autistic adults without ID had tremendous difficulty in qualifying for transition-related services (Anderson & Butt, 2018). Available services are more likely to enroll individuals with more significant or readily apparent disabilities such as ID, psychiatric disorders, or significant medical problems (Lorenc et al., 2018), leading to many autistic adults with average cognitive abilities being overlooked and underserved.
Transition support efforts at both stages appear to not adequately address the core challenges and needs of autistic individuals (Lorenc et al., 2018; Wong et al., 2015). Extant research suggests need areas related to social integration (e.g. skills to communicate effectively, find support, handle conflict), self-determination (SD; e.g. skills for self-advocacy, goal-setting), and self-regulation (SR; e.g. skills for managing daily challenges and stress, executive functioning, and changes in routines; Cai & Richdale, 2016; Elias et al., 2019; Elias & White, 2018; Sosnowy et al., 2018; White et al., 2016). Critical to all three domains is self-knowledge (SK; understanding one's disability and needs). There is a growing literature that suggests these specific mechanisms (i.e., SD, SK, and SR) can be targeted and improved via manualized interventions (Baker-Ericzén et al., 2018; Morán et al., 2021). Given concerns about the validity of using interventions not adapted for autistic people (Spain et al., 2023), it is imperative that interventions developed to improve the transition to adulthood are informed by autistic individuals’ goals and address the unique needs of autistic people.
STEPS (White et al., 2017; 2021) is a cognitive-behavioral intervention designed to improve the adult transition for autistic individuals. STEPS builds skills in SK, SD, and SR, with the goal of enhancing age-appropriate independence in preparation for self-reliance in adulthood (White et al., 2017; 2021). STEPS was originally developed based on needs identified by autistic individuals and their caregivers related to feeling more prepared for the transition to adulthood, and the intervention protocol was continuously informed by stakeholder feedback including expert consultant panels of clinical scientists, educators, autistic individuals, and caregivers who reviewed the intervention and provided feedback (see White et al., 2017 for more details on intervention development). In a randomized controlled trial (RCT), autistic adolescents and young adults randomized to STEPS experienced significant improvements in SK, SD, and SR, and felt more ready to transition to postsecondary education, compared to peers randomized to “transition services as usual” (White et al., 2021). STEPS was highly acceptable to autistic people and their caregivers, and it was delivered as intended when delivered by highly trained research staff in a university-affiliated clinic (White et al., 2021). Following the RCT, stakeholders were interviewed to understand barriers to future STEPS implementation, which informed the current STEPS iteration (manuscript currently under review). STEPS was also expanded to include goals related to educational and/or vocational pursuits, social relationships, or daily living skills at home to reflect the varying needs of autistic emerging adults; thus, this is the first test of a broadened STEPS in the target context, a community-based agency separate from program developers.
The current study
Since STEPS was expanded to include a wider range of goals that autistic adults could focus on while building skills in SK, SD, and SR and STEPS has not yet been tested in a community-based setting, we conducted a small pilot trial in the community to primarily examine feasibility (i.e. whether the intervention was do-able as designed in a community-based setting) and acceptability (i.e. how participants received the intervention, such as whether they liked it, found it helpful, completed homework for it, and wanted to finish it to completion), with a secondary aim to examine preliminary clinical impact (i.e. whether participants gained skills or improved in certain clinical constructs) of STEPS with autistic adolescents and adults. The purpose of this trial was to provide preliminary results of STEPS as a program that can be individualized and applied to a broader range of adult goals and to prepare for a larger-scale randomized, controlled trial of STEPS when delivered in the community, by community providers.
Method
Participants
Participants included 12 autistic adolescents and adults. Inclusion criteria were: (1) being 16 to 35 years old, inclusive, (2) having a prior autism diagnosis, confirmed using the Autism Diagnostic Observation Schedule, Second Edition (ADOS-2; Lord et al., 2012), and (3) being verbally fluent, measured by reports from autistic adults or their caregivers that the adult could have a communicative back-and-forth exchange using complete sentences. Exclusion criteria included: (1) presence of profound ID that would interfere with participation in STEPS, confirmed by the Wechsler Abbreviated Scales of Intelligence, Second Edition (WASI-II; Wechsler, 2011), (2) presence of suicidal intent, psychosis, or aggression warranting immediate care, or (3) currently receiving services redundant with STEPS (e.g. working with an individual therapist to increase daily living skills, participating in a similar program through the state vocational rehabilitation agency to build skills in SD and find a job). Three participants had an official ID diagnosis but were determined verbally fluent enough to participate. No one was excluded due to receiving services redundant to STEPS, and participants were allowed to continue any other services (e.g. emotion-focused therapy, day program services). Participants were recruited from the implementation site (see below; n = 8), and active advertising using the research team's existing research registry and social media (n = 4). This study was approved by the institutional review board, and participants and caregivers provided written consent/assent. See Table 1 for participant demographics.
Participant demographics and baseline data (n = 12).
Note. TRS = Transition Readiness Scale; ISKS = Integrated Self-Knowledge Scale; AIR-SD = American Institutes for Research-Self-Determination Scale; GSES = General Self-Efficacy Scale; EDI = Emotion Dysregulation Inventory.
aOnly race was reported, rather than both race and ethnicity, because all participants identified as not Hispanic/Latino.
Intervention
Before this pilot trial, STEPS was refined into one unified intervention protocol, appropriate for people regardless of age or employment/school status. The focus of this refinement was to consider factors related to implementation (e.g. training materials, improving the accessibility and utility of STEPS materials). The core content largely remained the same, specifically targeting skills in SK, SD, and SR and following a trajectory of exploring needs, goal setting, skill development, and skill practice to build adult autonomy skills and prepare autistic individuals for their unique goals. The protocol had a strong emphasis on individuation in order to meet the client's specific needs and goals, which informed goal setting and skill building. Given STEPS was developed specifically for autistic individuals, certain adaptations to standard cognitive-behavioral interventions were included such as increased use of visuals to deliver information, opportunities for modeling and role play to learn skills, use of a visual agenda for each session, additional time devoted to identifying goals important to the autistic client, content focused on understanding the specific client's autism including their strengths, challenges, and what the identity means to them, and accommodations for sensory, communication, and executive functioning needs. By design, caregiver involvement in STEPS was limited; caregivers were invited to attend the first session to learn about STEPS and participate in the participant's needs assessment if they wished. However, no caregivers joined the first session.
The refined STEPS involved 12 weekly one-on-one sessions delivered over 12–14 weeks. In these weekly sessions, participants and their STEPS therapist discussed and practiced content focused on skills necessary for adult autonomy, within the domains of SK, SD, and SR. Additionally, the therapist completed 4–6 “check-ins” by phone with participants between sessions to discuss homework practices and troubleshoot problems experienced while working toward goals. Online content was available to participants and caregivers on a password-protected site, and participants were encouraged to watch supplementary videos throughout STEPS. This online content consisted of six modules about the transition to adulthood, navigating services, and building SD and self-advocacy. Although caregivers were not involved in weekly STEPS sessions, they were encouraged to utilize the online content. Each participant also participated in an immersion, during which the participant and STEPS therapist met in the community to practice skills in situations that were directly tied to the participant's STEPS goals. These immersions lasted a few hours and involved several structured activities.
The implementation site was a local non-profit organization that provides services to autistic individuals from preschool through adulthood. They served the local community for approximately 20 years and are unaffiliated with the research team. Delivery occurred at the organization's main agency and at a nearby satellite campus that housed the adult day program. Since the organization provides services to autistic individuals year-round, individuals in their programs were informed about the trial and invited to join; however, to maintain generalizability to other populations (e.g. those who are unable to pay for a private service like the adult day program), participants were also recruited outside of the organization's client list. STEPS was delivered for free to all participants in person at the implementation site. A clinical psychology doctoral student served as the therapist who delivered STEPS for all participants, including the individual sessions, immersions, and weekly phone check-ins and was supervised by a clinical psychologist who developed STEPS. All sessions were videotaped for supervision and fidelity purposes.
Measures
Feasibility and acceptability outcomes
Fidelity to the STEPS manual (i.e. ≥ 90% of objectives met) was measured using a session fidelity rating form created prior to the trial. The session fidelity rating form assessed the therapist's adherence to the manual in delivering session objectives. The session fidelity rating forms were completed by the therapist following each session. Therapist-reported fidelity was also confirmed by two independent raters, who collectively co-coded 25% of all sessions, randomly selected, for fidelity to the manual. Training for independent raters consisted of a STEPS overview, including an explanation of objectives, and a practice coding session. During this session, the principal investigator (PI) and independent raters reviewed two session videotapes and rated the therapist's adherence to session objectives. To establish reliability, the PI and independent raters coded randomly selected session videotapes until they achieved exact agreement across three consecutive videos. After reliability was achieved, the independent raters co-coded a randomly selected 25% sample of the total videotapes for each participant (n = 3 videotapes per participant).
Alliance was measured at four timepoints throughout STEPS; thus, each participant had four videotapes coded (total n = 45 videos). Sessions within the four phases were randomly selected. Total alliance strength at each individual intervention phase and an average of the four timepoints were used as benchmarks of feasibility. Good feasibility was operationalized as having an average alliance score of ≥3, which indicates moderate to strong alliance. The coding team consisted of the PI and one independent rater. This independent rater was trained on the VTAS-R-SF and achieved coding reliability with the PI during a previous study (Brewe et al., 2021). The independent rater demonstrated reliability by coding three consecutive videos with ≥80 percent agreement (Cicchetti & Sparrow, 1981). After becoming reliable, the independent rater coded all videotapes (total n = 45 videos), and the PI co-coded a random 20% sample of sessions (n = 9 videos). Since the independent rater was familiar with STEPS, the rater was not masked to session number coded, although precautions were taken to avoid coder bias (e.g. coding tapes in random order, not coding the same participant or timepoint twice in a row). Since the PI also served as the therapist for all participants, only alliance data coded by the independent rater was used for analyses; the PI only assisted in training and reliability coding.
Acceptability. Acceptability included all aspects of consumer response to STEPS (i.e. participant retention, homework completion, and participant and caregiver satisfaction). Acceptability was operationalized as low attrition (i.e. < 20% attrition), high session attendance (i.e. average attendance of ≥80% of scheduled sessions), and high homework completion (i.e. average homework completion of ≥80% for assigned session homework). Additionally, acceptability was operationalized as moderate participant and caregiver satisfaction with STEPS, measured at endpoint via the program satisfaction scale (PSS). The PSS was adapted from the Consumer Satisfaction Survey (McMahon & Forehand, 1983) and assessed impressions of STEPS across several metrics: helpfulness and relevancy of content, satisfaction with participants’ progress, acceptability of format, and likelihood of referring another person to STEPS. Satisfaction with STEPS was measured as having an average PSS score of ≥3, which indicates moderate helpfulness, impact, and acceptability. Participants and their caregivers also provided qualitative feedback regarding their perceived strengths and weaknesses of STEPS.
Clinical outcomes
American Institutes for Research – Self-Determination Scale (AIR-SD). The AIR-SD (Wolman et al., 1994) is a tool that measures capacity and opportunity to be self-determined. The Capacity subscale (12 items) measures one's abilities, perceptions, and knowledge. The Opportunity subscale (12 items) measures how often an individual uses their knowledge and abilities to behave in goal-directed ways at home and school. Since participants’ goals were not solely focused on education, school questions were adapted to include school, work, and other environments participants are regularly in. Self-report included the full 24 items, and caregiver-report included 18 items, which excluded six Capacity items that measured their child's internal perceptions and knowledge. The summation of both subscales produces a Total Score, and higher scores suggest increased SD. The AIR-SD demonstrates adequate reliability and validity in measuring SD for students with and without disabilities (Mithaug et al., 2002; Wolman et al., 1994). Prior research with autistic individuals showed that the measure is internally consistent, and the two-subscale theoretical factor structure is supported (Chou et al., 2017). Internal consistency was high for caregiver- and self-report (α range = 0.91–0.98).
Emotion Dysregulation Inventory (EDI). Caregivers completed the EDI Reactivity (24 items) and EDI Dysphoria (6 items) and participants completed an alpha version of the self-report EDI that consisted of the caregiver-report form reworded to person-first language (Mazefsky et al., 2018). Both forms assess the intensity and frequency of emotion dysregulation in an individual over the past week. Dysphoria is characterized by low positive affect/motivation and a sad or nervous presentation, and reactivity is characterized by rapidly escalating and poorly regulated negative emotional reactions. Higher scores reflect higher emotion dysregulation. The EDI was created for use with autistic individuals and has good construct validity and sensitivity to change (Mazefsky et al., 2018; 2021). Internal consistency was high for caregiver- and self-report (α range = 0.88–0.99). Dysphoria and reactivity T-scores were used as measures of SR.
General Self-Efficacy Scale (GSES). The GSES (Gershon et al., 2013) is a 10-item self-report and informant-report form that measures one's self-efficacy, or belief in their ability to deal with a variety of stressful situations effectively. Higher total scores reflect increased self-efficacy. The GSES has strong internal consistency and convergent validity (Kupst et al., 2015; Salsman et al., 2019). In this study, internal consistency was high for caregiver- and self-report (α range = 0.88–0.97).
Integrative Self-Knowledge Scale (ISKS). The ISKS (Ghorbani et al., 2008) is a 12-item self-report measure of experiential and reflective SK, or understanding one's experience across development in relation to goals and desired outcomes. Higher total scores reflect increased SK. The measure is psychometrically strong, evidenced by strong internal reliability and convergent and discriminant validity (Ghorbani et al., 2008). Internal consistency was good at baseline (α = 0.81) and endpoint (α = 0.84).
Transition Readiness Scale (TRS). The TRS (Elias et al., 2023) is a 30-item caregiver-report and self-report form that measures participants’ readiness to enter college. It was adapted for the current study to measure general readiness for the transition to adulthood, including meeting one's goals in employment and/or education. The TRS measures readiness across three domains: cognitive, emotional, and behavioral. Higher scores reflect more transition readiness. The TRS has strong internal consistency and discriminant validity (White et al., 2021; Elias et al., 2023). Internal consistency was acceptable for caregiver- (α = 0.90 at baseline and endpoint) and self-report (α = 0.79 at baseline and α = 0.88 at endpoint).
Procedure
Individuals completed a brief phone screener to determine eligibility. Eligible participants and their caregivers completed informed consent and the in-person baseline assessment within two weeks of starting STEPS. Participants attended all individual STEPS sessions at the implementation site or their nearby satellite office. All sessions were videotaped for supervision and fidelity purposes. Within two weeks of STEPS completion, participants and their caregivers completed an endpoint assessment. Participants were compensated $30 per assessment, earning up to $60 total.
Community involvement
Members of the autism community (i.e. autistic adults, caregivers, and community providers of autism-focused services) participated in interviews that informed this iteration of STEPS (manuscript under review). Likewise, the implementation site, a community organization serving autistic individuals across the lifespan, was involved in planning for the trial.
Data analysis
Data were analyzed using SPSS 29 (IBM Corp, 2023). Descriptive analyses of all variables were computed to characterize participants (see Table 1). Frequency and descriptive analyses were used to test feasibility and acceptability. Given the small sample size and the preliminary nature of examining clinical impact in this pilot trial, clinical impact was evaluated using reliable change indices (RCIs). RCIs were chosen as the statistical method to measure change because they can account for imprecise measurement and determine the amount of change occurring from STEPS within participants using very small sample sizes (Bauer et al., 2004; Jacobson & Truax, 1991). These analyses evaluated whether participation in STEPS led to significant improvement in transition readiness (TRS), self-efficacy (GSES), and STEPS mechanisms: SK (ISKS), SD (AIR-SD), and SR (EDI). RCIs were calculated by dividing the difference in scores between baseline and endpoint by the standard difference, which accounts for test–retest reliability of the measure being used (Jacobson & Truax, 1991).
RCIs for each outcome were calculated separately. The test–retest reliabilities and SDs used to calculate the standard difference score were obtained from prior research using those outcome measures (Elias et al., 2023; Ghorbani et al., 2008; Mazefsky et al., 2021; Salsman et al., 2019; Wolman et al., 1994). The recommended RCI cutoff is 1.96 to infer clinically reliable and meaningful change (Jacobson & Truax, 1991). Change was examined in both directions to consider significant improvement and worsening. Only data from STEPS completers were used to examine clinical impact. See Figure 1 for RCIs for all measures.
Reliable change indices by participant for all outcome measures.
Results
Feasibility
The results supported feasibility of STEPS implementation, reflected by therapist-rated fidelity to the manual, as well as ratings of fidelity and alliance completed by trained independent raters uninvolved in implementation. Fidelity ratings completed by the STEPS therapist indicated that 93% of STEPS objectives were delivered as intended across 134 sessions. High fidelity to the manual was also confirmed by two independent raters who co-coded approximately 25% of session videotapes (n = 33 videos). Their fidelity ratings indicated that the therapist delivered the stated STEPS objectives as intended for 94.44% of the co-coded sessions.
Alliance ratings completed by an independent rater reflected moderately strong alliance averaged across the 12 STEPS sessions (M = 3.31, SD = 0.35). Looking at alliance per participant, results indicated at least moderate alliance (M > 3) for ten participants (83%), and weak-to-moderate alliance (M < 3) for two participants (17%). A one-way repeated-measures ANOVA showed there were no significant differences in alliance across the four timepoints [F(3, 41) = 0.74, p = 0.415]. Approximately 20% (n = 9) of the total number of videotaped sessions coded by the independent rater were randomly selected and co-coded by the PI to calculate interrater reliability. Interrater agreement was strong; across nine co-coded videos, the independent rater and PI agreed on 91.11% of ratings. Interrater agreement was further examined using intraclass correlation coefficients (ICCs) (2,1); a two-way random effects model was fit which examined absolute agreement among raters (Shrout & Fleiss, 1979). Results indicated adequate interrater agreement (ICC = 0.72).
Acceptability
Results supported acceptability of STEPS, reflected by program attrition, session attendance, homework completion, and program satisfaction ratings. Out of the 12 participants who started STEPS, 11 completed it (92% retention rate). The dropped participant withdrew after the second session, citing limited time to pursue STEPS amid existing extracurricular activities. Across all participants, session length averaged 54.10 min (SD = 8.12; range = 32–77 min). Immersion session length ranged 2–4.5 h. The mean length of the program was 20.14 weeks (SD = 4.15; range = 12–25 weeks). Two participants received one extra session and one participant received two extra sessions, due to concerns outside of STEPS (i.e. family emergency that prompted needing to end the session early, immediate emotional concerns that dominated session time). There were only five no-shows stemming from two participants, indicating a 96.27% attendance rate. Approximately 64.60% of homework assignments were fully completed and an additional 19.50% were partially completed.
Participants and caregivers indicated high acceptability, suggesting they found STEPS helpful, beneficial, acceptable, and relevant, and that they would recommend STEPS (see Table 2 for PSS ratings). They described what they found most useful about STEPS. Participants identified content related to coping with stress, self-advocacy, time management, and cognitive flexibility, whereas caregivers cited content related to goal setting, problem-solving, self-advocacy, and stress management. Program feedback was generally positive, although one participant suggested more review of past skills in session to help with memory and one caregiver suggested providing information about transition-related services following STEPS.
Program satisfaction scale ratings at endpoint.
Clinical impact: intervention outcomes
Considering transition readiness, four participants (36% of completers) and two caregivers (18% of completers) reported significant improvement in transition readiness. No participants or caregivers reported significant worsening. One participant demonstrated reliable improvement across both reporters. Considering self-efficacy, six participants (55% of completers) and three caregivers (27% of completers) reported significant improvement in self-efficacy, whereas one participant and two caregivers reported significant worsening. Of note, two participants demonstrated reliable improvement across both reporters.
Clinical impact: mechanisms
Considering the change in SK, no participants reported significant improvement and one participant reported significant worsening. Caregivers did not report on their child's SK. Considering SD, one participant and two caregivers reported significant improvement, whereas one caregiver reported significant worsening. Considering reactivity as a facet of SR, two participants and three caregivers reported significant improvement in reactivity, whereas one participant and three caregivers reported significant worsening. Of note, one participant demonstrated a significant worsening in reactivity across both reporters. Considering dysphoria as a facet of SR, two participants and three caregivers reported significant improvement in dysphoria, whereas four participants and three caregivers reported significant worsening. Of note, one participant demonstrated significant improvement, and two participants demonstrated significant worsening in dysphoria across both reporters.
Discussion
This study involved a small pilot test of STEPS to examine feasibility (i.e. treatment fidelity and therapeutic alliance) and acceptability (i.e. participant retention, homework completion, and satisfaction) when STEPS is implemented within a community organization serving autistic adults, in order to prepare for a larger RCT of STEPS delivered in the community, by community-based providers. Findings speak to successful implementation of STEPS. Results indicate high fidelity, with the therapist delivering 93% of objectives as intended, rated by both the therapist and two independent raters. Based on independent rater observations, therapeutic alliance was moderate or stronger. This finding represents the good therapeutic bond that was established between participants and the therapist, which is similar to and builds upon existing work exploring development and impact of alliance with autistic participants (Albaum et al., 2020; Brewe et al., 2021). Results on feasibility are also promising considering the study's lack of inclusion/exclusion criteria tied to cognitive ability. Clinical trials are often limited to either individuals with average or higher cognitive abilities or individuals with ID. However, the current study recruited individuals with a range of cognitive abilities, if they were communicative enough to participate in STEPS, to promote generalizability. Findings suggest that STEPS can be implemented with high fidelity when working with individuals with lower cognitive ability. Interestingly, the two participants who demonstrated weak-to-moderate alliance had the lowest IQ. Despite promising results about high fidelity, further testing using a larger sample is critical to explore alliance development with individuals with lower cognitive abilities and whether alliance influences program impact. It may be that individuals with lower IQ need additional adaptations to STEPS, such as additional sessions, more caregiver involvement, or additional time to clarify goals of their participation in STEPS.
Results suggest strong consumer acceptability, evidenced by low participant drop-out, high session attendance, moderate homework completion, and high program satisfaction. Only one participant withdrew, and withdrawal was due to external factors unrelated to STEPS. Although our sample size was fairly small, our retention rate falls within a similar range found in general psychotherapy regardless of diagnosis (Hamilton et al., 2011), intervention trials with autistic people (Conner et al., 2019; Rodgers et al., 2024), and the previous STEPS RCT (White et al., 2021). More than half of the homework assigned was completed, with an additional 20% of assignments partially completed. This suggests participants were somewhat motivated to practice skills outside of session, despite STEPS having limited caregiver involvement to routinely check in with participants about homework. Homework completion rate has not been well-documented for therapy with autistic people, although rates of homework completion in non-autistic samples are around 50%–80% for adults and about 50% for adolescents (Tang & Kreindler, 2017). Given the importance of between-session homework for improved treatment outcomes in cognitive-behavioral interventions (Kazantzis et al., 2016), future intervention efforts should assess ways to increase homework compliance. For example, it may be beneficial to discuss the importance of between-session practice, collaboratively identify homework goals, and address executive function challenges by planning when the client will practice skills and a method for remembering that practice (Kazantzis et al., 2016; Pappagianopoulos et al., 2024). Session attendance was high, with only five no-shows across all participants. Interestingly, one participant accounted for three no-shows and was one of the participants who did not experience reliable improvement in transition readiness for either reporter. It is impossible to directly attribute this poorer attendance to his TRS performance; nonetheless, it highlights an aspect of intervention engagement that may affect intervention impact. It may be that higher no-shows led to longer gaps between sessions, which made consistent progress difficult to maintain. It is also possible that this participant's higher no-shows represent their reduced motivation to participate in STEPS, which led to less buy-in to show up and practice skills, and thus, poorer TRS performance. Despite these no-shows, this participant still had 100% homework completion, suggesting he was relatively engaged. Future iterations of STEPS may benefit from regular check-ins about how engaged participants feel in STEPS and collaborative discussion about how STEPS may benefit them and their progress toward goals for independence. Participants reported high program satisfaction, evidenced by ratings of 4.5 or higher out of 5 across all items on the PSS, and this was corroborated by positive open response feedback on the PSS.
Although not the primary goal of this pilot trial, we also examined initial metrics of efficacy when STEPS was delivered in a community setting. These findings are inconclusive but encouraging. Approximately 36% of participants and 18% of caregivers reported significant improvement in transition readiness. Additionally, 55% of participants and 27% of caregivers reported significant improvement in self-efficacy. These findings suggest some movement in outcomes. This pilot trial did not include a follow-up assessment to gauge intervention effects that occur in the weeks or months following STEPS completion. Therefore, it is impossible to determine if changes in outcomes are sustained, or if there are delayed “sleeper” effects. Regardless, it is important to note that demonstrating clinical impact is not the primary goal of pilot studies; instead, pilot trials serve as valuable tools to optimize interventions, with a specific focus on feasibility and acceptability, and to consider barriers and facilitators that influence future larger scale implementation (Kistin & Silverstein, 2015; Vogel & Draper-Rodi, 2017). Thus, these results, despite the small sample size, suggest some clinical change in the context of pilot feasibility trials and provide a complicated picture of impact.
Analyses examining change in mechanisms were similarly inconclusive. There was a pattern of some improvement, worsening, and stability in scores across mechanisms. Results suggested that at least half of participants experienced no significant change, with more frequent cases of improvement than worsening. Given the context of this pilot trial, continued research on clinical impact using a larger sample is needed. The exception to these findings was change in SR, which suggested a higher percentage of participants experiencing worsening dysphoria. Although the EDI was used to measure SR, emotions were not the focus. It is likely that STEPS may not sufficiently address concerns like emotion dysregulation, beyond one session to identify coping skills for navigating emotions that may come up in pursuit of goals. Given the high co-occurrence of emotional and behavioral disorders (Hollocks et al., 2019), it may be beneficial to include content about mental health resources and discussion about emotional problems as they relate to self-advocacy and help-seeking. It may also be that participation led to increased SK about one's needs and goals for adult autonomy, and this gap analysis led to more feelings of unease and dissatisfaction with current abilities and progress. Although an objective of STEPS is to build an understanding of the stepwise, developmental nature of independence, further iterations should continue to consider ways to improve SK using a strengths-based approach that normalizes gaps in autonomy and is encouraging, rather than disheartening, to participants.
In addition to considering ways to improve acceptability and feasibility as discussed above, an important purpose of this pilot was to identify factors that may impact future implementation as we move STEPS from being delivered in a university research setting to a community-based setting. STEPS delivery highlighted several lessons for future implementation. First, STEPS was broadened to focus on several adult goals, not just postsecondary education, which allowed the STEPS therapist to individualize STEPS content to a variety of pursuits based on that participant's individual needs and goals. Thus, findings support feasibility and acceptability of a broadened STEPS and was a better fit with the implementation site, which serves a diverse clientele with varying cognitive abilities and needs. It also meant that participants were more readily able to collaborate with the therapist on goals that were meaningful for them across a range of domains (e.g. a participant desiring to enroll at a local community college and become more routine in her self-care at home). This pilot demonstrated the site's need for manualized transition support, given many participants were easily recruited from their existing clientele. However, limited physical space that was private, quiet, and appropriate for clients’ developmental levels was a barrier for delivery. The pilot also highlighted the importance of considering clients’ maturity and SD, particularly within agencies that may include both child- and adult-focused services. For example, feedback suggested some clients and caregivers desired more involvement of a close support person during program delivery to aid in skills practice and generalization. The team may need to re-work how STEPS fits with agencies serving autistic adolescents to respect the balance of autonomy and caregiver support. These implementation takeaways are important considerations for future STEPS delivery, particularly since pilot studies are not well suited for testing clinical impact or efficacy and rather serve as an opportunity to prepare for a larger trial (Leon et al., 2011; Vogel & Draper-Rodi, 2017).
Considering STEPS's broadened focus, future implementation research should consider ways to measure participants’ actionable progress towards their identified goals, in addition to change in constructs like SK, SD, and SR. It may be that the measurement tools used were insufficient to capture important changes in participant-identified goals, despite participants making meaningful progress or reaching those goals. For example, a participant achieving her goals in feeling more comfortable with cooking at home and enrolling in college courses are meaningful changes, but they may not have resulted in significant changes on the TRS. Given the range of desired goals and the degree of co-occurrence of emotional and executive functioning challenges, it may be that STEPS is more beneficial as a generalized service that leads into targeted intervention (e.g. an emotion-based therapy or executive functioning intervention). Connecting STEPS with a targeted service, such as job coaching, may also increase clinical impact and achieve participants’ functional goals.
Limitations
Participants were primarily White with no ethnic diversity, and the sample size was small, which limits generalizability. It will be important to examine how acceptability may change with a diversified sample. It should also be noted, given the small sample size, that this pilot trial reflects efforts to understand STEPS's impact when delivered in the community and to inform future implementation efforts, rather than to add to theory. Although STEPS was designed based on feedback from autistic people and other stakeholders and interviews with autistic adults and their caregivers conducted prior to the current feasibility trial further informed the intervention, autistic people were not involved in the design of this feasibility study. Methodologically, the TRS as the primary outcome reflects thoughtful consideration of the extant literature compared to goals that could be addressed by STEPS; however, it has not been fully validated as an outcome instrument. The extent to which the TRS may accurately capture change is important to interpretation of these findings, and continued research with this measure is needed to confidently determine clinical impact. The pattern of observed change across participants was not uniform, making it nearly impossible to draw definitive conclusions about clinical impact. Although clinical impact is not the primary goal of pilot studies (Kistin & Silverstein, 2015; Leon et al., 2011), future research should continue to investigate clinical impact, considering both psychological outcomes (e.g. changes in readiness, SK, SD, and SR) and functional outcomes. For example, qualitative interviews with participants who were considered ‘treatment responders’ versus not may provide rich data on participants’ perspectives of what led to positive or negative change for themselves in certain clinical outcomes or whether certain activities or aspects of STEPS affected change. The implementation setting and provider may limit generalizability in future iterations of implementation. The results may not generalize to agencies with different funding structures, clientele, and services. Likewise, the therapist was a clinical psychology doctoral student trained in manualized evidence-based interventions for autistic individuals and received weekly supervision from the STEPS developer; thus, generalizability to providers with less experience working with autistic individuals or delivering manualized interventions may be limited.
Conclusion
The transition to adulthood is a period of heightened risk for autistic people and many feel less prepared for this new stage of adult autonomy (Kirby et al., 2020b; Newman et al., 2011). There is limited availability of research-supported interventions that support the transition to adulthood, which is a problem when considering the quickly growing population of autistic adults needing transition support. This pilot trial of STEPS demonstrated strong feasibility and acceptability, although clinical impact was less clear with the preliminary nature of these results and small sample size. These results suggest that STEPS may be a suitable choice for programming in community-based organizations and can be individualized to a variety of unique needs and adult autonomy goals.
Future research should utilize innovative approaches (e.g. hybrid effectiveness-implementation trials; Curran et al., 2012) to test STEPS when delivered in organizations by their own providers. Given the inconclusive results on clinical impact, this research could examine change using a larger sample, and ideally, a follow-up period to examine long-term impact of STEPS. Outcome measures should consider both societal and autistic self-advocates’ definitions of positive outcomes (Autistic Self Advocacy Network, 2019; Taylor, 2017), to fully understand the extent of clinical impact of STEPS and to establish data on “successful outcomes” that can be used to market STEPS to community agencies.
Footnotes
Acknowledgements
This data was collected as part of a doctoral dissertation. As such, we would like to thank the first author's doctoral dissertation committee for their feedback on the project and the families who participated in this research.
Author contributions
A.M.B.: conceptualization, methodology, formal analysis, investigation, writing—original draft, writing—reviewing and editing, visualization, project administration, and funding acquisition. S.W.W.: conceptualization, methodology, supervision, writing—reviewing and editing, and funding acquisition.
Data availability
The data that support the findings of this study are available from the authors upon request.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Ethical approval
All research presented in this manuscript was approved by the University of Alabama Institutional Review Board and all participants and their caregivers provided written consent and/or assent.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: We received funding from the College Academy of Research, Scholarship, and Creative Activity at The University of Alabama (PI White).
