Guided self-help for depression in autistic adults: the ADEPT feasibility RCT

Inclusion criteria

• Was aged ≥ 18 years.

• Had a clinical diagnosis of ASD.

• Had current depression as measured by a PHQ-9 score of ≥ 10.

Exclusion criteria

• Did not speak English.

• Had literacy level such that the written materials would have been inaccessible.

• Had risk of suicide such that a low-intensity intervention would not be in line with clinical need.

• Had a history of psychosis.

• Had current alcohol/substance dependence.

• Had untreated epilepsy.

• Had attended more than six sessions of individual CBT during the previous 6 months.

Bristol Autism Spectrum Service

Potential participants were identified by clinicians following attendance at the clinic for diagnostic assessment. Participants were considered potentially suitable if they were given a clinical diagnosis of an ASD at the end of the assessment and were identified as having low mood following a clinical interview and/or a pre-clinic screening PHQ-9 score of ≥ 10, and routine risk assessment contraindicating low-intensity intervention. Potential participants were introduced to the study by clinicians using a brief expression of interest (EOI) form. If they were interested in the study, they could return the EOI form expressing interest and requesting follow-up contact from the research team. This further introduction to the study could occur face to face in the clinic, or by postal/telephone follow-up after the clinic attendance if the clinician considered this more appropriate. Follow-up contact from the research team was ordinarily over the telephone; full information about the study was provided at this stage in the study information sheet. Standard recruitment procedures were then initiated.

Everyone Included

The Everyone Included review panel (with independent membership) approved the study for inclusion in the approach and a draft of the research opportunity letter. Potential participants were identified from an automated search of the electronic patient record system by the information analysis team in the NHS trust and based on the study inclusion/exclusion criteria. Potential participants were sent the research opportunity letter by the local NHS research and development (R&D) team and, if they made contact, were provided with a study information sheet and asked their permission for the study research team to make contact with them. No information about potential participants was sent to the study research team without permission first being gained from the potential participant. Standard recruitment procedures were then initiated.

Adult Autism Clinic in Northumberland, Tyne and Wear NHS Foundation Trust

Potential participants were identified by clinicians as those given a diagnosis of an ASD following clinic attendance and assessed as having depression sufficient to warrant signposting/referral to Increasing Access to Psychological Therapies (IAPT) services as per routine clinical practice. They were introduced to the study using the brief EOI form at the end of the diagnostic assessment process, ordinarily at the second assessment appointment. Potential participants interested in the study who contacted the research team were asked to complete the PHQ-9 either by post or over the telephone, and those individuals with a score of ≥ 10 were provided with the full study information sheet. Standard recruitment procedures were then implemented.

In addition, patients formerly diagnosed by the NTW Autism Diagnosis Service who had agreed to be contacted about research studies were provided with information about the project (EOI form). Potential participants who contacted the research team were provided with comprehensive information about the study and potential suitability assessed using the PHQ-9. Standard recruitment procedures were then initiated.

Adult Autism Spectrum Cohort – UK at Newcastle University

The cohort research team approved the study on application, and an introductory letter was drafted by the Newcastle University research team. Potential participants were identified from a search of the database, constrained by NTW area postcode, age, intellectual disability and history of depression, and then cross-matched with the NTW Adult Autism Clinic research register to ensure that people were not introduced to the study more than once. Potential participants were sent the introductory letter and EOI form. Those who contacted the research team were provided with comprehensive information about the study and asked for their permission to screen them for the study, including completing the PHQ-9 by post or over the telephone. Standard recruitment procedures were then initiated.

Standard recruitment procedures

The outcome of screening each potential participant using the PHQ-9 was communicated to the individual’s general practitioner (GP) with their permission.

Potential participants with a PHQ-9 score of ≥ 10 who wanted to take part in the study were invited to an eligibility assessment with the research team.

Treatment as usual

Treatment as usual consisted of standard NHS care for depression as provided to the individual during the trial. This included no treatment, GP support, referral to IAPT services, referral to secondary mental health services and antidepressant medication. Participants were asked to report any treatment received at each follow-up meeting.

Guided self-help

The intervention developed in the current study adhered to the principles of a low-intensity psychological intervention for depression as outlined in NICE guidelines,29 namely 6–10 sessions of individual GSH based on the principles of CBT to include BA.

The GSH treatment developed for the study was a low-intensity intervention based on the principles of BA for depression, which was adapted for autistic people.

Patient Health Questionnaire-9 items

The PHQ-922 is a nine-item self-report measure of depression that is commonly used in primary care settings. The PHQ-9 has been found to be reliable (Cronbach’s α = 0.84–0.93), valid and sensitive to change in the general population. 53 To the authors’ knowledge, the psychometric properties of the scale have not been investigated in the autistic population.

Beck Depression Inventory-II

The BDI-II25 is a widely used 21-item self-report measure of depression. The BDI-II has been found to be reliable (Cronbach’s α = 0.92) for outpatients. 25 A validation study24 of 50 young people with ASD reported good internal consistency on the BDI-II (Cronbach’s alpha = 0.90).

GRID-Hamilton Depression Rating Scale

The GRID-HAMD54 is a version of the Hamilton Depression Rating Scale (HAM-D). 55 It is a 17-item clinician-administered interview that has been found to be reliable and valid in the general population but, to our knowledge, has not been investigated in the autistic population. 54 GRID-HAMD interviews were audio-recorded, with participant consent, for the purposes of inter-rater reliability. The first six HAM-D interviews by each interviewer were subject to a second independent rating, excluding items 8 and 9 (i.e. observation of psychomotor retardation and agitation), which need to be assessed face to face, to establish the reliability of each assessor. To establish reliability across the study, a random sample (20%) of GRID-HAMD recordings were independently rated.

Participant Global Rating of Change

Participants were asked to rate whether their depression was better, worse or much the same on a 5-point scale at the 16- and 24-week follow-ups.

Generalised Anxiety Disorder-7

The GAD-756 is a seven-item self-report measure of anxiety. The scale has been found to be reliable and valid in the typically developing population56 (Cronbach’s α = 0.92). However, its psychometric properties for use with autistic individuals are not known.

Obsessive–Compulsive Inventory-Revised

The Obsessive–Compulsive Inventory-Revised (OCI-R)57 is an 18-item self-report measure of the symptoms of obsessive–compulsive disorder, with items such as ‘I feel I have to repeat certain numbers’ scored on a five-point Likert scale (0–4), indicating increasing frequency. The scale has been found to be reliable and valid. 57 The OCI-R was found to have good psychometric properties in a sample of 225 autistic people. 58

Positive and Negative Affect Schedule

The PANAS59 is a 20-item self-report scale of positive and negative affect. The scale has been found to be reliable, with a Cronbach’s alpha of 0.89 for the positive affect scale and of 0.85 for the negative affect scale, but it has not been used with autistic individuals. 59

Work and Social Adjustment Scale

The Work and Social Adjustment Scale (WSAS)60 is a five-item self-report measure of impaired functioning that has been found to be reliable (Cronbach’s α = 0.7–0.94) and valid in typically developing individuals. It has been used with autism populations, but its psychometric properties have not been assessed.

EuroQol-5 Dimensions, five-level version

The EuroQol-5 Dimensions, five-level version (EQ-5D-5L)61 is a five-item self-report measure of health, with items measured on a 5-point Likert scale (1 to 5) indicating severity. It has been found to be reliable and valid in the typically developing population. 62

Short Form questionnaire-12 items

The Short Form questionnaire-12 items (SF-12)63 is a 12-item self-report measure of physical and mental health. It has been found to be a reliable and valid measure among people with severe mental health problems, but psychometric properties for autistic populations are not available. 64

Adult Repetitive Behaviour Questionnaire-2

The Adult Repetitive Behaviour Questionnaire-2 (RBQ-2A)65 is a 20-item self-report measure of repetitive behaviours. It has been found to have good internal consistency, with Cronbach alphas of between 0.73 and 0.83, and autistic people score significantly higher on the measure than typically developing individuals, suggesting that it is a valid measure of autism-specific repetitive behaviours.

Rumination–Reflection Questionnaire

The Rumination–Reflection Questionnaire (RRQ)66 is a 12-item self-report questionnaire comprising two subscales: rumination and reflection. The measure has good psychometric properties; the rumination scale has a Cronbach’s alpha of 0.90, and the reflection scale has a Cronbach’s alpha of 0.91.

Economic evaluation

We aimed to assess the feasibility of collecting data on statutory health and voluntary service use. This was assessed with a questionnaire collecting information on the use of other primary and community care services (NHS Direct, attendances at walk-in centres, use of community health-care services); secondary care related to mental health (number of outpatient visits, type of clinic and reason for visit, inpatient stays, length of stay and reason for stay); use of social services and disability payments received; personal costs related to mental health (expenditure on over-the-counter medication, expenditure on prescriptions, travel costs associated with health-care visits, loss of earnings, out-of-pocket expenditure on other services, e.g. private counselling or complementary or alternative therapies, child care and domestic help); time off work; and unpaid activities. We also accessed GP records to obtain information on the number of primary care consultations, by type (e.g. face to face, telephone), who was seen and prescribed medication.

Inter-rater reliability of the Hamilton Depression Rating Scale

The HAM-D is an observer-rated instrument, with seven assessors interviewing participants. During the study, it was important to establish inter-rater reliability by using two assessors to re-rate these interviews. It was also important for subsequent analyses to compare the sensitivity to change of the two self-reported depression measures (BDI-II and PHQ-9), with which the HAM-D was to be used as the gold-standard measure for comparison.

We assessed inter-rater reliability early in the study (June 2017); a second assessor independently rated the audio-recordings of the first six interviews completed by the initial assessor in Newcastle and the first eight interviews completed by the initial assessor in Bristol. Agreement between assessors by HAM-D item (excluding the two items that required direct observation, i.e. psychomotor agitation and psychomotor retardation) was calculated as a percentage of total number of agreements on ratings of individual items.

By the end of the study, the first eight HAM-D assessments for each of the seven assessors (three in Bristol and four in Newcastle) had been double rated and the inter-rater reliability of the HAM-D was assessed. As most of the second ratings were conducted by listening to an audio-recording of the interview with the participant, the two items of the HAM-D that require direct observation (psychomotor agitation and psychomotor retardation) were not included in the total score. Therefore, the HAM-D scores being compared were the total of 15 items in the HAM-D without prorating due to the established variability across individual items. 69 Cohen’s kappa was calculated for every pair of assessors. Agreement was defined as both of the assessors’ total score falling within the same three-point range of HAM-D scores, which has been suggested to be clinically relevant. 28 A value of Cohen’s kappa of ≥ 0.8 was prespecified as acceptable.

Sensitivity to change of depression outcome measures

We explored the sensitivity to change of the two self-reported measures of depression (PHQ-9 and BDI-II) in comparison with the observer-rated HAM-D (clinician-administered interview) to identify the most appropriate outcome measure(s) for the main trial. We also identified the time point that would be used to measure the primary outcome in the main trial (based on the time point at which most participants had completed the intervention).

For each of the self-reported depression outcome measures (PHQ-9 and BDI-II), we examined:

1. acceptability (proportion of individuals providing sufficient data)

2. agreement between continuous scores using the intracluster correlation coefficient (ICC) from a two-way mixed effect, repeated measures analysis of variance model (observations are random, outcome measure instrument is fixed)

3. Bland–Altman plots

4. sensitivity to change (defined as a binary outcome of at least 50% improvement in symptoms on HAM-D at the time measuring the primary outcome compared with baseline) using receiver operating characteristic (ROC) curve analyses separately for each measure.

We also compared the continuous scores on the depression outcome measures only (HAM-D, PHQ-9, BDI-II) between the groups.

Economic evaluation

The focus of this feasibility study was piloting the methods of economic data collection rather than comparing costs. Participants were asked to report details of their use of health services and voluntary services during the trial as part of the follow-up questionnaires at 10, 16 and 24 weeks. Data collected included medication use, use of primary care and secondary care [including outpatient appointments, inpatient stays and attendances at accident and emergency (A&E) for their mental health] and the use of autism support services. Participants were asked to recall their use of these services during the interval since the last contact (e.g. baseline to 10 weeks, 10–16 weeks and 16–24 weeks). In addition, data on medication and use of primary and secondary care services were collected from electronic medical records (EMRs) at the end of the study to inform the decision regarding the source of data for a larger-scale trial. These included data from 2 months before the participant entered the study until the date the participant ended the study, namely 24 weeks after randomisation.

The data on health and voluntary service use were described, using appropriate statistics, to inform which economic data were to be collected in the main trial. These included examining completeness of data; determining which source of data should be used in the main trial (when information from both self-report and medical records is available – antidepressant medication use, use of primary care services and NHS outpatient or community mental health team clinics for mental health problems); identifying any redundant questions; and identifying any additional questions required to enable costing in the main trial.

To analyse the self-reported data over the whole study period, each participant was recorded as using each health-care or voluntary service (outcome) if they had done so at least once over the three assessment points detailed above, even if they additionally recorded a negative or missing response at another assessment point. If a participant responded ‘no’ at all assessment points, or ‘no’ at some assessments and missing at others, they were recorded as not having that outcome.

Concordance between self-reported data and the EMR data was defined as the outcome being recorded as both yes (or both no) in the self-reported data and the EMR data, and data were regarded as missing if that outcome was not recorded. These analyses were possible only for the outcomes of antidepressant medication use, use of primary care services and NHS outpatient or community mental health team clinics for mental health problems.

Post hoc analysis

During the study an amendment was made to collect a Participant Global Rating of Change in depression (May 2017). This question was asked at the 16- and 24-week follow-up to find out how the participant felt their depression was compared with baseline (i.e. 16 or 24 weeks earlier).

Ten weeks

The mean scores on most of the outcome measures (PHQ-9, BDI-II, OCI-R, PANAS negative, GAD-7) were higher (worse outcome) or lower (PANAS positive, SF-12 physical function; worse outcome) for those in the TAU group than for those in the GSH group (Table 4). Across most measures, around 15% more participants responded in the GSH group than in the TAU group. The mean scores on the HAM-D, EQ-5D-5L and RRQ measures showed little difference between the groups.

Sixteen weeks

By 16 weeks post randomisation, the mean scores for several outcome measures (PHQ-9, BDI-II, OCI-R, HAM-D, PANAS negative and GAD-7) in the GSH treatment group had further decreased (lower score, better outcome) or had increased (SF-12 physical function; higher score, better outcome) compared with scores at 10 weeks (Table 5). The mean scores of the outcome measures in the TAU group had also changed (compared with scores at 10 weeks), but not by as much, and the score changes were more variable and not consistently in the same direction. Differences between the group mean scores on HAM-D and quality of life (as measured with the EQ-5D-5L) presented as slightly larger at 16 weeks than at 10 weeks. The trend of increases in scores from baseline continued, with both groups showing better outcomes across all measures, except SF-12 normalised physical function for those in the GSH treatment group. There was, however, a large difference between the groups in the proportion of participants who were followed up and completed the measures (22% more participants were followed up in the GSH group); therefore, the findings should be interpreted with caution given the potential for bias.

Twenty-four weeks

By 24 weeks (Table 6) the mean scores for several outcome measures (PHQ-9, BDI-II, HAM-D, OCI-R, PANAS negative, GAD-7, WSAS) had increased (higher score, worse outcome) or decreased (SF-12 physical function; lower score, worse outcome) in both groups compared with 16 weeks. SF-12 physical function and quality of life (EQ-5D-5L) had increased (higher score, better outcome) in the GSH group. The trend from baseline continued, with both groups showing better outcomes across all measures, except for the PANAS positive score for those in the TAU treatment group and the SF-12 normalised mental health for those in the GSH treatment group. However, as previously, these findings should be interpreted with caution given the large differences in the proportion of participants who completed the measures between the groups, with 18% more participants in the GSH group than in the TAU group.

Aim

To understand trial participants’ and GSH coaches’ views and experiences of the trial and of GSH or TAU.

Objectives

• To explore, using qualitative methods, trial participants’ views, experiences and beliefs about therapy for depression.

• To examine trial participants’ understanding and knowledge of GSH and their treatment expectations.

• To explore the feasibility and acceptability of the trial design.

• To understand trial participants’ and coaches’ experiences of the trial, including their perceptions, experience, opinions and acceptability of GSH and TAU.

• To use qualitative methods to understand the barriers to and facilitators of GSH to enable intervention to be refined for a future effectiveness trial.

Study design

In-depth interviews were conducted with participants (from both arms of the trial) 10 weeks after randomisation (after the 10-week outcome measurement was complete). These interviews considered and compared views and experiences of the trial and the acceptability of the GSH intervention. All coaches delivering the GSH intervention were approached for an interview towards the end of the trial to illuminate the perceived effectiveness and acceptability of GSH and explore any barriers to its uptake.

Recruitment and sampling

Purposive sampling was used to select trial participant interviewees in an attempt to capture maximum variation in views and experiences in order that they adequately reflected those of a range of trial participants. All trial participants had consented at baseline to being approached by a qualitative researcher. From participants who had indicated that they would be willing to be contacted, a purposive sample was drawn in relation to (1) the trial site, (2) the trial arm and (3) sociodemographic variables such as age, gender, ethnicity and socioeconomic status (with participants selected from areas of high and low socioeconomic deprivation). Interviews were analysed in batches, and sampling continued until no new themes emerged from the interviews.

Interview conduct

All interviews were conducted over the telephone or face to face in a location of the participant’s choice. Interviews followed topic guides developed at the start of the study based on the literature, the qualitative study aims and input from the Trial Management Group (TMG). The topic guide was used to ensure that primary issues were covered during all interviews, but without dictating data collection, allowing participants to introduce unanticipated issues. All interviews were conducted by two clinical psychologists in training (Hannah Harvey and Lauren Evans) under the supervision of a senior qualitative researcher (Jeremy Horwood). Neither interviewer had been previously known to participants. Interviews lasted between 26 and 65 minutes (mean 47 minutes).

In telephone interviews, consent was taken verbally prior to the interview. In face-to-face interviews, participants gave written consent. Field notes were taken during and after the interviews and these were integrated into the analysis. All interviews were digitally recorded, professionally transcribed verbatim, anonymised and checked for accuracy prior to analysis. Transcripts and field notes were imported into NVivo version 10 software (QSR International, Warrington, UK) for analysis. 71

Participants

Five therapist ‘coaches’ were interviewed: three from the Bristol site and two from the Newcastle site. All coaches interviewed were assistant psychologists, except one who was a clinical psychologist in training. Two coaches had not previously worked with adults with ASD. Therapists had conducted the GSH intervention with between 1 and 13 trial participants (mean seven trial participants).

Twenty-one trial participants were interviewed, 14 randomised to the GSH arm and seven to the TAU arm (age range 21–58 years, mean 40 years) (Table 11). All trial participants interviewed were white British. Eighteen participants had received an ASD diagnosis in the previous 6 years and nine had received it in the previous year. Eighteen trial participants had previously engaged in therapy for anxiety and/or depression, with 10 previously receiving CBT. All initials of participants refer to pseudonyms, with trial participants being Mr or Ms A–U and coaches C1–C5.

Previous therapy

The majority of participants (19/21) had previously attended psychological therapy before participating in ADEPT, with CBT being the most frequent model of treatment. Two participants had also previously engaged in counselling and another had participated in eye movement desensitisation and reprocessing and positive psychology in addition to CBT. One participant had experience of psychotherapy and another had tried computerised CBT. The duration and amount of previous therapy varied. Some participants had attended three sessions of treatment while others had been engaged for up to 2 years. The time frame also varied, with some participants engaging in other therapies shortly before entering the ADEPT trial whereas others had not done so for > 10 years.

Two participants talked positively of their previous experiences of CBT, with one participant highlighting the positive impact it had on their mood and another describing how they still found some of the strategies helpful. However, some participants did not find their previous experiences of CBT helpful. The reasons for this included finding it difficult to describe their feelings and to identify triggering situations, difficulties engaging because of other factors in their life, the treatment involving not many practical aspects, and being asked to do things that they did not feel ready for. When compared with previous therapy GSH participants appreciated taking part in treatment that was tailored to people with ASD:

But the CBT you’re weighing up your thinking and your beliefs and your prophecies almost [inaudible 0:04:22.8]. And then the action planning and testing. That I wasn’t very good at and I’m still not very good at but the way in what that . . . that’s something I still use.

Ms F

Just someone who maybe understood a little bit about how I can understand and communicate and things like that. I found it [previous CBT] hard to get any perspective and so I was unable to communicate exactly what was going on. I would go into sessions and just wail for an hour and so yes, this [ADEPT] was new, to have someone who understood and even if I wasn’t feeling up to it I was able to come and just sit and eventually I’d come out of me shell a little bit more whereas when I was accessing services before I was probably masking still. It was nice to be able to engage with something and actually feel like it was doing me good.

Ms G

The last time I did CBT it was useful up to a certain point. It didn’t really seem to consist of many practical exercises. It seemed to consist mainly on reciting information back to myself and it was less focused on actually going out and then performing different kinds of behaviour . . . It got to a point where I was being told to basically go to a public place of some kind and just start talking to strangers. That wasn’t something that I was particularly comfortable with doing and I didn’t think that that was going to solve any particular problem.

Mr E

Since being diagnosed I’ve just been trying to find some help that’s a bit more autism specific because some of the stuff that I’d had before wasn’t necessarily – it just didn’t really feel like it was working.

Mr C

Trial understanding

The majority of participants said that they were given the right amount of information about the trial and that this information was understandable. One participant commented that it would have been helpful to have additional information illustrating people with ASD’s previous experiences of CBT as a way of explaining the rationale for the adapted GSH. On the other hand, one participant felt that there could have been slightly less information given in the information sheet as it replicated what she had already been told about the trial by a researcher. One participant commented that it was helpful to have someone to go through the information with them to aid their understanding:

There would be two groups. Anyone taking part would be in one of the two groups: a control group, what is the standard therapy and the more adapted therapy for people with autism.

Mr D

I grasped it fully. My partner did too. It was easy to register my interest and easy to get onto the trial. The information sheet was fab. It was everything you needed really.

Ms G

Just how many sessions, what roughly it would be like and then obviously the follow-up sessions as well, so I knew exactly what it was going to be, because if it had been vague and that, it would be hard to sort of gauge what it would be like.

Mr B

I suppose it might have been interesting to have had some information on other people’s reactions to CBT, other people with something like Asperger’s syndrome: how they had reacted to CBT and what their experiences had been, if they’d had CBT in the past.

Mr E

My social worker explained to me really well. This was my learning thing you know, I don’t always process everything and I read information or when things are being said to me I have to have it repeated and if it’s too complex . . . you know she did explain it fully to me so I knew everything that was going on.

Ms I

The majority of trial participants had a clear understanding of what taking part in the trial involved, although there was some uncertainty around the content of the GSH therapy:

Attending sessions or telephone conversations on a regular basis over a period of time a minimum of like 10 weeks plus the follow-up. You know, it was all fairly well.

Mr C

I wasn’t too sure really, but having done CBT before a long time ago I thought it might be something not too dissimilar to that, possibly with some new elements which I hadn’t encountered before.

Mr E

Well I wasn’t sure because I think she said there was elements of CBT but that there were also differences to make it more accessible and then also maybe some other techniques brought into it as well. So I didn’t really have any set idea in mind.

Ms F

I think there was some explanation. It was very clear to me I was getting involved in a talking therapy where I’d have some sessions with another person and I went in a room with no idea about what that would be and I was fine with that. That’s perfectly reasonable.

Mr Q

Three participants reflected that they had felt nervous about taking part in the trial as they had not known what to expect, they had worried about having to try another treatment aimed at improving their mood, or they had difficulties trusting other people:

Yeah, a little bit nervous to start with . . . As to what is was going to entail. What was going to be needed of me. What was going to be asked of me. How personal it was going to be and not knowing I didn’t think initially as to whether it was going to be many people and whether one to one because although I come across as quite good in groups, I don’t really like group work. I much more prefer sort of one to one and am more comfortable.

Mr C

Well as with everything I was quite . . . I suppose you could say anxious or stressed about having to do something else to help my mood, help me get better or get back to where I used to be. But then when I thought about it and it did say as you say, I could stop doing it at any time I wanted or whatever, so. I was . . . basically I was just thinking about well what’s the next step and then get onto that step and then I can ask the questions and then, you know, like that sort of thing, take one step at a time, or one meeting at a time sort of thing.

Mr K

The majority of trial participants commented that they thought that the trial was evaluating whether or not a newly adapted intervention is effective for low mood in people with ASD, and that this was being compared with standard NHS therapy:

By my understanding it was to see if a different form of therapy could be adapted for people with autism as there’s quite a big group of people with autism where standard therapy doesn’t really help with so this was more of a tailored therapy for people with autism.

Mr D

To see if certain adaptations could be made to an exact mix of CBT to see if it could be made to be more effective for people with an autistic spectrum condition.

Mr E

One participant thought that the trial was comparing experiences of depression in people with ASD and those of people who do not have ASD, and another thought that the trial was comparing self-help with CBT:

The difference between a person . . . an ordinary person who doesn’t have autism, how depression can differ with someone like that to a person with depression that’s got autism and if there’s any differences in the experiences.

Ms I

Difference between self-help and CBT. Trying to understand the difference, which one would be the best of the two.

Mr R

All but three participants knew that participants were randomly allocated to either GSH or TAU at random. Two participants said that they did not know how participants were allocated. One participant thought that group allocation depended on their initial answers to the questionnaires:

I understood it was randomised and I also understand why you need to have that controlled group in there.

Mr B

My understanding was that a lady called [name] who I saw first off before I did this went and pressed a button on a computer and it randomly told her.

Mr Q

It was explained that my information would be put into a computer program, which would decide on what group to put me into.

Mr D

Well basically we had a questionnaire or something to fill in or questions were asked and it was from that so I gather but I don’t know what the criteria would have been for deciding.

Mr K

Some participants questioned how a comparison could be made between the GSH and TAU groups because those in the TAU groups potentially had to wait a long time to start their treatment:

I think the only thing that I ever questioned and that was with somebody that wasn’t well or was aware of what study I was on, but I thought, I think I mentioned it to [ADEPT researcher] before we’d even got to the point of me being not part of the study where I just said, well if you’re starting the 10 weeks as soon as you kind of can for guided self-help, then it’s always it’s kind of making the study a bit difficult to kind of work out what’s helping and what’s not, because there’s such a delay. If you’d have had like a certain amount of people like psychologists or whatever there, ready to do both sides of the study then it would be a little bit fairer.

Mr J

In a way one thing I don’t know how you’re gonna get an equal comparison in the trial, you’ve got people who have basically fast tracked into it rather than have to sit around and wait for a year or two. The treatment-as-usual people probably won’t have been treated. I honestly don’t know how you’re going to evaluate the trial on that basis. Perhaps something will actually happen as distinct from it happening sometime never like things in the health service have a way of doing. But that’s not a criticism of the health service as you realise.

Mr Q

Well they could have waited until I’d had my CBT before comparing this thing ‘cause I hadn’t had any treatment. I’ve never been treated until the 20th October, no 13th, start next week. So we only receive the last bit – my last one – there’ll be any improvement. You’ll see the first bit, but don’t know if it’s improved through the process and drop back off again.

Mr R

My first appointment was only a few weeks ago, I think it was just after the 10 weeks when I got the phone call to organise the appointment.

Mr J

Trial participation reasons

Most trial participants said that they had taken part in the trial because they were hoping that the GSH intervention would lead to improvements in their mood. Six participants commented that they wanted to contribute to research that may be helpful for other people with similar difficulties:

First of all I wanted to do it because there were certain problems that I was having and also it seemed like it would be a useful thing to do. It might help other people having similar problems.

Mr E

It feels like I’m getting something and contributing something all at the same time. Because I’m autistic I spend a lot of time – on my particular presentation of autism – I spend a lot of time feeling uninvolved and disconnected from stuff. Also somebody wanting to be involved in something that’s going to be useful seems like a good idea.

Mr Q

That’s a really good question. I would like to think, being quite interested in research anyway and things like this and I thought it was a good way to kind of make a contribution and also kind of face the diagnosis as well and maybe learn a little more about myself. It wasn’t altogether altruistic because I was hopeful that it may lead to some beneficial treatment for me as well.

Mr U

Three participants said that they had taken part in the trial to have the opportunity to try a new intervention that was different from what they had experienced previously and that focused on the needs of people with ASD:

So I was really interested in taking part, because I had a tiny bit of experience of talking therapy before and I had only just started some CBT, just prior to hearing about the research, so I was very interested in trying something that was geared more towards people with autism.

Mr B

Really hoping that a new form of therapy would help me more than the standard therapy, what hasn’t really helped me much so that was what I was trying to get out of it. Even if I wasn’t put on the new therapy group it would possibly lead to a new therapy being adapted and implemented so that I would in the future go on it.

Mr D

To see if there was anything that would help me and the idea of it being tailored to autism. I was interested to see what was different about it from the therapies I’d already had.

Ms F

When deciding whether or not to take part in the trial, there was variation in whether or not participants discussed their decision with others. Those who had not discussed the trial with others found the study information clear and the decision to take part relatively simple:

I did the research on it before accepting it and it seems all pretty straightforward so I just did it.

Mr D

Eight participants interviewed did discuss their decision to participate in the trial with others. Seven participants discussed their decision with a family member and one discussed it with their social worker. Two participants wanted reassurance from others that it was worth taking part in the trial, whereas one participant wanted to clarify that he was appropriate for the trial. One participant said that he could not understand the study information and so he asked his father:

Although I knew it in myself that I wasn’t happy with just everything in life at the moment, I didn’t know whether I was classed or going to be classed as low mood and so on and so forth. So I just wanted to double check with my wife as to whether she thought that how I was feeling at the time, so yeah so . . . Yeah, yeah just to make sure it was worthwhile and that I wasn’t wasting your time and stuff.

Mr C

I couldn’t understand most of it so I just asked me dad and he just explained it. He says it will be good for you to get involved with it.

Mr L

Preference for guided self-help

In terms of group preference, all but three trial participants interviewed said that they had wanted to be allocated to the GSH group. The other three participants said that they had not had a preference for a group. Reasons for wanting to be in the GSH group included to see whether or not GSH was effective; to try a new treatment as they had already tried other forms of therapy, such as CBT, and not found them helpful; to have access to a treatment that took into account the needs of people with ASD; and because they would find it difficult to explain their difficulties to their GP to be able gain access to treatment. One participant commented that his allocation to the TAU group (and his previous negative experience of mainstream services) was the reason he withdrew from the study:

I certainly preferred to be in the group that I was in because I’d done regular CBT before . . . I thought it would be a good idea, especially considering that I’d tried CBT in the past and not found it all that helpful, which I thought may have been to do with the fact that I had Asperger’s syndrome. I hadn’t been diagnosed with Asperger’s syndrome when I’d first done CBT so I thought that that may have been part of the reason why it wasn’t so effective. I thought it would probably be a good opportunity. It would be a way of helping basically.

Mr E

I think, if I hadn’t been in the group I’ve been in, I think I would have perhaps dropped out because what I had previously been given wasn’t helping and wasn’t altering anything.

Mr M

I’d previously had the other service and that wasn’t helpful so it would be back to square one so I was kind of hoping and fortunately for me, it’s worked out that way and it was the new stuff.

Mr P

Yes I did rather hope I would end up in the group with the new treatment which is where I did end up. I suppose that was partly ‘cause otherwise I’d then go and have to have a discussion with the GP and explain what I’m doing. I don’t like explaining things to people.

Mr Q

It’s a two-to-one chance that it’s going to work or that I’m going to have to withdraw. I remember saying that very, very specifically, not that I might get in or I might not. It was literally I either will or I will withdraw, so pretty much as soon as I’d seen how the sampling was going to work, it was kind of ‘Oh yeah, oh this is going to be a one or the other’ . . . The allocation didn’t come until quite late in proceedings and it wasn’t until then that I realised ‘Well, I actually might end up in a group here that is where I was 2 years ago and I found of no benefit whatsoever’ and then that’s pretty much what happened. I ended up in that group and thinking ‘Well, this is gonna entail 10 days of leave from work so something that I know doesn’t work for me’ so I just couldn’t. A terrible, you know, concrete way of looking at things, but that literally was it. ‘This is gonna be of no benefit and whilst I’d love to help the study, I can’t take 10 days of leave to do so.’

Mr U

Acceptability of the trial

The majority of participants said that they found the trial acceptable and in line with what they had expected. Two participants commented on how they did not feel like they were taking part in a trial, as trial participation mainly involved them seeing a coach:

Not too bad, I guess. I was in the control group so I didn’t really have much involvement in the ADEPT study really, just the scoring of how it helped.

Mr U

Yeah I suppose so. That part was fairly easy enough because everything was explained to me every time I went sort of thing, so it’s not as though I’m thinking about things and thinking, well they’re not telling me this or not telling me that. I found it quite helpful.

Mr K

It hasn’t really necessarily felt like taking part in a trial simply because it’s just involved me seeing a CBT therapist. There’s not been any problem with it. It’s been fine.

Mr E

I’ve found it quite interesting actually. It didn’t mostly feel like a trial. It was presented by a person doing – and accepted by me – as a treatment. The fact that it was trial. I mean what we would very occasionally refer to that. But it kind of didn’t feel like that so I think it’s as it should be.

Mr Q

Five participants had difficulties with the location of their appointments, which related to difficulties parking or finding the right building. These participants would have liked to have been given directions and more information on where to park:

I knew I was coming to [name of building] House but I didn’t know what the buildings were called. I don’t think there was a map. I knew to go in and just ask and I did but I felt like an idiot and then you feel like you’re on a wild goose chase and have I got the right place? Just explain that there’s a number of buildings, you’ve got the [name of company] bit or whatever, just a little bit more of an understanding. It’s good to look on Google maps [Google Inc., Mountain View, CA, USA] and stuff but just to know that you can get parked or that there is an issue with spaces, come and give plenty of time or find somewhere nearby that you could . . . I think, on one of the occasions, I rang my practitioner to say, ‘I can’t get parked. I don’t know what to do,’ and she said, ‘well try over the road in the school, you can get parked in there,’ and I didn’t know that and I’d been coming for weeks and so just knowing that would [over speaking 0:37:59.4] have been helpful.

Ms G

I think maybe the location wasn’t ideal, but the only reason I had to fight through traffic every morning. I used to get stressed. I was quite calm when I left the house but getting to [XXXX] and then trying to find a car park really got me stressed, so obviously when I walked in there I was trying to calm down.

Mr M

One participant said that because timing of the appointments was such that they did not interfere with his work, this made participating in the trial easier:

It hasn’t impact on my work that much either. I was a bit worried in the beginning that I would have to take days off and holidays, but I think I got appointments more or less half nine in the morning, quarter past nine, which meant I could go back to work without having to take a day’s holiday, but I think I would have struggled if it had have been nine to five hours and I couldn’t have the flexibility in my job. If it had have been evenings, that would have been a bit better, but yeah, I think that’s what kept me going as well. If it had have started impacting on my work, I would have had to stop.

Mr M

Situations map

Three participants reported that they had found the situations map particularly helpful. Most participants said that they had used the map to document where they had been over the previous week and how often they had been there. The majority reported that using the map had increased their awareness of their behaviour and level of activity during the week, highlighting that this was not something they would normally focus on. Several participants commented that using the map also focused their mind on the impact of changes in routine, or exceptions to their normal routine, on their mood:

A visual aid like that [situations map] I think is a very useful tool I was able to use. I was able to explain it, the way I’d written . . . drawn it.

Mr K

And the other little map as well was good, to see, even when I was thinking I wasn’t doing a lot, when you write it down in a week and you can show how many times you have been to this certain place and looking at that guide, you sort of think, ‘oh yeah, I’ve actually done more than I thought I had done’.

Mr B

It just made me think a bit more about the routines and patterns that I tend to stick to because those aren’t really things that I would consciously think about otherwise. It made it easier to focus on making changes to my behaviour and things.

Mr E

Several participants created their own maps using pen and paper or adapted the copy from the materials. However, another participant commented that the layout and size of the map in the materials pack made it difficult to use as the layout of locations did not resemble their own life. They suggested that providing sticky notes for each image that an individual wanted to include on their map so that they could arrange it themselves would also support personalisation:

They’re not great maps, don’t get me wrong. I just found that easier to do my own and also because I do work and I don’t know . . . because I’m involved in the church as well I brought quite a few different locations each week so I just found it easier to do a new one each week and it helped me remember what I’d been doing then.

Ms F

The map was OK but I found myself having to reorder everything and rename everything because it didn’t fit with what I did. Most of my stuff was in the house so I had to try to expand the picture of the house to try and add more things into it of what I did.

Mr S

Just the size and the shape of it [situations map] and the actual images are very kind of, what’s the word, I would say basic but I found it hard to use because it was all very close together and in my personal life the house isn’t as close to this as to that and so forth so I did try and I was going to draw my own map, which was fine and I went home after this session and I tried drawing out but I actually used an actual map to locate and put it and work positions as to where I was working, my daughter’s college, the schools, the shops and whatever. But then it was all too small so I kind of got annoyed with myself and then we just kind of just sort of just – well whatever, and just used that particular printed copy and stuck some triangles on it.

Mr C

Situation triangles/granularity

The majority of participants commented that they had found the situation triangles helpful. The majority stated that the most helpful thing about the situation triangles had been breaking situations down, focusing on the link between feelings and behaviour. One participant said that writing down particular situations had helped them to externalise their feelings:

But yeah, definitely breaking it down and just looking at the things I was doing a bit more closely, was a big help.

Mr B

Probably doing triangles. I hate to admit it, but it’s probably true . . . working out what you could have done differently.

Mr O

So straight away I was able to spot that I was listening to music and they were the moments that my mood was picking up and that I was able to say, ‘oh well, that activity made me feel better as well just like the music did’. So it really helped me put the theory into practice.

Ms G

A minority of participants specifically mentioned the concept of granularity (the process of chaining several situation triangles). They reported that this helped them to focus on a whole situation rather than on one negative moment. One commented that introducing the triangles slowly, focusing on one aspect and then adding another, had supported their understanding and use of the technique. Several participants commented that they had applied the concepts they had learnt in sessions regarding the situation triangles to other situations in their lives (e.g. to understand more negative situations and to approach situations they had encountered after sessions):

The breaking it down like that it meant that I was looking at experiences rather than . . . I was looking at the whole thing and finding the many benefits rather than just the bit that I hadn’t liked.

Ms F

It’s just, because there’s all this triangulation of you’re doing your little triangles and stuff and all that and situations, behavioural stuff and that and trying to, trying to not I mean because say I’d go in, I’d had a week where I’d go in and I’d say I’ve had a terrible week, this has happened, but then I’d talk to my therapist and she’d say, ‘well, OK, let’s talk about the rest of the week’ and it transpired that one thing had gone wrong, but the rest of the week had been OK and I’d just focused on the negative thing and ignored all of the positives and she made me think, ‘no, you’ve got to try and not think about that, think about all the other things’, and I did, I have been doing that a bit more.

Mr N

I think just breaking down a positive event into where it enters the smaller bits and thinking about it because I noticed that, once I did this positive thing and how all these little things made it a positive thing, how I then went on and did other positive things.

Ms T

Feelings chart

Participants commented on the range of methods offered with which they could rate their mood – images, colours, numbers and words – and that they could choose which method suited their individual needs. The majority of participants valued the feelings chart. However, a minority of participants commented that they had found it difficult to decide where to rate their mood when asked and wondered whether or not this was down to their ASD. One participant said that they found the feelings chart difficult to use as they found that their feelings were difficult to explain and they wanted to focus on their thoughts rather than their feelings:

I think we ended up using words with that so yeah on a number of [inaudible 0:26:36.6] could use words or number or whatever and whichever may be more comfortable. For me personally using words. Yeah that worked for me. Yeah. In the mood chart, yeah. I think that was fine you know if you could individualise it or personalise it.

Mr C

Yes, well we filled that out and then it was like nought to ten and stuff and I filled, I filled that out with you know all of the relevant – we put our own words on to like and then when I was in my triangles and that I put the numbers down and referred back to the slots quite – that was quite useful actually.

Mr N

But I think, as soon as you have different grades of something that’s when it gets a bit tricky. I am sure it’s not just me, other people with autism have problems where there’s a lot of choice. I think if it’s binary or a very small choice, I think sometimes it’s a lot easier. Yeah, but on the spectrum where we have trouble picking things from the spectrum.

Mr B

Maslow’s hierarchy of needs

The majority of participants had a positive view of the inclusion of Maslow’s hierarchy of needs in the treatment protocol. Many described that they had found its inclusion helpful because it had provided them with a different perspective on aspects of their life that might influence their feelings and behaviour. One participant said that, although she had found discussing the hierarchy useful, observing gaps at the base of the triangle (e.g. financial security) had been a powerful but difficult realisation:

Yes, the hierarchy of needs was useful. That was useful because it makes sense as a concept but if I’m doing something I wouldn’t normally think about which part of that hierarchy of needs it’s relating to. I think, again, that’s something that helped me to think about what I’m doing in a different way, in a more focused way.

Mr E

Yes, helpful in terms of hierarchy of needs one. Yes, I found that useful . . . I just found it – what was helpful was just to sort of organise needs better and think more. I think that was part of the thinking in terms of planning events to meet my needs rather than just lurching from one minute to the next.

Mr P

It put things into perspective for me in the sense of showing me how important the higher up things are, how they’re not just extras and add-ons but they’re just as important but everything underneath it needs to happen as well for those things to happen.

Ms T

Looking at it realised I’m basically going to stay anxious and depressed until I’m financially secure and there’s no end in sight for that happening so I left feeling really, really upset that day. That’s no reflection on the study . . . no reflection at all just the realisation about what I need to prioritise because I think they do a lot of stuff that’s towards the top of the triangle without underpinning it with this level of personal security and that’s where my life sort of seems to be going wrong and so it hit me pretty hard realising that.

Ms F

Views and experiences of guided self-help between-session tasks

Most participants had a positive view of the content of the homework tasks. Participants said that they felt the homework reinforced in-session content and that this helped them to complete homework tasks. The majority of participants commented that homework had been helpful, with some commenting that homework tasks had prompted them to try out new activities and notice the impact of them on their mood. The majority of participants commented that the amount of homework set was acceptable and presented a sufficient level of challenge. Furthermore, two participants stated that the homework tasks felt more manageable than those set in previous therapeutic experiences:

I think the activities were not complex ones and, to be honest, I don’t think it wasn’t independent either, but in the main what I was being asked to do was achievable both physically and financially so I think realistically it was good. There are a lot of possibilities as well. If you don’t really want to do something, do something different.

Mr M

I guess they were things that I would have wanted to be doing regardless. It wasn’t like I just came and picked stuff out of the blue. It was things that I had going on in my life and used this to reinforce them so it wasn’t ever like, ‘Oh, I’ve got to go and do my homework now.’ It was more like, ‘I’m doing this anyway. Now I can reflect on it.’

Ms T

A minority of participants wanted more information on what was required for some of the between-session tasks, including examples of what they could do to complete the tasks. Some participants stated that they found the initial homework task (which included noting down their behaviours in each of their visited locations) difficult to complete because they did not understand why they had been asked to do this. They suggested that completing an example in the session and knowing that behaviours would be linked with feelings in later sessions would have facilitated their understanding and increased their motivation to complete the task:

If you could make clear what the goal is for the task, why you’re doing a task and make sure that’s made more clear, that might help.

Mr K

I don’t know. I think one of the problems was trying to discover what tasks are supposed to be helpful, that makes you feel good, but I don’t know, I don’t quite know what the task was supposed to be. If it was something that was productive, you would feel proud after you did it, but I don’t think I can think of anything that I ever particularly wanted to do in that sense. That might be something personal for me that I found it difficult to do the tasks, but it might be something that other people feel as well.

Mr H

Well just some information about why we need to make a map of where you’ve been that week and why you need to make a map of where you’ve been and what you’ve been doing that week. Because actually it was really only that third week when you put it together with feelings that it makes sense. And obviously when things don’t make sense it does affect your motivation to do it.

Ms F

Most participants found the homework tasks manageable alongside their day-to-day commitments. However, for a few participants, finding time to complete homework tasks alongside other life demands (e.g. family commitments) and negative moods acted as a barrier to completing homework as planned. Participants talked about using tools such as a scheduling assistant to help them remember to complete homework tasks. One participant said that he used his mobile phone to note down where he had been to aid homework relating to the map:

I struggled to do them mainly. My mood varies so much and you’re supposed to schedule them for a certain time, and I found I had no way of knowing whether I’d be in a mood that was particularly good, in which case I felt, you know, I don’t necessarily need to do . . . or I did a task, but I was feeling OK at the time anyway, or if I was feeling bad, in which case I didn’t really feel like doing a task. That made it difficult to do them.

Mr H

Although I did put things on my phone, wherever I had been, so I just sort of made a note of it, so I could look back and see where I had been. How many times I had been to a certain place, so that did help.

Mr B

Most participants commented that they discussed their choice of homework activity with their coach beforehand. Some participants stated that they knew that they could discuss any difficulties in completing the tasks with their coach in their next session, and that they valued this support. One participant suggested that it would have been helpful if the coach had placed more emphasis on the purpose of the tasks:

The fact that it wasn’t too much. I could sit down across the week and just take a couple of minutes to sit and think – reflect – and then get that written down and I like the challenge that each week it was slightly different but not too different, not too new that I would have gone, ‘oh this is scary and I’m gonna do it wrong’. So it felt like it wasn’t too challenging and it wasn’t like you’re gonna get caught out or anything and it wouldn’t be the end of the world if I did get it wrong because my worker would help me with it so I think it felt a lot more approachable.

Ms G

You know it’s like you think how’s this going to help, how’s this going to help me, but then you did it and then you went to the next session and then you would talk through it and then you realised that some things you’d written that you didn’t think were important actually meant something, so you know you realised that it was useful.

Mr N

Guided self-help: sessional goals

The majority of participants noted that their primary goal for treatment was to improve their low mood. One participant reported changing his patterns of behaviour to achieve this. Most participants commented that their goals had been met by the end of treatment. However, two participants commented that not all of their goals had been met yet, whereas another participant did not feel that the treatment had helped him to view things differently:

Yes in the fact that in the first session we set out what goals I expected to take away or the goals obviously I wanted to meet by the end of the sessions. And then on the last session we just sort of looked over as to whether I had managed to do that. It definitely appeared to have been that that I’d taken away strategies to help myself to raise my mood. I was obviously not in a low mood as I had been prior to the study. So in that respect, yes, I took way what I’d expected in the goals, yeah.

Mr C

I think as time has gone by I’ve got a lot of better, a lot more positive, a bit more focus on life.

Mr M

I don’t think I really had a goal from it, I was just going to go with it and see how it worked. I guess I was hoping it might change the way I see things at the moment but I didn’t find it did that.

Mr H

A couple of participants commented on the way goals were reviewed throughout treatment. One participant felt that including a review at the start and end was helpful in noticing change, whereas another felt that it would have been helpful to review her progress towards her goals with the coach at every session:

Yeah. I think the goals should have really been referred to each week to some extent. Because actually you could have been looking at how the activities that we’re looking at how they impacted on the goals because actually I kept forgetting the goals. And certainly some of the goals I made more progress on than others because one of them was about I wanted to spend better time with my son but he’s not that keen since he’s 16 and smelly! [laugh] So he didn’t really play ball with that so really I maybe shouldn’t have chosen a goal that was so reliant on somebody else. But then again one of my goals was about finances so that ended up being very pertinent and made progress against that prompted by setting that goal.

Ms F

When discussing reviewing goals in sessions, one participant recalled:

I don’t think we did it every session but there were times where we did, I think more towards the end.

Do you think you did it enough?

It wasn’t – yeah, for me, because the thing is I think sometimes I can get a bit obsessive about things and it takes over whereas I think in a way the whole thing made life seem a bit better which made those things less daunting.

Views of guided self-help coaches

All participants expressed positive views about the coach. Some participants described positive aspects of the coach’s style, highlighting how the coach appeared relaxed, calm and interesting, which made it easy for participants to talk to them. When asked about the coach’s understanding of their autism, some participants described how the coach explained concepts in detail so that it was understandable for them, which was helpful. One participant commented that it took a few weeks for the coach to adapt their style to accommodate the participant’s level of autism. All participants apart from one said that they felt it was useful to have a coach present during the GSH to keep them focused, to check their understanding of key concepts and so that they could ask questions, if needed, to aid their understanding. The majority of participants valued the presence of the coach and commented that this helped them complete the course:

It was just very easy to talk with her and stuff so I always felt better knowing that I was going to the session or I was at the session. That was always very good. She obviously spoke through it clearly and precisely and sort of very easy to talk with. So all the strategies she’s presented to me are all very clear and made a lot of sense and were easy to follow.

Mr C

I do. I could tell that she was clearly reading the notes and reflecting on the sessions and just, ‘you said this,’ or ‘I know you talk about doing that’. She’d often ask me stuff that I’d said the week before . . . and that was really really lovely. It shows that she actually was quite interested and that we got on and I really, really appreciated that.

Ms G

I felt like he understood because sometimes the reading material – there would be quite a lot of stuff that was overly detailed and he’d be like, ‘oh, I can see that you understand that. We can just jump ahead.’ And I’d appreciate that, that he didn’t have to bore me with stuff that I already understood. I felt like he understood where I was at.

Ms T

Like I feel, it’s a hard one to explain, but it feels like, if I go to a class or I see a coach or a trainer or something like that, it feels like I’m properly doing it and it feels more like I have a right to do it, if that makes sense?

Mr K

Because if you had the material and you maybe got a bit stuck, and maybe started to not see what the purpose was. I think if you didn’t have that contact and that regular contact, it might just fall by the wayside, you might just give up.

Mr B

Acceptability and experience of the trial

All coaches were enthusiastic about the trial because they acknowledged the current lack of NHS mental health services tailored to the needs of people with ASD diagnoses:

I wasn’t really surprised that there wasn’t a particular like autism services in mental health. So when I did read the job advert and I saw that this is specially designed to meet those needs I actually really engaged with it.

C1

I have an awareness that there was definitely a gap in terms of NHS provision so the first thing that struck me was the need for this intervention or an intervention is for this particular population.

C2

I was interested just to see what that was like in practice, but also, really pleased that there was a movement in terms of just an acknowledgement that there needs to be, not necessarily an alternative, but something that is more meaningful to somebody with autism so they can do the same kind of work as everybody else I guess.

C3

I wasn’t aware that there’s so little out there with regard to specialised interventions for people with autism. We had a little bit of training about how to adapt CBT intervention for people with autism, but I didn’t realise until looking at the trial that actually this is quite another area. I thought it was needed and I think from my limited experience people might often struggle because people might not be aware that the clients have autism and they may also perceive them as just being difficult or non-co-operative when actually it’s more about specific communication needs.

C4

All coaches also felt that the design of the GSH intervention was appropriate for people with ASD diagnoses, in particular the focus on BA:

I was really interested in the study, because I’d noticed from assessing in IAPTs, that actually I would speak to people and it was very clear that people were experiencing the difficulties and they were interested in engaging with therapy for, but actually then trying to pass that person on to someone that understood autism or using the cognitive–behavioural therapy with somebody with autism, it can be hard for the person with autism, because of the heavy cognitive element of it essentially. So, I suppose one of the main reasons I was really interested in it, just to see if it worked and how people responded to it, was because this was something that was mirroring work that I’d done in IAPTs but was moving away from the cognitive focus and more on to the behavioural situation because you’re seeing the content.

C3

The actual design felt that it was appropriately ambitious, if that make sense. Then I read the literature that supports behavioural activations in adult depression and that’s very well founded so it seemed like a really sensible approach to maybe use an adapted version of something that we know works with the typical population.

C2

I think behavioural activation was a good model, it makes sense to me to use something like that rather than getting caught up with cognitions and feelings of people who don’t necessarily have the emotional literacy to pull all that together and getting into things like that.

C5

All coaches praised the comprehensive training and supervision they received during the trial, in particular the focus on communication approaches to use with participants and the details provided of the underlying theory of the GSH intervention. All coaches described how, after receiving the training, they felt knowledgeable, confident and well prepared to deliver the intervention:

I was offered ample training and then ongoing supervision, even though I hadn’t started seeing patients yet. So I was definitely very well prepared, did a lot of role-plays and then offered ongoing supervision every week as well.

C1

I think that the information and the openness of the team was brilliant actually. I felt that everybody that I spoke to but also from a personal perspective, I always felt that I could ask questions and that we were appropriately support. I thought that the information given was great if you were to go to reread that information or find out more . . . I thought that the training I received on the interventions and some training that I had on amending communication approaches to adults with autism prepared me.

C2

I actually thought the training was really good and the supervision was amazing as well . . . In the handouts that you go through with people is a series that hopefully makes sense to someone with autism, and I think what the training did was go beyond that and explain why this is even relevant in the first place, and here’s the theory behind that as well. So, we did lots of work around cognitive–behavioural therapy and behavioural activation in general and then honed in on the specific treatment. I felt that was really good for my learning. I find it much better to be oriented in where this has come from rather than drop down in the middle of something and deliver it in a more regurgitated basis . . . it gave a deeper understanding of what you are delivering. When you were then in treatment and people asked you questions, that maybe deviated from the content, you were much better prepared for explaining because they’d given you a context and approach as well.

C3

Views and experiences of the guided self-help intervention

All coaches valued and appreciated the structured design of the GSH intervention and the accompanying material and all followed the order of the sessions as designed:

I think it was good in terms of the structure and a planning meeting first and I liked the folder and the materials seemed all very well designed and straightforward to use. I think having certain measures before each session worked well as well.

C4

I think the structured method offered, having the different booklets and visual tools and things to actually work through questions, answer exercises to do, I think that works really, really well. It helps keep it focused and it also allows both of you to be looking at the page and focusing on that rather than there being pressure of lots of eye contact and more social interactions. So I think that works really well and people, by and large, I think really engaged with it and picked up lots of new things.

C1

I thought it felt comfortable for me in terms of the order, absolutely fine.

C3

Coaches commented on the importance of the initial planning session for demonstrating to the participants that they had a good understanding of ASD and issues relating to communication differences, particularly for those participants who had previous negative experiences with mainstream NHS mental health services:

I think some of them came in a little bit resistant and sceptical but at the same time because they were there, I guess, they kind of knew they were hopeful at the same time and they all kept coming back, so . . . Just having had previous bad experiences, I think, like they had depression for quite a long time and nothing really lifted it, they’d done a little bit of psychology before and not really connected well with it. I guess a lot of them were people who got their diagnosis in maybe their 30s, 40s, like 6 months previous to coming in to see me. So they’re people who’ve kind of been failed by their school system, their workplace, just kind of society in general not really recognising their needs. A few of them had to actually fight really hard to get an assessment for the autism diagnosis as well. So a lot of them, yeah, came in not really having had this streamlined and pleasant experience before . . . I think, I mean, I just then took time to sort of listen to what had happened before and showed empathy for it and then just really explained why we were doing what we were doing and what we hoped it was gonna achieve and encourage them to give it a try, and they did. So yeah, just had to be quite careful then to attend to those experiences rather than kind of rush in and be like, ‘Right, well, we’re gonna focus on positive feelings now’. Yeah, it took time to really show that I’m gonna be here for you, this is what the aims are.

C1

I think absolutely without question the most important part of delivery is their coach or therapist having an awareness of individuals with autism and what their needs for adaption in terms of how future approaches are likely to be. I think obviously the intervention itself is adapted for that population but I think if you were just to follow the material by route but not take your individual understanding then . . . the impact or the potential value of the intervention will be lost. And that’s again based on things participants said to me which was that most of them had experienced former CBT through an IAPT service but they found it ill-fitting for their needs and the therapist didn’t show them any aware of understanding between the likely interaction between ASD and their depression. That just meant that they disengaged and didn’t complete the treatment so I think that first planning session that’s often 90 minutes long is so key because that’s your opportunity as a therapist to show the participant that you do understand them and their needs. And you can individual this approach, albeit there’s pre-set material but you can individualise this to meet their needs and just show understanding in terms of communication differences. I think that’s the most important part of this intervention, definitely.

C2

Sessions 1–3 built on from each other, with each focusing on situations, behaviours and feelings, respectively. The majority of coaches commented that introducing the concept of situations on its own was challenging for some participants, who wanted to also talk about behaviours and/or feelings:

I thought it flowed really nicely. I could see how the content was building on the theory as you went through the weeks. I felt that at the start, when you are suddenly talking about a situation, people were often a bit initially taken aback by that. That needed a little bit of explaining in my experience, just to orientate that person to the treatment. But I think that has more to do with the fact that it’s the first thing you’re introducing rather than the actual concept itself.

C5

I think it worked really well. I think that something that I noticed was the word situation, meaning like the place that the person goes, could get a little bit confusing. Because the people and myself, when I initially started the training in it, they saw the situation as sort of an overall event, like an all encompassing thing of the place, the behaviour and how it made me feel, whereas in the training it just means the place. When they were trying to do the feelings triangles they would get really confused about the situation, they’d start writing down what they were doing and then it would all be a little bit muddled . . . So when they’re trying to write situation, behaviour, feelings, for situation they sort of start to write everything. Does that make sense? . . . So it’s like they take situation as the overall event as opposed to only just the place.

C1

Quite a few people I worked with felt that session 1 to 3 they’re not quite sure what’s going on because it’s split up quite a bit isn’t it and you only really find out about the feelings triangles and what that means from about session 3 so sometimes people are a little bit unsure at the start.

C4

. . . when you started to talk about situations and then behaviours, everybody was itching to tell you how that felt and so it just felt a bit unusual at times to be trying to segregate them. But often you had so much supervision to explain why the function is that. So, sometimes you could just naturally feel that the next section, they would come in and people were ready to integrate that.

C3

Situations

The situation map was introduced in the first session and then used throughout the intervention. Some coaches commented that participants preferred to draw their own map as the one provided either did not have the right locations or did not have the locations in the right order. Others commented that the map worked initially to explain the concept, but after a while participants preferred to list their activities rather than use the map:

It worked well, I only had one person that drew their own. I think after a few sessions people sort of stopped using it quite as much, didn’t really refer back to it the same way. So at the start of each session you’d be like, ‘OK, so where did you go? Would [sic] did you do? What gave you the most pleasurable feelings this week?’ They wouldn’t have mapped that out, they would just be able to then remember where they went. So after a few sessions it didn’t really get used.

C1

The thing that people in my experience challenged more was the map, because they didn’t want somebody else’s, they wanted to create their own. I’d be interested to hear what the other coaches are saying. I think everybody that I worked with created their own map, or didn’t like the map and wanted a list or something like that . . . Some of them created some really amazing maps, and I suppose from a practical perspective, and this is my opinion, and as a patient mechanism, I don’t think it’s big enough. You’re filling it with more and more things and often the person will go to the same place many times and you need to be able to put on lots of feelings. So, I think it could do with . . . one of the maps that somebody made was probably four times the size of the one that is included in the handouts. So, possibly some changes to the map or just making it OK to create your own if you want to.

C3

Yeah, and a couple of people, maybe three out of the people I saw, they wouldn’t engage very well with the map at all and just preferred to write things like an activity schedule as you would in a more standard CBT-type model but some people the visual for the map was really good . . . Yeah. The task would always be to put a triangle on your map. Sometimes people wouldn’t want to do that, but they would make lists of things that they’d done and then they would highlight it a different colour and things like that so a good one would be a pink or they would prefer to use lists or words rather than using the maps.

C5

Feelings

All of the coaches said that the materials provided to prompt participants to discuss their feelings were valuable and worked well to stimulate discussion. The flexibility in what could be used for the scale for the feelings chart was appreciated by coaches and allowed them to individualise the intervention based on the participant’s preference. However, the majority of coaches commented that the use of pictures, such as Winston Churchill’s black dog, was not well received, and this was thought to be confusing for participants because of the use of metaphors:

The only other comment would be the feelings chart, where it has Winston Churchill’s black dog on it, I don’t know anybody that used that. It was either numbers or colours. So, I don’t know if that is to do with the metaphorical side of things not being well received.

C3

With the feeling thermometers there was a bit of confusion about you could pick which one you want to use. Then generally people finding that quite tricky. The colour one worked really well. One guy was thinking very much about his colour intensities, his feeling intensities in terms of colours. The dog one, as I said, the black dog of depression, I explained where that came from but that was not on the person’s . . . ‘I like dogs. That’s a really cute dog. Why is it . . .?’

C4

I was a little bit dubious about was the emotions part, the feelings and that is where the difficulties were. A lot of people didn’t recognise what feelings they were having in the moment and things like that, but I think definitely the way the feelings triangle worked breaking it down with the situations and going through that process of doing and the feelings thermometer sessions and all of those sorts of things really helped people to engage with that.

C5

Granularity

Several coaches described how some participants initially struggled with the concept of granularity, but commented on the importance of the intervention being guided by the coaches to help demonstrate the concept. The granularity exercise was, however, described by some coaches as a critical point in participants’ learning about their own feelings:

I think granularity was hard without somebody there. I don’t think it’s because it was disliked or anything like that. It just is more theoretical. It does need somebody there often to help you and I know that trying to do that on your own when you already struggle to think about your thinking and mentalise, that was hard.

C3

. . . granularity aspect of really focusing in on, ‘OK, I like doing this, but what is it about this that is pleasurable?’ and think about all the different aspects. Would it be the same if this wasn’t there, or this wasn’t there? The magnifying glass metaphor of really focusing in on the stuff that is enjoyable and I guess just rating it. I thought that was a quite an eye-opener to me.

C4

Granularity I think felt a little bit weird for them in session 2 because it’s quite an unusual thing to go out and do but they generally go with it even though they’re not quite sure what it’s leading to. Some people said it was a bit silly I think but they would have a bit more context around what it was leading to I think they would have been more onboard with it. All the written questions were generally quite useful.

C5

. . . the feedback I’ve got from people, when they come to the very final session they reflect on what’s been the most important piece of learning for them and a lot would say the behaviours one, where they do this granularity exercise and break down different situations and see all the little small moments in it and how their feelings can fluctuate so much within just a small amount of time.

C1

Maslow’s hierarchy of needs

All coaches described the hierarchy of needs task in session 5 as a pivotal point in the intervention and a ‘lightbulb moment’ for participants’ understanding of what changes to make in their life. Coaches commented that participants engaged with the task well and that it drew together all of the learning and understanding from the previous weeks:

It was the hierarchy of need session. The participant that I worked with, it made so much sense to her that she wanted to talk about it in depth and she had kind of a lightbulb moment in terms of understanding where she’d been putting most of her time but where it might make more sense to change and focus and so that was a big moment for her. I think she had prior experience of exposure to the hierarchy of needs. Maybe if someone hasn’t I think, for me in terms of delivering the intervention, that session is a real turning point in terms of clarity. I think that’s the point where the intervention moves from almost exploring current to considering change to make sure that the participant is still with you and that that session makes sense because it’s key for the remaining sessions.

C1

. . . the one element that seemed to be universal with the people that I worked with was the Maslow hierarchy of needs. So, not so much in the sense of . . . most people had heard of it before, but explaining that the relationship between that and compassion, and linking that then to situations, new situations, new behaviours, and if you do the same behaviours over and over again, or if they become more restricted then you’re taking less of these needs higher up. That seemed to work with the people that I worked with . . . it just seemed to really get the message across. It was a chart. People were ticking things off. They had a chart and a map of what they were currently doing. They could compare that to this tangible list of needs that [inaudible – 0:18:33] mood improving, and it then created a tangible goal for the next week, because I could see that they haven’t met this need in this box, therefore I need to go to this situation or do this behaviour. So, it often seemed to be the session that would pull things together like that.

C3

. . . the hierarchy of need session, just because I think people could really see. One of my clients was actually very active. He didn’t need to schedule more behaviours but it was about finding a better balance. We actually found that a lot of his stuff . . . He drew this lovely picture of where he said, ‘I’ve got my achievement oriented stuff and I’ve got my basic safety and everyday needs, but the social level was just really empty’. So he drew that like a pyramid where in the social part there were just a couple of unwritten columns. This is obviously where I need to find some stuff to fill it in. I think that was really, really useful in helping people recognise, especially if people have quite a few activities going on, to think ‘Yes, this is all good for my achievement and my self-esteem, but actually is really stressful as well and I maybe need to wind down and attend to my safety needs or attend to my social needs instead’. I think that worked really, really well.

C4

. . . hierarchy of needs and relating it. It was like the penny dropped sort of moment when they realised oh, this is why this is so important, this relates to this and it provided a framework I suppose so after that made much more readily kind of say, ‘Oh I can identify what needs I’m not needing and what I need to meet,’ and it kind of, why it’s so important for them to be doing some of things that they were if that makes sense . . . I think sometimes some people are a little bit unsure that they got the granularity and they knew the concept of the feelings triangles and everything, but it wasn’t until we did session 5 [hierarchy of needs] and looked at it in the context of we have needs that we need to meet in order to keep ourselves . . . our well-being in check and all those sorts of things that they made that connection . . . I guess it sort of just helps giving more context to people in terms of why going to all these different places and bringing about all these good feelings were important. It gave it a framework I think, the context.

C5

Somebody did their own hierarchy of needs as an extra task. They came back with a sheet of paper with a colour-coded hierarchy of needs with all of their activities within it, so it was things that I hadn’t actually asked them to do but they just developed their own understanding and took it upon themselves to do.

C5

Between-session tasks

All coaches described and provided examples of trial participants appreciating and understanding that the intervention was self-help and that the emphasis was on the participant to work through the between-session tasks, with the coach being there only to help and guide them. Coaches explained that these therapeutic principles were appropriate for this client group:

Yeah, I think so, like increasing positive feelings. Yeah, because they, yeah, they picked up pretty quickly that it sort of had to be focused just on that, like these positive feelings, positive behaviours, and it might come in a little bit down at the start of the session and like how was your week? They would know not to talk too much about another problem or they would say, ‘Oh, I’m not really supposed to go into this too much, am I?’ I’m like, ‘Oh well, it’s OK for a couple of minutes and we’ll then look at this’.

C1

It’s continuous throughout all the weeks that we’re working together as the backbone of our work. It’s not something to just do and then forget about it. I think that there’s an opportunity there to set up the fact that the home practice is continuous. The therapy continues outside of the sessions. The sessions are guided. They are not the entirety of the intervention. I think you’ve got a real opportunity to get all of those messages across and I think there’s a risk that if you don’t do that skilfully then, that maybe it won’t flow as well, or you won’t have that backbone to your profession that you need for it . . . So, in terms of one person I can think of who I felt really did understand the treatment aims, they were continuously working towards things that they used to do, that they used to enjoy and they could use an event within that week where they were starting to make a distinction between feeling happy and striving towards feeling content. They started to create through their situations map. That brought contentment areas that brought the opposite of that and how could they make those harder situations feel better and we were doing subtle changes in their behaviour, to try and make their feelings better . . . So, for example, one person worked out that driving to a particular place if they were a passenger was worse than being the driver, so they would drive. I think that shows real insight in to the elements of the intervention and behaviour situations and how changes to those could change feelings and really felt like this person grasped that and you could see the impact of how they were moderating their behaviour in that way.

C3

I think it just needed a bit of explanation to say ‘I’m your coach. I’m not your therapist. This is for you to work through and for me to help you with the experience and the main work happens between sessions’. I think what was important possibly to say for people with autism because there was a few times in the manual that there were options. You can use different kinds of feelings, thermometers and that kind of thing to just say, ‘These are resources that you may find helpful’, because I think there is a tendency for some people with autism at least to work through everything by the book or to find it difficult to . . . You can actually pick and choose rather than you have to do everything in a particular way . . . I think especially when somebody is depressed, to have that personal connection is absolutely essential. I don’t think if you just give out the folders for people to read in their own time, it wouldn’t be the same sort of engagement.

C4

I think the element of having it as guided self-help has probably been really useful for the people with autism. I think that for someone like myself being there to be able to kind of suggest things or prompt things probably really helped in that respect . . . I think there was one man that I worked with who, we went through . . . in his situation map we mapped out each for him and then the next week he came back, and he had done a really comprehensive schedule of everything he’d been doing in the week and from that I’d pulled out the bits that had been good activities for him. That’s not a very good example actually but I think that was how I knew that he’d understood though what we were trying to do there. There was one guy who really engaged with the session 5 and then the next week he’d made a list of the different areas in his life that he felt he was making progress on and that he wanted to and things. People used the materials but then did their own little bit to work alongside it sometimes. I guess made it personal to them. People were coming back with homework and if I could see that they were able to do that in really granular detail in the sessions and things like that.

C5

Coaches described that, on the whole, participants engaged well with the between-session tasks, with engagement varying between participants and between tasks depending on the individuals’ involvement in the task. If the task had not been completed between sessions, coaches took time during the session to complete the task or to explore why it had not been done:

Some were good at doing the homework and some not. Doing a scheduled activity worked really well, that was session 4 we introduce that, and that got carried on after every session. So people would do that and only some of them would come in having done a feelings triangle for that and so this is how it met their needs. So if they came in without having done it I would still spend the first 5, 10 minutes talking over the past week and what needs were being met.

C1

I think the first activity scheduling that was really good, because I thought then when people had actually . . . The planning worked really well and the breaking it down, what works as planning and then to reviewing that. Those were popular sessions where I spent quite a lot of time as well, so where I went up to an hour to actually set some activities and then come back and do them or not do them and then troubleshoot around why, what got in the way or what could be done instead. I thought that went quite well.

C4

I think in the one experience I had of doing it [homework], it was in the new behaviours, new places because the participant had an understanding of the likely hoped value that that change would bring. She really really engaged. She drew posters and put them on her kitchen wall to remind her of the things, the new activities she was going to do that week. She kept a diary of when she did them. She involved her partner. She came back to me really kind of keen to report the things that she’d done and how she’d found them. I think certainly in terms of how this participant engaged, I felt that she was using lots of the skills and tools that we’d worked through before to understand why this activity or this change was going to be potentially of value. I was worried when I first looked at the intervention. I was worried by the word new because the training that I’d been giving on working with adults with autism was that quite a lot of people enjoy regularity and routine and change makes them anxious so I thought, ‘Gosh, this whole thing is building up to this point where they have to change something’.

C2

Session length

All coaches described sometimes struggling to fit some sessions into the allotted 30 minutes, and all felt that 45 minutes was more realistic. The sessions that were most likely to take longer were those introducing the concept of granularity and the hierarchy of needs task and also when participants had not completed their between-session tasks:

I think for some people they would get through each session really quickly and some I would take 45 minutes maybe, it just sort of depended on how well they got the concept. So yeah, it just really depended on each person. Sometimes 30 minutes just wasn’t long enough for sessions that had like a really big component to them. So like the session called ‘the hierarchy of needs’, that’s session 5, some got it really, really quickly and for some other people it was like a pretty long session, they really had to think about it and they didn’t quite understand it right away. But then I didn’t want to send them home not having got it.

C1

I think some of the sessions are a lot busier than other sessions but on the whole I found it difficult to complete the prescribed activities tasks within 30 minutes. I think that on the whole that is difficult and it varies hugely. If the individual has done their homework then the session kind of works as planned but if they haven’t there isn’t much room to address that and then complete the next task, which is often dependent on the homework having been completed . . . I did still find some of the sessions very full and I think most of my sessions were around 40–50 minutes generally.

C2

Yes, I found you needed longer than that definitely. I think if you want to be . . . I think you could do it in half an hour. I think if you were literally just going through the content, then yes. But if you wanted to allow the person to show you what they’ve done that week, talk about their situations which is often in the guidelines for each session, then weaving that in to their goals and stopping to get a sense of where you could help them out a bit in that session, it’s those bits that take it a little bit longer. I don’t think you’re talking about an hour. I think most of my sessions were around the 45-minute mark, which was the upper end of what I was allowed . . . the boundaries of what was acceptable. But I think half an hour is short.

C3

The length of time was a bit of a challenge. From what I remember I think it was session 3 which was quite full on. I think that’s where you introduce granularity. After that, the later sessions, new behaviours, new situations, I think that could have been one session. That was relatively quick because it’s repeating. It’s useful that it’s two and it’s building up on things, but that said, 2 was in 25 minutes whereas some of the other sessions I went up to a minimum 45 minutes and up to an hour.

C4

The session length actually that’s one thing, 30 minutes just is not enough really you have to quite a bit of explaining and sometimes people haven’t done the work in between the sessions and you spend a little bit of time just trying to look at that with them and I think most of my sessions were more like 40–45 minutes . . . Sessions 3 and 5 for example actually were really long in comparison to often the first in session 1, not 0, session 4 I think particularly and then the later ones but the later ones that makes sense because you’re generally just going away in the week and doing it and then coming back and reviewing it. I guess something about was there too much in session 3 or 5 could [inaudible 27:29]. I think three in particular stood out as that there was quite a lot.

C5

Treatment goals

The majority of coaches described that setting treatment goals with the participants was a challenge. This was because some participants chose long-term life goals that would not be achievable over the course of the intervention and others were not able to think of goals on the spot. It was suggested that thinking of goals could be given as a between-session task between the planning meeting and session 1 to allow time for participants to consider what could be achievable over the course of the intervention. It was also suggested that additional material for coaches to guide the participants to set goals would be welcomed. Coaches also suggested that revisiting the goals more regularly during the intervention would be a useful addition:

. . . you do your three goals in the planning session and then there’s one more session before the very final one that asks you for them again. I think it might be helpful to have two sessions, at least, between the very first one and the very last one that asks them about their goals and their progress through them, ‘cause I think it’s quite easy to forget what they are.

C1

I think it’s really important, and the material does guide the therapist to do this but it’s really important that before the goals are set the participant understand the parameters of capability within the intervention. And I know it’s the same in any therapeutic goals setting but I think it’s about finding a balance between something that’s attractive for the participant but then it’s also possible within the time frame and the likely progress that’s going to be made. I struggled with that, to strike a balance and so took it to supervision and I actually went back to the goals in the first active treatment session to just talk about them again. I know later on it was introduced in the material that does goal reviews. I forget which sessions they were introduced in. I think that’s a really good idea but possibly in the material for the concept of goal settings to be introduced and discussed at that planning meeting and then homework to be for the participant to consider what those goals might be. Then come back to, I think to meet the new participant, for them to completely grasp what you’re going to be doing for the next eight sessions, for them to accept the realms of the possibility in terms of improvement and then commit to three goals all within that first session is probably a bit too much. Maybe to introduce goals and then agree them in the following session might be quite nice for them.

C2

I found goal-setting quite difficult. I’ve got to say, I think one of the reasons for that is because I have never set goals when I haven’t assessed a person and normally, I suppose if you were going to do an intervention you would do a bit of a formulation with them, and that gives you a much . . . you start to build and start to visualise those goals together before you set them. So, I suppose it’s about jumping in a bit, and you don’t see the notes from the research assessment. So, you don’t know much. So, I suppose it felt a bit disjointed or a bit disconnected to set goals then . . . I think one thing would be to have an explicit point where you revisit the goals. I know it’s meant to be woven in to most sessions and you’re meant to skilfully go back, but if you could have an explicit point where you both know that you’re going to be just checking it’s still going in the right direction that would be helpful for motivation and mutual understanding.

C3

When it came to revisiting the goals, actually people’s ideas of what was the most important goal had changed in both cases. Again, I think it’s good to have it there, but as long as there’s an understanding this can be flexible and people don’t get upset about, ‘Oh, I haven’t done anything towards this goal’, because actually the situation had changed. The one guy had had a change in job and the other person had one of the goals was to go to school regularly and then obviously that went after the start of the summer holiday. It’s good to have them there, but to treat them flexibly.

C4

In session 4 we review the goals and I really didn’t like doing that part because hardly anybody would report a change at that point because it was almost like at session 3 we were still going through what the concepts were and getting a feel for it and then their one session in touch and understanding what it’s all about and we’re reviewing goals it’s felt too early. It’s good to revisit them and have that in mind but asking to rate them at that point really wasn’t helpful with a lot of the people I work with and if anything, it kind of side-tracked them to ruminate on not having got very far with them . . . I almost feel like it should have been done from session 5. That sounds quite late on to do it unless they were revisited every week so it was just something you did every week but just for them to be in zero and then suddenly in 4 we just got the conduct together we’ve done the first week . . . in fact I don’t think at that point we’ve scheduled any activity, session 4 is the activity setting, it just felt a bit premature and wasn’t always helpful in terms of people . . . they weren’t very positive about that a lot of the time.

C5

Difficult for some people. That would be the only bit that I would have said I could have done with some more materials around or something because some people would come in with very clear ideas and it would be a case of trying to help them shape it into a kind of SMART goal that was appropriate, but some people really wouldn’t and for you to spend a lot of time trying to [thrash] that out with them. I’m not sure what would have been helpful but maybe some sort of activity to get people to think a bit more clearly about what it is they want to get out of it and what a realistic smaller goal would look like that would be appropriate for the 10 weeks that you’d be working together.

C5

Sensitivity to change

Twenty-nine per cent (12/41) had at least a 50% improvement in symptoms on HAM-D at 16 weeks compared with baseline. The results in Table 14 show that the PHQ-9 was better able to distinguish between those who did and those who did not have at least a 50% improvement in symptoms on HAM-D as the area under the curve was higher (0.93, 95% CI 0.83 to 1.00; Figure 3) than the area under the curve of the BDI-II (Figure 4).

FIGURE 3.

The ROC curve of PHQ-9 against HAM-D scores at 16 weeks. Area under curve = 0.9272 SE (area) = 0.0475. SE, standard error.

FIGURE 4.

The ROC curve of BDI-II against HAM-D scores at 16 weeks. Area under curve = 0.8213 SE (area) = 0.0738. SE, standard error.