Psychoeducational Group Intervention for Adults with Autism Spectrum Disorder and Their Close Relations (Prisma) – An Open Feasibility Study


 Background: Autism spectrum disorder (ASD) in adulthood is associated with severe impairments in functioning and poor health, while ASD is also affecting close relations. Accessible first-line interventions addressing the complex clinical needs and care coordination are lacking. Methods: This study investigated the feasibility and preliminary effects of a new psychoeducational group intervention (Prisma) developed for intellectually able adults with ASD and their close relations in an outpatient setting. Results: Completion rate was 77% (n=71) of the 92 adults with ASD and 73% (n=69) of the 94 close relations. Participants considered Prisma to be an acceptable intervention and their feedback will be used to further improve the Prisma for an upcoming RCT. Preliminary analyses of effects showed promising results with an increase in knowledge of ASD. Conclusions: Overall, results indicate that the Prisma is a feasible first-line intervention in a stepped-care process in outpatient services. Trial registration: Clinicaltrials.org (NCT04460976).

of each session. The intervention consists of four 2-hour sessions (including breaks, time for questions, and structured mapping of individual needs). For descriptions of content, themes, and focus of the four Prisma sessions, see Table 1.

Treatment delity
To increase treatment delity, a half-day training course, including an introduction to the intervention and course contents were given to course leaders. Also, course leaders received on-going support from project coordinators (via email, digital platform, telephone, or visits at the respective clinic/center) throughout the study, regarding all issues related to the intervention as well as the course leader's role. All study and intervention materials were thoroughly structured and made available to the course leaders via a digital platform. Furthermore, the course leaders had access to a manual containing all the parts that were included in the course leader training, answers to FAQs, a structured checklist for time planning, and ready-made suggestions for administration.

Participants and recruitment
This was an open feasibility study in a clinical outpatient context including adults with ASD and their close relations. The data collection was conducted in 2017 at eight adult outpatient psychiatric clinics and four habilitation clinics [1] in the Stockholm area. Each clinic organized a group of approximately [10][11][12][13][14][15] participants (M=13) with ASD and 1-2 close relations per patient. In total, 13 groups received the psychoeducational intervention Prisma. Information about how to participate in Prisma was given to patients at each clinic through information brochures in the waiting rooms and/or by clinicians. All patients that expressed interest in participation were contacted by one of the course leaders. A structured screening interview including information about the content of Prisma, as well as a brief assessment of inclusion and exclusion criteria, was conducted by telephone or at the clinic. Individuals who ful lled the requirements were invited to an information meeting with their close relations where they received more information about the intervention, gave consent for participation, and completed the baseline measurement. An additional individual and nal ascertainment of eligibility performed with each patient and his or her close relations to ensure that the prerequisites for participation were met.

Inclusion and exclusion criteria
For inclusion, patients had to meet DSM-IV and/or DSM 5 criteria for ASD and/or ICD-10 criteria for one of the ASD diagnoses under F84, assessed within the Swedish health care system before the study participation. Both patients and the close relations had to be 18 years or older. Su cient knowledge of the Swedish language was required to understand the course contents. Close relations could be a parent, sibling, partner, friend, or whomever the participant with ASD thought of as a close relation. Having ASD or other diagnoses were not considered exclusion criteria for close relations.
Exclusion criteria for the participants with ASD were not being able to participate with a close relation, intellectual disability, mental or psychosocial instability to a degree that made participation impossible as judged by the course leader or experienced health care professionals (i.e. severe psychiatric comorbidity such as ongoing substance use disorder, manic episodes, psychosis, and acute suicidality). Further reasons for exclusion were the inability to participate in a group, or severe life situations (e.g. homelessness). Parallel treatments and interventions like pharmacological or occupational therapy were not an exclusion criterion.

Measures
Demographic data Case histories and sociodemographic data for participants with ASD were extracted from medical records. The participants also completed a questionnaire "Current Life Situation Form" covering demographic information and current stressors in different areas of life [24]. A modi ed version of this questionnaire was used to assess demographic characteristics of the close relation. ASD symptoms were measured using the Ritvo Autism Asperger Diagnostic Scale-14 screen (RAADS-14) [25]. Background and demographic data are described in Table 2 for adults with ASD and in Table 3 for close relations.
Primary outcome: feasibility The focus of this open study was the feasibility of the Prisma intervention in a clinical outpatient setting. Outcome measures were gathered through selfratings before, during, and after Prisma. In addition, for preliminary estimation of treatment effects, self-rating questionnaires were administered at baseline, i.e. 1-2 weeks before the intervention started, and post-intervention, i.e. 1-2 weeks after the last session.
Treatment completion was the proportion of participants who completed the intervention. To be regarded as a completer, the participant had to attend at least 3 out of 4 course sessions.
other two assess the participant's experience of taking part in group discussions/exchange of experiences and could also be answered "not applicable" if the participant did not share or discuss experiences. At the end of each form, the participants could write comments about the session.
A modi ed 12-item version of the Patient Evaluation Form (PEF) [28] was distributed at the end of the last session regarding the participant appraisal of the course as a whole. Six items are scored on a Likert scale ranging from 0-4 (Cronbach's alfa 0.78, n = 132). Four items are open questions for participants to further develop their answers ("How did the intervention help me?"; "How can the intervention improve?"; "What could I have done differently?"; "Is there anything else you want to share about the intervention?"). For the open questions in the PEF, a thematic analysis was performed based on previous suggestions [29]. As the non-completers only participated to a limited extent and did not assess the full content of the intervention, only data from completers were included in the main thematic analysis. However, a separate analysis was conducted on non-completers. As participants who answered the open questions at times gave more than one answer, the percentage in Table 6 re ects the proportion of answers rather than individuals. To be considered a theme, a minimum of four similar answers had to be present. One of the authors (DS) analyzed and categorized the answers in themes and another author (NH) con rmed them. The agreement was very high and the few deviating analyses were discussed and placed in the theme's authors agreed upon. Yet another author (TH) reviewed and con rmed the categorizations. After a consensus discussion with all three researchers (NH, DS, TH), slight changes were made to the description of the themes to clarify what they re ected.
Adverse events were de ned as spontaneous oral complaints or instances when patients stated that they experienced negative or unwanted effects during the intervention period. Serious adverse events were de ned as events that involved hospital care/hospitalization, etc. due to the intervention. Adverse events and serious adverse events were recorded in the case report form (CRF) folder, and it was assessed by course leaders if these were caused by the intervention.

Preliminary effectiveness outcomes
All scales that measure preliminary effectiveness were administrated to participants before and after the intervention except for the Burden Assessment Scale (BAS) [30] which was only lled in by close relations.
Acquired knowledge of ASD and support and services was measured using the ASD 20 Questions (Appendix 1). ASD 20 Questions is a knowledge quiz with 20 true/false/don't know scored items created for this study. Moreover, a separate question was included where participants were asked to list all support and treatment interventions helpful for adults with ASD and their close relations. Similar knowledge quizzes have been used in previous studies [15,27]. Internal consistency of the quiz using Kruder-Richardson 20, at pre-intervention was 0.78 (n = 138). High values indicate more acquired knowledge.
Relationship quality was measured both from the perspective of the participant with ASD and the close relation with The Questions About Family Members (QAFM) [31]. Mental health. The Hospital Anxiety and Depression Scale (HADS) [32] was used to measure mental health on the two subscales: "depression" Cronbach's alfa pre-intervention 0.84 (n = 139) and "anxiety" Cronbach's alfa pre-intervention 0.88 (n = 138). The subscales contain seven items each and were scored on a 0-3 Likert scale with a maximum score of 21 points. Higher scores indicate high symptoms of anxiety and/or depression. Furthermore, two items from the PEF covering participants' well-being before and after the intervention on a Likert scale of 1-10 ("How would you rate your well-being before the intervention?", "How would you rate your current well-being") was used to measure general well-being. High values indicate high well-being.
Quality of life was measured using the Satisfaction With Life Scale (SWLS) [33]. SWLS contains ve items, scored on a 7-point Likert scale Cronbach's alfa pre-intervention 0.91 (n = 139). High values indicate higher satisfaction with life.
Acceptance of diagnosis, adapted from the Acceptance and Action Questionnaire -II [34], was measured using "What I think about my diagnosis" for adults with ASD and "What I think about my close relation's diagnosis" for close relations. Cronbach's alfa at pre-intervention was 0.86 for participants with ASD (n = 68) and 0.87 (n = 68) close relations about the adult with ASD's diagnosis. Both questionnaires contain 7 items, scored on a 7-point Likert scale where lower values indicate higher acceptance.
The burden of care on close relations was assessed using the Burden Assessment Scale (BAS) [30] consisting of 19 items scored on a 4-point Likert scale Cronbach's alfa at pre-intervention 0.87 (n = 68). BAS was used to measure to what extent the close relations experienced a subjective (e.g. emotional distress) and objective (e.g. economic consequences) burden of care because of the relationship with the person with ASD. High values indicate a greater burden.

Statistical analysis
Statistical outliers in all outcome measures were screened using boxplots in SPSS version 27. The few extreme outliers (1st/3rd Quartile ± 1,5) that were identi ed did not signi cantly affect the results and were therefore retained in the subsequent analyses. Outcome data were approximately normally distributed. The main statistical analyses regarding feasibility and e cacy-related measures were performed on all participants attending at least 3 out of 4 sessions and who had completed pre-(T1) and post-measurement (T2).
Comparisons between completers and non-completers were performed on baseline data using unpaired t-tests, Fisher's exact test, and Pearson's chi-squared test to detect possible predictors of drop-outs. Unpaired t-tests were used to examine differences post Prisma between individuals with ASD and close relations on the SEF and the PEF. Paired t-tests (pre and post) were used to test preliminary e cacy both for individuals with ASD and the close relations for all outcomes. The effect size was interpreted according to Cohens d: 0.2 = small effect size, 0.5 = medium effect size and 0.8 = large effect size [35]. [1] The Stockholm's healthcare services (like several other regions in Sweden) is organized so that patients with ASD and other enduring disabilities receive support from habilitation services for functional impairments, and from psychiatric services for severe comorbid mental illness.

Feasibility
Treatment completion The ow of study participants is presented in Fig. 1. A total of 143 adults with ASD were screened for eligibility to participate. Out of these, 12% (n=17) could not be included in the study due to that attending with a close relation was a requirement and 24% (n=34) declined participation, with the most common reason for this being that the date/time of the upcoming intervention did not t with other activities in their lives. Of the adults with ASD screened for intervention, 64% (n=92) and 94 close relations were included in the intervention. Of those enrolled in Prisma, 77% (n= 71out of 92) of the adults with ASD and 73% (n= 69 out of 94) of the close relations met the prede ned criteria for completion (attending at least 3 out of 4 sessions). There were no signi cant differences between completers and non-completers in the background and demographic data ( Table 2 for adults with ASD and Table 3 for close relations).
However, there was a trend (p = .051) towards a difference for the age of participants with ASD, where non-completers were younger (mean age 27) than completers (mean age 32).

Acceptability
Treatment credibility. Both ASD participants and their close relation reported signi cantly higher credibility post-intervention compared to before starting (Table  4).
Treatment satisfaction. All items and all sessions were rated as being satisfactory "to some extent" on average (mean 2.52 -3.53 on a scale from 0-4). Levels of satisfaction were the same between ASD participates and their close relations with three exceptions rated higher by the close relations (see Table 5).
For all sessions, 17% -42% of the ASD participants and 42% -48% of close relations answered "Not applicable" to the statement "It was helpful to share experiences with other participants", thus indicating that they did not participate in experience sharing during the sessions.  Figure 2. Participants also rated the intervention as a whole (0 = "Not approved", 1= "Approved", 2 = "Well approved" or 3 = "Very well approved") with a mean score of 1.97 for participates with ASD and 2.13 for close relations.
A thematic content analysis of the participants' answers for the open questions in the PEF is presented in table 6.
The major themes that emerged in these analyses suggest that the intervention both gave more knowledge as well as acceptance towards the diagnosis but that participants wanted more interactive activities. A separate analysis was performed with participants that dropped out during the intervention. The themes were similar to what was observed for completers, but as very few answers were given, these results should be interpreted with caution.
Adverse events and serious adverse events. No serious adverse events were reported. There was one adverse event that was judged to be related to taking part in Prisma. One participant with ASD reported a deteriorated mood from participating in the intervention and therefore chose not to continue. In addition, 8% (n=7) of the patients with ASD reported increased levels of depression and/or anxiety, but it was not speci cally reported that it was linked to taking part in the intervention. In two of these 7 cases, external causes of the deteriorating mood were speci ed by the participants.
Group leaders' rating of treatment credibility. To evaluate the clinicians' perspectives on the Prisma program, group leaders who administrated the intervention completed an adjusted version of the Treatment Credibility Scale (TCS, see above). Course leaders' evaluation of Prisma was just above 7 on a 1-10 scale of the TCS both pre-and post-intervention.

Preliminary effectiveness
Descriptive statistics and results from pre-to post-intervention for all preliminary effectiveness measures are reported in Table 7 for both participants with ASD and their close relations.
Acquired knowledge of ASD and support and services. Knowledge of ASD increased for both participants with ASD and close relations with large effect sizes. Knowledge about support and services also improved with medium-sized effects.
Relationship quality. Measurement on QAFM indicates small signi cant positive effects between pre-and post-assessments for close relations on perceived emotional involvement, critical remarks, and perceived emotional over-involvement. No signi cant changes were observed in patients with ASD.
Mental health. Both depression and anxiety signi cantly decreased for participants with ASD (small effect sizes) whereas there was only a signi cant decrease in depression for close relations (small effect size). Also, participants were asked after the intervention to rate their well-being both before and after the intervention using the PEF (see above Quality of life. There was a small signi cant effect indicating that both groups experienced a better quality of life after receiving the intervention. However, this effect was only signi cant in patients with ASD and not their close relations when analyzed separately for the two groups.
Acceptance of diagnosis. Participants with ASD reported slightly better acceptance of their diagnosis after receiving the intervention.
The burden of care on close relations. A small decrease in the subjective burden of care was observed for close relations while the objective burden was unchanged from pre-to post-intervention.

Discussion
This study evaluated the feasibility of the novel manualized psychoeducational intervention Prisma for adults with ASD and their close relations, thus addressing the need for the development of scienti cally evaluated interventions [11]. The vast majority completed Prisma and overall, it was perceived as an acceptable intervention by both adults with ASD and their close relations. Moreover, preliminary analyses of effectiveness indicated increased knowledge and well-being. However, this study also identi ed areas of improvement such as the intervention's ability to enhance active participation.

Treatment completion
Of the adults with ASD who were included 77% (n=71 of 92) completed the intervention (attended ≥ 3 lectures), which was slightly more than for the close relations 73% (n=69 out of 94). Compared to the general attrition in regular psychiatric services, which varies substantially from 26% to 82% [36] the attrition in Prisma was in the lower range. Regarding adults with ASD, there is a lack of studies speci cally reporting attrition; however, the observed attrition in the current study was similar to what has been reported for autistic adolescents [15]. This suggests that the attrition rate for the Prisma intervention was acceptable, despite that the study was conducted in a clinical outpatient context not only recruiting from the clinics' usual patient base, but also administered by staff members at these clinics. We speculate that one of the reasons for the high treatment completion was participating with a close relation. Participating with a close relation may provide an opportunity for supporting each other in overcoming obstacles for participation during the intervention and after in the continued care process. Another important factor might have been that the clinical staff perceived Prisma as creditable. Again, this should be considered as promising for possible future implementation as these were regular staff members in the healthcare settings where Prisma is intended.
The background and demographic characteristics of the included sample (e.g. educational level, or psychiatric comorbidity) corresponded well to previous studies on clinical samples [37][38][39] thus indicating that the results may generalize to similar clinical outpatient contexts for adults with ASD. The gender balance was 49% female, i.e. relatively close to the male-to-female ratio 3:2 reported for adult patients (18 years of age and older) in the Stockholm Region between 2012-2016, who received an ASD diagnosis without intellectual disability according to the Center for Epidemiology and Community Medicine, Region Stockholm (personal communication, 21 May 2021). Dropout analyses showed a trend towards signi cance regarding age, indicating that those who did not complete Prisma were a few years younger. Participants who did not complete Prisma gave very few comments in the evaluation limiting our ability to identify possible areas of improvement regarding how to make Prisma more acceptable for young adults.

Acceptability
An important outcome of Prisma was acceptability (i.e. how participants react to the intervention). One challenge was building general trust for health care providers as previous research has shown that knowledge and awareness of ASD can be limited also among professionals [10] and autistic individuals nd it hard to access mental health support and experience high levels of stigma [8]. The increase in perceived credibility from pre-to post-intervention should therefore beregarded as important, especially as it is associated with post-treatment outcomes [40]. Furthermore, the reported treatment satisfaction was high in both autistic adults and their close relations. However, the close relations rated certain aspects as more satisfactory than adults with ASD. Based on the qualitative analysis of re ections from participants, the intervention could be improved especially for adults with ASD who wanted more possibilities to share their own and listen to others' experiences. This is in line with research showing that communication is more effective and motivating when autistic individuals share information [41]. Moreover, close to half of the participants in this study reported that they did not participate in the discussions. This indicates that Prisma needs to be revised to promote active participation in the intervention and the subsequent health care process.

Preliminary effectiveness
A secondary aim was to investigate the preliminary effects of Prisma. Knowledge of ASD and of available support/interventions improved for both groups with medium to large effect sizes, which is comparable to what has been reported for adults with ADHD and their close relations [19], as well as adolescents and young adults with ASD [15] after receiving psychoeducation. Means and individual scores of correct answers pre-and post-intervention were on the higher end of the distribution for knowledge of ASD, thus indicating possible ceiling effects that need to be adjusted in the coming RCT.
Regarding the relationship quality, close relations perceived the relationship with the person with ASD as more positive after the intervention regarding emotional involvement, criticism, and emotional over-involvement. Also, close relations reported feeling a decrease in the subjective burden of care (i.e. emotional distress), despite the remaining objective consequences such as negative effects on the economy or the close relations' time and activities.
However, from the perspective of the individuals with ASD, we saw no changes in relationship quality, thus indicating the need for further changes to the intervention to enhance mutual active participation. However, preliminary results related to relationship quality (perceived criticism and perceived emotional involvement) should be interpreted with caution as these scales showed poor internal consistency.
Preliminary analysis of depression and anxiety symptoms showed signi cant decreases with small effect sizes. Even though participants generally wanted more possibility to interact with others, we speculate that it is possible the intervention in its current structure also have led to recognition between participants, a sense of belonging, acceptance, and therefore, fewer symptoms of depression and anxiety. This is also supported by that large number of the answers from adults with ASD indicated that the intervention was helpful by increasing the acceptance of diagnosis, courage, recognition, and self-awareness (see Table 6). Similarly, the decrease in depression and anxiety symptoms was consistent with the increase in general well-being.

Future revision of the Prisma intervention
Since the adults with ASD and their close relations were not part of the group that developed Prisma, it was crucial to conduct the present feasibility study so that we could further improve the acceptability and accessibility of the Prisma program based on the participants' feedback. In line with this feedback, an extensive revision was carried out. The opportunities to exchange experiences were increased and updated including structured discussions in small groups with predetermined topics. These discussions now also include a method ("communication tra c light system") for how participants can choose to participate to the extent that they prefer. Furthermore, to increase treatment acceptability, a revision was done with an increased focus on strength to balance the focus on ASD-related obstacles, following re ections from participants with ASD in the current study. The course leader manual was updated including more concrete ways in which they could support the participants. Other updates have been made emphasizing that course leaders need to conduct individual post-intervention follow-ups regarding a plan for the continued health care process after this rst-line group-based psychoeducational program. How these changes are received and how this will affect the effectiveness will be evaluated in an ongoing RCT.

Limitations
The results of this study need to be considered in light of some limitations. First, the relatively small size of the group of non-completers might have limited our ability to detect signi cant differences. This will however be addressed in the next study phase including a much larger sample. Second, this study did not include any controls and we cannot be sure that all observed effects are related speci cally to Prisma.

Conclusion
In summary, the overall patterns in the results indicate Prisma to be feasible and the preliminary effects are promising. Areas of improvement and limitations were identi ed, which have been addressed in the revised Prisma program used in forthcoming study phases. Hence, we, therefore, feel encouraged to make the suggested adjustments and continue with a RCT.

Declarations
Ethics approval and consent to participate The study was approved by the Regional Ethics Committee in Stockholm, Sweden (2017/1065-31/1). It was performed in accordance with the ethical standards of the Declaration of Helsinki (1964) and its later amendments. Informed consent was obtained from all individual participants included in the study.

Consent for publication
Not applicable Availability of data and materials The datasets used and/or analysed during the current study are not made available due to the regulations in the ethical permission for this study.

Competing interests
Three of the authors (NH, CB, AB) participated in the development of the intervention materials included in this study. Two of the authors participated in the revision of the intervention materials and co-authored the intervention manual following current study (HA, TH). Intervention materials and manual are owned by Region Stockholm, why there are no nancial interests to declare. None of the other authors have any con ict of interest.

Funding
The project received funding from the Swedish Association for Local Authorities and Regions (Mission Mental Health for the product development, education and implementation. Research funding was received from Region Stockholm funds for clinical research (ALF).   -Interest organizations, getting in touch with similar others online or in real life, and links for more information. Separate columns are reported for participants who completed Prisma and for those who dropped out including signi cance testing of baseline differences. Separate columns are reported for participants who completed Prisma and for those who dropped out including signi cance testing of baseline differences.  • Asked questions: 6 (17%)