The Emotional Support Plan: Feasibility trials of a brief, telehealth-based mobile intervention to support coping for autistic adults

Abstract

Difficulties regulating emotions during periods of distress may contribute to the high rates of co-occurring depression and anxiety in autistic adults. The Emotional Support Plan (ESP) is a brief intervention designed to support autistic adults to use positive coping skills during periods of distress. Thirty-six autistic adults participated in studies to assess the acceptability of the ESP intervention to cope with stressors during the COVID-19 pandemic and postsecondary education and feasibility of the study design elements to inform future trials. Most participants reported using strategies from their ESP; 86–89% reported the intervention had a positive impact on them and 67–71% would recommend it to another person. Completion of weekly monitoring and outcome assessments were high; ecological momentary assessment was more variable. The current results provide preliminary support for the acceptability of the ESP intervention. Important insights were also gained to inform feasibility of the design to assess its potential efficacy in future studies. While further research is clearly needed, the brief nature of the ESP may provide benefit as a starting point for those who may be proactively seeking support to cope with anticipated stressors or those who would benefit from guidance to promote emotion regulation during stressful life events.

Lay abstract

Autistic adults may have difficulty coping during stressful periods, which could make them more vulnerable to depression and anxiety. We designed the Emotional Support Plan (ESP) to help autistic people find ways to cope in stressful situations. Thirty-six autistic adults created an ESP and answered questions about their opinions of the ESP. Most autistic adults found the ESP to have a positive impact on them and many would recommend the ESP to another person. Feedback from autistic adults suggested ways that we might test the ESP in future studies. Overall, autistic adults in this study found the ESP to be useful and a worthwhile intervention to study more in the future. While more research is clearly needed, we hope that the brief nature of the ESP will make it helpful for autistic people who are trying to handle negative feelings during stressful life events.

Autistic adults show increased risk for co-occurring anxiety and depression (e.g., Hollocks et al., 2019) and are three times more likely than non-autistic adults to visit the emergency department for psychiatric reasons, including suicidal ideation (Vohra et al., 2016). Research indicates that many autistic adults struggle with emotion regulation and may rely on maladaptive coping strategies (e.g., self-injury, Goldfarb et al., 2021), making it more challenging to adapt to daily stressors (Cai et al., 2018; Mazefsky, 2015). This leaves autistic adults particularly vulnerable to mental health symptoms during stressful periods, such as the transition to postsecondary education (PSE) as well as the COVID-19 pandemic (Gunin et al., 2020; denHouting 2020). These statistics demonstrate an urgent need for studies focused on developing and testing mental health interventions designed specifically for autistic people who are coping with mental health symptoms and stressors.

There is a limited body of research on interventions for autistic individuals (Brugha et al., 2015; Maddox et al., 2021, White et al., 2018). More specifically, to our knowledge, there are no empirically-validated, brief interventions focused on supporting autistic adults in coping with acute emotional distress that may be present during stressful periods, a need that has become increasingly pressing since the onset of the COVID-19 pandemic. Reports highlighting the impact of the pandemic on mental health and well-being of autistic adults (Adams et al., 2021; Bal et al., 2021; Hedley et al., 2021) underscore the need for brief, problem-focused interventions to support autistic adults through periods of uncertainty and constant change. This may be particularly relevant as new variants continue to pose risk in the absence of widespread mandates and the expiration of the COVID-19 public health emergency in the United States.

Beyond COVID-19, life transitions, such as the transition into PSE, pose numerous potential stressors, particularly for autistic individuals (Mazefsky, 2015; Vanbergeijk, Klin, & Volkmar, 2008). Changes associated with PSE, such as increased workload, new social dynamics, and a potentially different location all are likely to create increased stress and more frequent instances that require the use of healthy coping skills (Elias & White, 2018; Gunin et al. 2021; Jackson, Hart, Brown, & Volkmar, 2018). Considering the notable service cliff experienced by autistic adults, wherein after leaving high school these individuals receive substantially less support (Shattuck et al., 2012), autistic adults may benefit from having a brief, flexible tool that can be tailored to promote coping as needed.

Without interventions to directly address emotion regulation when faced with stressors, autistic adults will continue to be a high-risk population for adverse mental health outcomes. To begin to fill this gap, the Emotional Support Plan (ESP), was designed as a brief intervention to support autistic adults in the use of positive coping skills during periods of distress to regulate negative emotions. This brief telehealth and mobile intervention package was modeled after the Stanley-Brown Safety Plan (Stanley & Brown, 2012), a highly studied suicide risk intervention. Although not all autistic adults have suicidal thoughts during “meltdowns” or other negative emotional experiences, the pragmatic, problem-focused nature of the Safety Plan was attractive for several reasons. First, this approach was chosen for its brevity and concrete nature. Like the Safety Plan, the ESP provides a structure that is used to pre-emptively identify stressors and outline specific strategies that could be used to cope in the moment, lending consideration to strategies that would be feasible in the context of the identified stressor(s). Also, like the Safety Plan, the ESP is written out and can be made easily accessible via a web link through a smartphone or tablet, allowing it to provide support or intervention in the moment when it is needed most. This seemed particularly important, as being able to pull up the plan on one’s phone may help reduce the need for reliance on memory and other executive functions that may be impeded during periods of high stress (Phung et al., 2021; Lewis & Stevens, 2023). Importantly, the ESP is intended for use in any stressful period, not solely at times of suicidal risk, as these elements are comparable to those described by autistic people when feeling overwhelmed or having “meltdowns” (Lewis & Stevens, 2023). While risk planning may be viewed as a secondary intervention approach (Horowitz et al., 2020), the ESP may have utility as either a primary/preventative approach (e.g., administered to all individuals before an anticipated stressful transition, such as beginning a new college semester) or secondary-intervention approach (i.e., delivered during a high stress period, such as the COVID-19 pandemic).

Mirroring the structure of the Stanley-Brown Safety Plan, the ESP has six main sections (Table S1):

• 1Warning signs. This section focused on helping individuals identify signs that they are beginning to feel stressed or upset, such as shoulders tensing or fists clenching. For people who were not able to identify physiological experiences, this section could also be used to identify situations that are anticipated to bring on stress, such as interactions with a professor or navigating a busy grocery store.
• 2Things I can do to manage my emotions. This section aimed to help the participant identify coping strategies. Participants were presented a “menu” of seven broad categories of coping techniques, plus an eighth “personalize it” category to encourage unique interpretations of coping strategies (Fig.S1). If a participant chose to further explore any area, a brief list of strategies were shown and additional explanation or examples (YouTube videos) were provided. For example, under Mindfulness Exercises, examples included mindful breathing, progressive muscle relaxation, etc.
• 3Positive supports. Here the focus was on people that can serve as contacts to provide support. These may be trusted individuals with whom the participant can talk through the distressing situation, or simply contacts that may serve as a positive distraction in the moment.
• 4Reasons for controlling my emotions. This section asks individuals to consider possible consequences or outcomes associated with losing control due to feelings of stress. Examples may include getting fired from a job or asked to leave the class.
• 5Ways that I can make my environment less stressful. The focus of this section was to identify changes to the environment to reduce stress. This could include specific sensory stimuli, such as taking a cold shower or lighting a candle to comfort oneself, changing locations (e.g., leaving a crowded room) or putting on noise-cancelling headphones.
• 6Professional or other help. Here, professional sources of support and the associated contact information are noted. Examples include a therapist or community/national resources (e.g., a crisis text line).

Two final sections are drawn from the Stanley-Brown Safety Plan and are optional for individuals with a history of suicide risk to promote safety during those specific circumstances.

• 7Reasons for living. Individuals are asked to think about reasons to live and things that are important to them. These may include friends and life goals they want to achieve.
• 8Ways that I will make my environment safer. The focus is on efforts to identify and remove the individual’s access to lethal or dangerous items.

Consistent with the nature of the suicide safety plan, the ESP was designed to be flexible enough to modify for use in a variety of different situations. It may be useful as a facet of primary prevention efforts to promote emotional well-being and prepare individuals for coping during periods or specific situations that are likely to be stressful, or as a secondary intervention strategy to reduce the effects of acute distress. Recognizing that many autistic adults are seeking interventions that could be self-initiated or managed (Benevides et al., 2020), both clinician- and self-guided versions of the ESP were designed and feasibility-tested. The aim of the current studies are to describe the acceptability and feasibility of the ESP intervention and study design for future studies evaluating its efficacy; this was seen as a necessary first step considering the novel nature of the ESP. Data presented here are from two separate studies designed to assess the ESP’s acceptability for coping with distress in different contexts (i.e., during the pandemic and during PSE); preliminary clinical outcomes are described elsewhere (Mournet et al., under review).

Method

Participants

Participants were recruited to ESP feasibility studies registered with ClinicalTrials.gov: 1) for coping with stress during the COVID-19 pandemic (COVID; NCT04460677) and 2) to support postsecondary students (PSE; NCT04926090). Structurally, these studies were quite similar and thus we combine them in this present manuscript; differences are highlighted below in Procedures. A total of 36 participants were eligible and consented to participate in the study (see Fig.1 CONSORT). Thirty (COVID=21, PSE=9) participants were guided through the ESP intervention by a clinician (“clinician-guided” = COVID-C; PSE-C). The remaining six were PSE students who completed a self-guided version of the intervention (“self-guided group” = PSE-S). All students attending the university where the study took place were assigned to receive the PSE-C and students from other universities were assigned to the PSE-S. Eligibility criteria across the studies included having a self-reported professional autism spectrum disorder diagnosis (including Autistic Disorder; Asperger’s and PDD-NOS), being able to speak English, and having access to a smartphone device. For individuals in the COVID study, participants had to be 18 years of age or older, have a verbal intelligent quotient (VIQ) of 70 or higher or past/current enrollment in postsecondary education and living in the US. For the PSE study, reading comprehension skills of 6^th grade or higher and enrollment in a New Jersey postsecondary institution was required; due to difficulties recruiting locally, eligibility was expanded to include students at any US-based institution in October 2021. In both studies, independent adults provided consent to participation and legally dependent adults could be enrolled with their guardian’s consent. In the PSE study, 17-year olds could be enrolled with their parent’s consent. Participant characteristics are shown in Table 1.

Figure 1.

CONSORT Diagram

Note: ASD=Autism spectrum disorders; COVID-C=COVID-19 clinician-guided group; EMA=ecological momentary assessment; PSE-C=Postsecondary education clinician-guided group; PSE-S= Postsecondary education self-guided group

Table 1.

Demographic characteristics

	COVID-C (N=21) N (%) / M (SD)	PSE-C (N=9) N (%) / M (SD)	PSE-S (N=6) N (%) / M (SD)	Statistical differences
Age	25.1 (7.1)	19.5 (6.5)	21.5 (3.1)	F=2.80; p=.076
Gender				χ2=4.65; p=.325
Men	11 (52.4%)	6 (66.7%)	1 (16.7%)
Non-binary / Genderqueer	2 (9.5%)	1 (11.1%)	2 (33.3%)
Women	8 (38.4%)	2 (22.2%)	3 (50%)
Race				χ2=5.85; p=.828
Asian	1 (4.8%)	1 (11.1%)	0 (0%)
Black	1 (4.8%)	0 (0%)	1 (16.7%)
Native Hawaiian / Pacific Islander	1 (4.8%)	0 (0%)	0 (0%)
Other	2 (9.5%)	0 (0%)	0 (0%)
White	15 (71.4%)	8 (88.9%)	5 (83.3%)
Decline to report	1 (4.8%)	0 (0%)	0 (0%)
Ethnicity				χ2=1.6; p=.449
Hispanic/Latinx	3 (14.3%)	1 (11.1%)	1 (16.7%)
Education level				χ2=22.03; p=.02
Some high school	2 (9.5%)	0 (0%)	0 (0%)
High school	5 (23.8%)	0 (0%)	0 (0%)
Enrolled: Undergrad	6 (28.6%)	8 (88.9%)	4 (66.7%)
Bachelor degree	5 (23.8%)	0 (0%)	0 (0%)
Enrolled: Graduate	0 (0%)	1 (11.1%)	2 (33.3%)
Grad/Prof degree	3 (14.3%)	0 (0%)	0 (0%)
Age of diagnosis	14.0 (12.8)	8.9 (6.9)	12.1 (10.0)	F=0.72; p=.496
Household income				χ2=8.44; p=.392
< $35,000	2 (9.5%)	3 (33.3%)	1 (16.7%)
$36,000-$100,000	1 (4.8%)	0 (0%)	1 (16.7%)
Over $101,000	3 (14.3%)	0 (0%)	0 (0%)
Declined	9 (42.9%)	4 (44.4%)	4 (66.7%)
Did not respond	6 (28.6%)	2 (22.2%)	0 (0%)
Receiving mental health servicesa				χ2=8.4; p=.077
Yes	10 (47.6%)	6 (66.7%)	4 (66.7%)
No	0 (0%)	3 (33.3%)	2 (33.3%)
Unknown	11 (52.4%)	0 (0%)	0 (0%)

Note: COVID-C=COVID-19 clinician-guided group; PSE-C=Postsecondary education clinician-guided group; PSE-S= Postsecondary education self-guided group

^aIncludes counseling/therapy and/or psychiatry services

Procedures

Questionnaires were collected via REDCap. All visits and procedures were implemented via telehealth by a clinician who was a psychology doctoral student or a trained research assistant; study procedures were supervised by a licensed clinical psychologist. Beyond differences in target samples, the primary differences between the COVID-19 and PSE studies were the length of time participants spent in the monitoring phase and how they were assigned to the EMA. These differences are highlighted in Figure 2 and described in detail in the respective sections below.

Figure 2.

Overview of Study Design

Baseline Assessment.

All participants completed a battery of questionnaires about their emotional functioning and a telehealth baseline eligibility assessment which included review of questionnaires and previous mental health diagnoses and identification of a collateral contact (i.e., someone that could facilitate contact if the participant could not be reached).

Intervention.

Clinician-guided ESP.

Participants participated in 1–2 intervention sessions to create their ESP. The first part of ESP creation was to identify coping strategies. Participants were asked about the types of strategies they used to cope with stress. They were shown the coping strategy menu (Fig.S1) and given the option to explore each category with the clinician. In the second part of ESP creation, clinicians guided participants through creating each part of their ESP (Table S1), documenting the information on a Qualtrics survey. If participants indicated risk for suicidal behaviors, they were prompted to include specific resources in Section 6 (e.g., suicide hotline) and completed sections 7 and 8. At the end of the session, clinicians showed participants how to save the Qualtrics survey on the home screen of their device for easy access. Following ESP creation, participants completed weekly questionnaires about their clinical symptoms and activities and met with clinicians for brief (10–20 minute) telehealth monitoring sessions.

Self-guided ESP.

Participants were sent a link with step-by-step instructions to create their ESP. This included a series of videos (Introduction/Overview of the ESP), Brainstorming Coping Strategies, ESP Creation Guide, two examples of situations/strategies, and How to Install the ESP on your phone). These videos were intended to provide the same information typically provided by clinicians in the ESP-C, such as explaining each ESP section, describing and providing examples of the different types of coping strategies and supports one might choose to put on their ESP. To ensure comparability in information provided, videos were designed and created by autistic students and study clinicians. All videos were accompanied by a transcript of the audio and PDF slides shown during the video. Two worksheets were also provided to support participants in documenting their thoughts while watching the videos and to inform consideration of where different strategies may be useful (i.e., public, private, or both). They were asked to upload their final ESP to the website and complete a questionnaire about their experience creating the ESP.

Monitoring Period.

As shown in Figure 2, monitoring procedures varied across studies. The COVID-C monitoring period was seven weeks and consisted of weekly questionnaires about clinical symptoms and activities and brief monitoring visits, during which participants reported on their weekly activities, how stressful their week was (0-not at all to 5-very much) and ESP use. The PSE monitoring period was designed to span the entire 15-week semester, with weekly questionnaires for the first six weeks, and then biweekly questionnaires for the remaining four weeks of the semester. The ESP-C participants also completed monitoring visits, whereas ESP-S only completed questionnaires.

Assessments.

Ecological Momentary Assessment (EMA).

Some participants (COVID-C=4; PSE-C=4; PSE-S=6) were non-randomly assigned to receive four EMA prompts to report on daily experiences for their monitoring period (COVID-C: seven weeks; PSE: first six weeks). Alerts were sent via MetricWire at 9:00 AM, 8:00 PM and two other random times throughout the day with a minimum of 2 hours between each alert. EMA periods were longer than the 2–4 week periods used in other research, we prioritized assessing feasibility of EMA for monitoring purposes in order to gain insights into whether daily prompts would promote ESP use. EMA prompts included information on ESP use since the most recent survey, affective states, and quality of sleep the night prior. Participants were also asked about suicidal thoughts and behaviors each evening. Automated alerts notified the study team if participants endorsed suicidal thoughts or behaviors or elevated negative moods. The on-call team monitored risk alerts between 9:00AM-9:00PM, seven days per week. If an alert was received, the study team reviewed the data and, if needed, initiated contact to conduct a risk assessment. Due to budget constraints and intensity of staffing required for monitoring EMA responses for risk and follow-up as needed, EMA assignment was based on study team availability. This resulted in the last four participants of the COVID-19 study being assigned to EMA. In the PSE study, every other PSE-C participant and all participants in the PSE-S were assigned to EMA; the latter decision was influenced by a desire to monitor all PSE-S participants more closely because they were not interacting with clinicians on a weekly basis.

Outcome Assessment.

At the end of the monitoring period, all participants completed a set of questionnaires (including a clinical battery, described in Mournet et al., under review) and a satisfaction survey (detailed below in Measures & Data Analysis section). A follow-up telehealth interview was conducted to obtain their previous week’s stress and ESP use and responses to the satisfaction survey. Participants were also reminded that they would have continued access to their ESP and were encouraged to use it in the future as needed.

Follow-up Assessment.

Participants of the COVID-C arm of the study also engaged in a follow-up session one month after their outcome session. At this point participants repeated questionnaires from the Outcome assessment.

Risk Assessment.

If suicide risk was endorsed on any questionnaires throughout the study period, the study team followed up on these responses at their next scheduled visit within 24-hours or less. If they did not have a visit scheduled, or their visit was not in the next day, the study team initiated contact in order to conduct a risk assessment.

Acceptability Measures

Satisfaction Survey.

Participants were asked to rate: how helpful the intervention sessions and individual ESP sections were; acceptability of the telehealth format and weekly monitoring; content relevancy; level of benefit and overall satisfaction with the ESP and whether they would recommend the intervention to others. EMA participants were asked to rate acceptability of daily prompts (i.e., receiving notifications four times per day, timing of the prompts, and number of questions asked/time it took to complete each prompt), whether the daily prompts benefited them, and if receiving the prompts helped to remind them that the ESP was available. All items were rated on a 1-(Strongly Disagree)-to-5 (Strongly Agree) scale. Participants were given space to elaborate on their answers. During the outcome visit, clinicians asked open-ended follow-up questions to obtain clarification or elaboration.

Self-guided ESP Creation Survey.

Participants reported their overall experience creating their ESP (1-Very Difficult-to-5-Very Easy), how long it took them to create it, which materials they used and, if used, how helpful (Not at all, Somewhat, Very Helpful) that material was and if anyone had helped them create their ESP.

Data Analysis

Feasibility was assessed through recruitment and enrollment during the study period and survey and visit completion rates. Logistic regression was used to investigate whether higher weekly self-reported stress predicted ESP use. Acceptability was assessed through average ratings of satisfaction survey items and qualitative responses.

Qualitative content analysis was used to derive themes from transcriptions of the participant’s responses to open-ended questions during outcome sessions (Henninger & Taylor, 2014; Hsieh & Shannon, 2005). T.G. and A.M. independently used inductive coding to identify themes from a subset of participants. A codebook with examples was co-developed and refined with V.H.B for clarity. T.G. and A.M. then independently coded three full interview transcripts, discussed, and established consensus on codes. Based on these initial codings, it was decided that all transcripts would be double coded independently and discussed; consensus codes are reported. Inter-rater reliability was 84.8%.

As the focus of this paper was feasibility and acceptability, baseline clinical characteristics and clinical outcomes are described elsewhere (Mournet et al., under review).

Community Involvement

Two autistic graduate students were members of the study team and coauthors of this manuscript. They participated in co-development of the ESP-S videos and other materials and with the monitoring visits for the PSE study. One student also participated in qualitative coding.

Results: COVID-C Study

Feasibility [Recruitment, monitoring]

Recruitment.

The study aimed to recruit 20 autistic adults between August 2020 and July 2021. In the initial recruitment phase (August-December 2020), study flyers were distributed via social media, autism recruitment websites and professional networks. From these flyers, 48 interest forms were completed. Of these, 18 did not respond to contact for screening, 6 were deemed ineligible, one declined participation and the remaining 22 consented. Study enrollment was paused January-June 2021 due to limited staffing during the COVID pandemic. In the second recruitment phase (July-December 2021), 28 additional interest forms were completed in response to flyers circulating online. Of these, four participants were consented; the remaining 24 were not screened or consented due to ongoing staffing shortage and were notified that enrollment had reached capacity for that period.

Screening.

Of the 26 participants consented, two withdrew due to personal circumstances and one declined to provide a collateral contact. At baseline, one participant was not eligible and the other was referred to higher intensity clinical services. The remaining 21 created their ESP.

Monitoring and Outcome.

One participant withdrew after week 1 due to personal circumstances. For the remaining 20, monitoring visit (88%) and questionnaire (92%) completion was high and took place across 6.1 weeks (SD=0.42). Outcome completion was 95% for questionnaire and 91% for visits (M=1.19, SD=0.37 weeks after last monitoring visit). Eighteen participants (85.7%) completed follow-up questionnaires (M=4.75, SD=2.03 weeks after outcome).

Acceptability

Overall satisfaction with the ESP-C intervention (i.e., accessibility, content) was somewhat mixed (M=3.71, SD=1.01; Fig.3a). Qualitative analyses below provide additional context on areas in which participants noted desired improvements. Most participants indicated that the ESP had a positive impact/benefit (M=4.00, SD=0.84), content was relevant (M=4.05, SD=0.80) and the telehealth format (M=4.29, SD=0.56) and weekly monitoring were acceptable (M=4.33, SD=0.58). Visits to create the ESP were generally rated to be helpful; when asked about specific ESP sections, things I can do to manage my emotions were rated as helpful by the most people (Fig.4a). Most participants would recommend the ESP-C intervention to another person (Fig.3a).

Figure 3.

Satisfaction and acceptability ratings for the ESP by study.

Figure 4.

Satisfaction with individual components of the ESP and EMA by study.

Weekly Stress and ESP Use

Stress ratings ranged from 0.5–3.33, with a sample mean of 2.06 (SD=0.98). Participants reported using their ESP 4.55 weeks (SD=2.89) of 8 weeks queried (never: n=2; 1–3 weeks: n=6; 4–5 weeks: n=5; 7–8 weeks: n=7). Eight participants spontaneously reported that they did not actually refer to the physical ESP on their device, but recalled strategies from memory. The association between weekly stress level and ESP-use was not significant (OR=1.59, 95%CI=0.92–2.76, p=.099)

Results: PSE Study

Feasibility [Recruitment, monitoring]

PSE-C.

Recruitment

The study aimed to recruit 30 autistic PSE students for PSE-C between July and October 2021. Information about the study was provided to 26 students enrolled in an academic support program for autistic undergraduates (July-September 2021). Seven completed screening forms, three reported being too busy or overwhelmed to participate; one student said they may be interested later but did not respond to follow-up; the remaining 15 declined participation (reasons not indicated, though conversations suggested students were busy with family vacations before the semester or the start of classes). Four additional students responded to posted flyers and notices disseminated to other on-campus departments (September-October 2021).

Screening.

Of 11 students who completed interest forms, one was deemed ineligible, and one did not respond to follow-up. The remaining nine participants consented and completed the Baseline and Intervention visits.

Monitoring and Outcome.

Completion rates for the 14-week monitoring protocol were high (91% for monitoring visits and 93% for questionnaires), occurring over an average of 13.28 weeks (SD=0.66). All participants completed outcome visits and questionnaires. Outcome visit occurred 1.63 weeks (SD=0.70) after the last monitoring visit.

Caregiver involvement.

All 9 consenting participants agreed to inviting their parent. Of these, 6 caregivers consented. Five completed baseline and three completed the outcome questionnaires.

PSE-S.

Recruitment.

The study aimed to recruit 15 students to PSE-S between July and October 2021. Flyers were distributed via social media, autism recruitment websites, local college autism programs and professional networks; recruitment was extended to February 2022 due to low enrollment. Of 28 interest forms, 18 participants did not respond to attempts to contact them within one week. Ten students consented; one withdrew due to personal reasons before baseline assessment.

Screening.

Of nine Baseline assessments, one was lost to follow-up, one was not eligible, and one withdrew due to the nature of the questionnaires emphasizing dysfunction in daily life. The remaining six completed the ESP-S.

Monitoring and Outcome.

Completion rates were high (monitoring questionnaires: 92%, satisfaction surveys: 100%, and clinical outcome surveys:83%). Outcome visits were conducted for 83% and occurred 1.34 weeks (SD=0.62) after the last monitoring visit.

Acceptability

PSE-C.

Overall satisfaction (M=3.78, SD=0.44), positive impact/benefit (M=3.67, SD=1.00), content relevance (M=4.11, SD=0.60) and acceptability of telehealth format (M=3.89, SD=0.33) were in the neutral-to-positive range (Fig.3b). Acceptability of weekly monitoring was neutral (M=3.33, SD=0.50). Visit(s) to create the ESP-C were rated as helpful and when asked about the specific ESP sections, things I can do to manage my emotions had the highest average ratings of helpfulness (Fig.4b). Most would recommend the ESP-C intervention to another person (Fig.3b).

PSE-S.

Creation of ESP.

Participants rated their experience of creating the ESP-S as Neutral (n=3) or Easy/Very Easy (n=3), with average time to create it reported to be 56 minutes (range=10–120 minutes). Worksheets to facilitate Brainstorming and creating the ESP were used and rated as Very Helpful by all 6 participants. Transcripts and/or videos for brainstorming coping strategies were used and rated as helpful by 5 participants.

Overall satisfaction (M=3.33, SD=0.82; Fig.3c), positive impact/benefit (M=3.33, SD=0.82), and content relevance (M=3.33, SD=1.03) was rated as neutral. Telehealth format (M=4.17, SD=0.75) and weekly monitoring were generally acceptable (M=4.00, SD=0.63; Fig.3c). Things I can do to manage my emotions had the highest ratings (Fig.4c). Three students would recommend ESP-S to another person (M=3.33, SD=0.82).

Weekly Stress and ESP Use

PSE-C.

Average stress ratings across the study period ranged from 1.00–3.57 (M= 2.39, SD=0.91). All participants reported using the ESP (M=5.78, SD=2.86; of 11 weeks queried: 1–2 weeks: n=2; 6–8 weeks: n=6). Three participants reported use of strategies without referring to the ESP. Higher weekly stress levels predicted ESP-use (OR=1.82, 95%CI=1.79–1.85, p<.001).

PSE-S.

Weekly stress ratings were not collected in the self-report questionnaires due to study error. Reported ESP use ranged from 0–11 weeks (M=4.83, SD=4.92; of 11 weeks queried: never: n=2; 2 weeks: n=1; 6–11 weeks: n=3).

Results: Combined Studies

Although there were small differences in the EMA period and compensation (Figure 2), EMA results are combined across studies due to small sample size.

EMA feasibility.

The overall compliance rate for EMA surveys was 40% (SD=22%). The average number of surveys completed was 88.6 surveys (SD=49.9). No significant differences in compliance rates were detected across COVID-C, PSE-C, or PSE-S study group (F(2,11)=0.135; p=.875). See Table S2 for participants’ weekly response rates.

EMA acceptability.

Average acceptability of completing EMA surveys multiple times daily was neutral (M=3.00, SD=1.32; Figure 4a–c.). Those who disagreed reported disliking the frequency of prompts each day. Regarding helpfulness of the EMA surveys as a reminder that the ESP was available and benefit of the EMA, average ratings were neutral (M=3.00; SD=0.97; M=3.00, SD=1.11, respectively).

Qualitative Feedback.

Qualitative content analysis across studies yielded five themes highlighting participants’ views of ESP intervention (see Table S3 for quotes).

1. Increased Emotion Regulation. 64% indicated that the ESP contributed to increasing their emotion regulation (i.e., developing skills to regulate, manage and cope with stressors).

2. Usefulness of having a toolbox of coping skills. 51% indicated that steps outlining ways to cope with distress were particularly helpful. Some noted not referring to the physical ESP, but knowing that it was there or recalling information from putting it together as being helpful.

3. Improved ability to identify triggers/warning signs. 36% commented that the Warning Signs section of the ESP helped them to be more aware of triggers and/or signs that they may (or were already starting to) become distressed.

4. Benefits of psychoeducation. 21% indicated that the process of creating the ESP had simply taught them something that was helpful.

5. Starting Point. 9% of participants indicated viewing the ESP was a starting or jumping-off point for determining what works best for the individual.

Reasons for ESP not being of benefit.

Qualitative responses were also reviewed for four participants who indicated Disagree with items regarding overall satisfaction or positive benefit. One participant’s response did not elaborate on negative aspects of the intervention. The other three suggested strategies were not helpful, with two indicated they already knew information outlined in the ESP (one person had a therapist, the other had hoped new strategies might be identified), but that the ESP might be helpful to others. One participant elaborated that calling sections “steps” was not quite correct, because they did not feel that the ESP was about completing a task.

Discussion

Preliminary results across the two studies support the acceptability of the clinician-guided ESP, with at least 86–89% of participants reporting the intervention had a positive impact on them and 67–71% indicating they would recommend the ESP to another person. Acceptability of the self-guided ESP was more mixed, with 3 of the 6 participants indicating positive benefit and that they would recommend the ESP. Feedback suggested that two of participants who were less satisfied with the ESP-S indicated they already knew the strategies outlined in the ESP. This underscores the point that a small number of participants highlighted in their feedback: that the ESP is likely best viewed as a starting point to explore what works best for an individual. Thus, while caution is exercised in interpreting findings from such a small sample, this feedback suggests that this very brief intervention may not be appropriate for people who have current/prior therapy experience. It is possible that the ESP-C would be more beneficial for such people, where a clinician (their current therapist, if they have one) would offer more flexible exploration of new strategies. Future research is needed to explore the potential uses of the ESP and to directly investigate for whom a clinician vs. a self-guided version may be more appropriate.

Across both ESP-C and ESP-S, content was rated as relevant to most participants. There was some variability in the ratings of the individual sections of the ESP. Participants rated the coping strategies section as the most helpful. Indeed, identifying coping strategies is a core focus of the ESP intervention; for both the clinician- and self-guided versions, creation of the ESP starts with brainstorming coping strategies. Qualitative feedback reinforced this focus, with emerging themes highlighting the benefits of increased emotion regulation and having a toolbox of coping skills. Additionally, participants noted improved ability to identify triggers or warning signs that these strategies may be useful, though quantitative ratings of how helpful the Warning Signs part of the ESP were variable. In future adaptations of the ESP intervention, it may be beneficial to develop other resources to improve this section of the ESP.

As is customary for feasibility studies, target sample sizes were determined by availability of resources and sufficient numbers of participants to assess preliminary feasibility and acceptability and estimate effect sizes to inform power analyses for future trials. The first notable factor regarding feasibility was the difficulty recruiting enough college students for the PSE-C and PSE-S to meet target sample sizes during the enrollment period. Notably, many students approached for the PSE-C were already engaged in a program providing academic supports for autistic students; in addition to being busy, they may feel less need for this intervention because they had other supports in place. It is important to note that recruitment for the PSE study occurred during Fall 2021, which was the first semester that many campuses were returning to in-person activities. Though the ESP was intended to help cope with stress, participating in a research study may have been viewed as another stressor when students were already navigating a busy and potentially stressful transition. Further engagement with autistic PSE students will be important to better understand whether there is a need for more clarity around how the ESP may be useful for PSE students, or if timing is the most important factor to be addressed. For example, it may be that providing the ESP as a part of university orientation or enrollment in the college support program may help to mitigate feelings of burden during a busy transition time.

For those who did enroll, compliance rates were high across monitoring and follow-up visits and surveys. Coupled with participants positively rating the acceptability of the telehealth format support the feasibility of the study protocol. Notably, while the weekly monitoring was acceptable to most COVID-C and PSE-S participants, two-thirds of PSE-C participants responded neutrally. This may be because these participants already had weekly academic supports from the college program. That most PSE-S participants agreed with the weekly monitoring being acceptable may reflect that they were only asked to complete questionnaires; for busy college students, the need for monitoring visits should be considered. The interactive component may have been more acceptable to participants during COVID, when people were still experiencing varying levels of social distancing.

Regarding EMA feasibility and acceptability, more than two-thirds of participants agreed or were neutral that receiving multiple daily prompts was acceptable. Notably, the 40% response rate, was within the range of compliance rates seen in other EMA studies of the general population 23–94% (Williams et al., 2021). Though somewhat low at face value, previous research has demonstrated that response rates commonly vary considerably between participants (Williams et al., 2021), including autistic youth (Khor, Gray, Reid, & Melvin, 2014). Consistent with that, the current adult sample had a standard deviation of 22% for response rate. While the lengthy study period had bearing on some people (i.e., four participants had high response rates in week 1 and a noticeable drop in week 2), individual variability seemed to be more important, with 10/14 participants showing relatively constant – some high, some low – response rates across the 6–8 weeks. Interestingly, response rates did not differ across the two studies, despite differences in monetary compensation, warranting further exploration on motivating factors to complete EMA. Most importantly, considering that few participants agreed that the EMA served as a helpful reminder or had other benefit, it seems clear that sending prompts along the entire duration was unnecessary. Focus groups with autistic people are needed to gain insights into the utility of EMA and optimal frequency with different subgroups of autistic adults. Shorter periods would impact resources required to conduct future studies, as the staffing required to provide 12-hour monitoring, 7-days per week, for such extended periods limited us to enrolling only a small number of EMA participants.

In line with EMA not being a particularly helpful reminder, many participants indicated that they recalled ESP strategies without actually accessing the physical plan on their device. That the novel technological components (i.e., EMA; saving and accessing the ESP on a device) originally thought to be important to the utility of the ESP may not be essential are important insights gained from this feasibility trial. Not having to maintain a website or app to host individual ESPs or provide EMA prompts facilitates dissemination and enhances accessibility and scalability.

Responses from a few participants who indicated less satisfaction or benefit from the ESP, highlighted the limitations of this brief intervention, either being redundant with therapy or perhaps not sufficiently in-depth to uncover novel strategies. Notably, the self-guided intervention received more neutral overall feedback, though ratings regarding the materials to support creation of their ESP-S were generally positive. Considering the current nationwide shortage of mental health professionals (Bureau of Labor Statistics, 2020), and a gap in professionals specifically trained to support the mental health needs of autistic adults (Jadav & Bal, 2022; Maddox et al., 2021), the ESP intervention may also be beneficial in situations where stress may be experienced and/or anticipated (e.g., transitions, major events). Additionally, the ESP may be useful in clinical settings, perhaps as a resource for individuals when waitlists for more formal treatment preclude timely support or as a supplement to treatment to promote implementation of strategies that are sometimes difficult to generalize outside the therapy session (White, Connor, Beck, & Mazefsky, 2020). As noted above, further testing is needed to evaluate its utility in different settings and subgroups who may derive benefit from using it.

Limitations

There was gender diversity across studies and a range of educational backgrounds represented in the COVID study, limited racial diversity in the context of very small samples in a non-randomized feasibility trial are considered preliminary. Additionally, we were limited to collecting EMA from a small subsample of participants; larger samples are needed to further test the use of EMA to monitor ESP use and outcomes. Despite limitations, combining feasibility data across COVID and PSE studies provided important insights to inform future adaptations to the ESP and studies to test its efficacy in different settings and subgroups of autistic adults.

Conclusions

In sum, these studies provide preliminary support for the acceptability of the ESP intervention and feasibility of the design to assess its potential efficacy in future studies. While the ESP is not intended as a clinical intervention that will fill the gap in treatments needed to address high rates of co-occurring conditions, we hope that its brief nature will promote accessibility and provide a starting point for autistic adults who may be proactively seeking support to cope with anticipated stressors or would benefit from targeted strategies during stressful life events. Future studies are needed to test the ESP’s use as a primary or secondary preventative strategy, both for promoting positive coping during specific periods and mitigating longer-term mental health burden. While many participants reported recalling the ESP strategies from memory, rather than accessing it on their device as intended, it is possible that autistic adults will benefit from the structured approach to creating this plan in a range of settings.