SCHRES-07055; No of Pages 8

Schizophrenia Research xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

Schizophrenia Research

journal homepage: www.elsevier.com/locate/schres

A parallel group randomised open blinded evaluation of Acceptance and

Commitment Therapy for depression after psychosis: Pilot trial
outcomes (ADAPT)
Andrew Gumley a,⁎, Ross White a,b, Andy Briggs c, Ian Ford d, Sarah Barry d, Corinna Stewart a, Sara Beedie a,
Jacqueline McTaggart e, Caoimhe Clarke a, Rachel MacLeod e, Emma Lidstone e, Bruno Salgado Riveros c,
Robin Young d, Hamish McLeod a
a
Mental Health and Wellbeing, Institute of Health and Wellbeing, University of Glasgow, United Kingdom
b
Institute of Psychology, Health and Society, University of Liverpool, United Kingdom
c
Health Economics and Health Technology Assessment, Institute of Health and Wellbeing, University of Glasgow, United Kingdom
d
Robertson Centre for Biostatistics, Institute of Health and Wellbeing, University of Glasgow, United Kingdom
e
NHS Greater Glasgow & Clyde, United Kingdom

a r t i c l e i n f o a b s t r a c t

Article history: Background: Depression is one of the major contributors to poorer quality of life amongst individuals with psy-
Received 15 July 2016 chosis and schizophrenia. The study was designed as a Pilot Trial to determine the parameters of a larger, defin-
Received in revised form 18 November 2016 itive pragmatic multi-centre randomised controlled trial of Acceptance and Commitment Therapy for depression
Accepted 18 November 2016 after psychosis (ACTdp) for individuals with a diagnosis of schizophrenia who also meet diagnostic criteria for
Available online xxxx
major depression.
Methods: Participants were required to meet criteria for schizophrenia and major depression. Blinded follow-ups
Keywords:
Schizophrenia
were undertaken at 5-months (end of treatment) and at 10-months (5-months posttreatment). Primary out-
Depression comes were depression as measured by the Calgary Depression Scale for Schizophrenia (CDSS) and the Beck De-
Randomised controlled trial pression Inventory (BDI).
Pilot Results: A total of 29 participants were randomised to ACTdp + Standard Care (SC) (n = 15) or SC alone (n = 14).
Acceptance and Commitment Therapy We did not observe significant differences between groups on the CDSS total score at 5-months (Coeff = −1.43,
95%CI −5.17, 2.32, p = 0.45) or at 10-months (Coeff = 1.8, 95%CI −2.10, 5.69, p = 0.36). In terms of BDI, we
noted a statistically significant effect in favour of ACTdp + SC at 5-months (Coeff = − 8.38, 95%CI − 15.49,
−1.27, p = 0.02) but not at 10-months (Coeff = −4.85, 95%CI −12.10, 2.39, p = 0.18). We also observed sig-
nificant effects on psychological flexibility at 5-months (Coeff = −8.83, 95%CI −14.94, −2.71, p b 0.01) but not
10-months (Coeff = −4.92, 95%CI −11.09, 1.25, p = 0.11).
Implications: In this first RCT of a psychological therapy with depression as the primary outcome, ACT is a prom-
ising intervention for depression in the context of psychosis. A further large-scale definitive randomised con-
trolled trial is required to determine effectiveness.
Trial registration: ISRCTN: 33306437
© 2016 Published by Elsevier B.V.

1. Background depression (Whitehead et al., 2002; Birchwood et al., 2000). Depression

is associated with greater risk of suicide (Drake et al., 1985), poorer ad-
Depression contributes to poorer quality of life amongst individuals herence to treatment (Conley, 2009), greater interpersonal problems
with psychosis and schizophrenia (Saarni et al., 2010). Prevalence rates (Rocca et al., 2005) and greater insight (Mintz et al., 2003). Shame is
suggest between 33% and 50% of individuals meet criteria for major key feature in the emergence of depression (Gumley et al., 2010),
where individuals' envisage a bleak future status, through stigma and
the loss of social, interpersonal and vocational roles (Birchwood et al.,
2000).
⁎ Corresponding author at: Mental Health and Wellbeing Research Group, Institute of
Health and Wellbeing, University of Glasgow, Gartnavel Royal Hospital, Glasgow G12
There is a lack of robust evidence supporting the use of antidepres-
0XH, United Kingdom. sant (Whitehead et al., 2002) and psychological (Wykes et al., 2008) in-
E-mail address: [email protected] (A. Gumley). terventions for depression in context of schizophrenia. Although there

http://dx.doi.org/10.1016/j.schres.2016.11.026
0920-9964/© 2016 Published by Elsevier B.V.

Please cite this article as: Gumley, A., et al., A parallel group randomised open blinded evaluation of Acceptance and Commitment Therapy for
depression after psychosis: Pilot tri..., Schizophr. Res. (2016), http://dx.doi.org/10.1016/j.schres.2016.11.026
2 A. Gumley et al. / Schizophrenia Research xxx (2016) xxx–xxx

is preliminary evidence that depression symptoms improve in people 2.2. Participants

receiving CBT for psychosis (CBTp) (Wykes et al., 2008) this important
outcome domain is not typically assessed in CBTp trials (Jauhar et al., Participants were consecutively recruited, assessed and randomised.
2014) so there is a need to build the treatment evidence base. Accep- Inpatients or outpatients, aged 16 or over and receiving (a) anti-psy-
tance and Commitment Therapy (ACT) could offer a promising psycho- chotic medication (b) psychiatric follow-up and (c) follow-up from a
logical intervention that helps reduce unhelpful coping strategies secondary mental health care community based services. Participants
including rumination and avoidance, and enables commitment to be- met DSM-IV-TR criteria for schizophrenia and major depression (con-
havioural changes consistent with personal values. Randomised con- firmed by Structured Clinical Interview for DSM/SCID-I & Calgary De-
trolled trials show that ACT reduces depression in non-psychotic pression Scale/CDSS for Schizophrenia; score N 7; Kim et al., 2006).
populations (Hacker et al., 2016). Individuals with substance use problems were eligible for inclusion
Previous non-blind randomised controlled trials investigating but those with significant learning disability, or who were unable to
ACT for psychosis with rehospitalisation as the primary outcome speak English were not included.
have shown therapeutic promise in US acute inpatient settings
(Bach and Hayes, 2002; Gaudiano and Herbert, 2006). Gaudiano 2.3. Entry
and Herbert (2006) found a small effect of ACT on negative emotion
in people with psychosis. White et al. (2011) investigated ACT for Potential participants were identified by their clinical team or by
psychosis with the primary outcome focused on emotional distress. self-referral. Following signed informed consent a Research Assistant
They found a significant difference between the ACT and Treatment (RA) conducted baseline assessments. The Psychosis and Mood Epi-
as Usual (TAU) groups for negative symptoms. There was also a sodes section of the SCID-I was used to assess diagnostic eligibility.
trend on the limit of significance for differences between the groups Inter-rater reliability for the SCID-I was 90% (range 82–100%).
in depression (p = 0.051) and improvement in depression was asso-
ciated with self-reported increases in mindfulness. In a later analysis, 2.4. Randomisation
White et al. (2015) found that ACT was associated with significantly
greater likelihood of achieving a clinically significant improvement Eligible participants were randomised following completion of the
in depression. baseline assessments. Randomisation (at the individual level) was
To summarise, feasibility research informed the present study that organised at the Robertson Centre for Biostatistics, a fully registered
(a) ACT could be delivered to outpatients with psychosis and was highly and NIHR approved UK Clinical Trials Unit. Randomisation was stratified
acceptable, (b) that ACT appeared to improve symptoms of depression for early (b2 years duration) versus established psychosis (N2 years
and negative symptoms and (c) that improvements in depression duration).
were associated with ACT relevant mechanisms of change (increased
mindfulness skills). Therefore this study was designed as a Pilot Trial 2.5. Follow-up
(Craig et al., 2008) to determine the parameters of a larger, definitive
pragmatic multi-centre (UK wide) randomised controlled trial of Accep- Participants were assessed at entry pre-randomisation, 5 months
tance and Commitment Therapy for depression after psychosis (ACTdp) and 10-months by a blinded RA. All assessments were audio-recorded
for individuals with a diagnosis of schizophrenia who also meet diag- with consent by the participant. Management of blind breaks, inter-
nostic criteria for major depression. In this manuscript we address the rater reliability and safety reporting were governed by specifically de-
following questions: signed Standard Operating Procedures.

a) What are the potential numbers of participants who fulfil eligibility 2.6. Protecting the blind
criteria?
b) What proportion of potential participants provides fully informed RAs entered participant details via a secure web based portal. Fol-
consent to participate in the Trial? lowing allocation, another member of the research group received the
c) What is the overall rate of follow-up in the first 5-months and at 10- outcome of randomisation and informed the participant and their treat-
months? ment team. Blinding was maintained using a wide range of measures
d) What proportion of participants engages with ACTdp? (e.g. separate offices for therapists and researchers, protocols for an-
e) What rates of improvement in primary (depression) and secondary swering phones, message taking and secretarial support, forbidding
outcomes (Positive and Negative Syndrome Scale and Questionnaire any discussions of participants between RAs and therapists following
for Personal Recovery) are observed at 5-month (end of treatment) randomisation, separate diaries and security for electronic
follow-up and 10-month follow-up? randomisation information).
f) What are the associations with ACT specific mechanisms (mindful-
ness skills, psychological flexibility) and outcome in terms of 2.7. Blind breaks
depression?
When a blind break occurred, the RA affected informed the Chief In-
vestigator (CI) within two working days. In this event a second assessor
2. Methods/design (blind, where possible, to the participant's allocation) undertook an in-
dependent rating of the assessment. Following independent rating, as-
2.1. Design sessors met to discuss and resolve any discrepancies. The original
assessor entered agreed ratings into the Case Record Form (CRF).
The study was a Parallel group Randomised Open Blinded Evaluation Where possible, another team member, who remained blind to
(PROBE) of Acceptance and Commitment Therapy for depression after randomisation status, completed the remaining assessments for that
psychosis (ACTdp). The study protocol was registered before initiating participant.
recruitment (ISRCTN: 33306437). The protocol was published before
treatment codes were broken and data were analysed (Gumley et al., 2.8. Inter-rater reliability
2015). Ethical approval was provided by West of Scotland Research
Ethics Committee (12/WS/0311). Managerial approval was provided Inter-rater reliability meetings were held on a monthly basis. RAs
by NHS Greater Glasgow & Clyde (GN11CP470). were trained in all of the study measures before commencing data

Please cite this article as: Gumley, A., et al., A parallel group randomised open blinded evaluation of Acceptance and Commitment Therapy for
depression after psychosis: Pilot tri..., Schizophr. Res. (2016), http://dx.doi.org/10.1016/j.schres.2016.11.026
 A. Gumley et al. / Schizophrenia Research xxx (2016) xxx–xxx 3

collection. This included observation of a live assessment of psychopa- 2.13. Treatments

thology ratings (PANSS, CDSS, SCID) with an experienced research as-
sessor. Audio recordings were used for ongoing RA training, secondary 2.13.1. Acceptance and Commitment Therapy for depression after psychosis
assessment when blind breaks occurred, and fidelity checking to detect (ACTdp)
and correct rater drift. Individuals received up to 5 months of individual ACTdp. ACTdp is
based on the rationale that the experience of psychosis can threaten
progress in valued life domains. The ACTdp intervention protocol was
2.9. Serious adverse events used to enhance engagement with valued life activities through values
clarification, increasing mindfulness and psychological flexibility and
In accordance with NIHR trial management standards (Trial reducing experiential avoidance and fusion with experiences. An indi-
Managers Network, 2014), a Serious Adverse Event (SAE) was defined vidual formulation based on the six ACT core processes was developed
as any occurrence that (a) resulted in death, (b) was life threatening, for each participant in the treatment arm. Fidelity of treatment strategy
(c) required hospitalisation or prolongation of existing hospitalisation, use and consistency with the ACT model were determined via weekly
(d) resulted in persistent or significant disability or incapacity, or (e) supervision by a senior clinician. The details of intervention have been
was considered otherwise medically significant by the Chief described in greater detail elsewhere (Gumley et al., 2015; White,
Investigator. 2015).

2.13.2. Standard Care (SC)

2.10. Primary outcomes Treatment received by all participants in the trial was examined in
order to establish the parameters of Standard Care. For inclusion, all par-
The Calgary Depression Scale for Schizophrenia (CDSS, Addington et ticipants had to be in receipt of antipsychotic medication and follow-up
al., 1990) assessed severity of depression. The CDSS was better the BDI, from a secondary specialist mental health service.
PANSS – Depression subscale, and HAM-D for both sensitivity (94%) and
specificity (89%) of SCID Depression diagnosis (Kim et al., 2006). Inter-
2.14. Statistical analyses
rater reliability for the CDSS was 92% (range 67–100%). In this study
the Intraclass Correlation Coefficient was ICC = 0.68 (95%CI, 0.46–
As a pilot study we aimed to assess the variability in the outcome
0.82). CDSS was also associated with PANSS Emotional Distress (rs =
data and look for suggestive trends in order to estimate parameters
0.78, 95%CI, 0.67–0.85)
for a definitive multi-site RCT. To this end, we planned exploratory re-
The Beck Depression Scale (BDI-II; Beck et al., 1996) was used as a
peated measures regression models, adjusting for baseline measures in-
well-established self-report measure of depression with excellent reli-
cluding the stratification variable to be fitted to assess treatment effects
ability and validity. In this study Cronbach's Alpha was α = 0.91
on the main outcome measures and the evolution of these treatment ef-
(95%CI, 0.88–0.94). The CDSS and BDI-II were correlated rs = 0.70
fects over time and to estimate residual variability. We aimed to explore
(95%CI, 0.56–0.80). This is consistent with Addington et al. (1993)
outcome measures for strong treatment signals. Therefore we report
who found a correlation of rs = 0.69 amongst outpatients with a diag-
both ITT and Per Protocol Sample (PPS).
nosis of Schizophrenia.

3. Results
2.11. Therapy mechanisms
3.1. Population
The Kentucky Inventory of Mindfulness Skills (KIMS; Baer et al.,
2004) was used to assess four mindfulness skills: observing, describ- The flow of participants into the study is described in Fig. 1. We iden-
ing, acting with awareness, and accepting without judgement. The tified 92 potential participants of whom 55 were referred to the study.
KIMS has good internal consistency, test retest reliability and con- Of this group, 38 gave their informed consent to enter the study and 7
struct validity. In this study Cronbach's Alpha was α = 0.82 were not eligible. Two participants, who initially gave their informed
(95%CI, 0.76–0.88). consent subsequently changed their mind before randomization. This
Psychological flexibility was assessed with the 7 item version of left 29 participants who were fully assessed before being randomised
the Acceptance and Action Questionnaire (AAQ-II; Bond et al., to ACTdp + SC (n = 15) or SC alone (n = 14). In terms of follow-up
2011). This measure has satisfactory structure, reliability, and of ACTdp + SC, two participants declined follow-up at 5-months and
validity. In this study Cronbach's Alpha was α = 0.88 (95%CI, 0.84– we were unable to follow-up 1 further participant at 10-months. In
0.92). SC, 1 participant declined follow-up at 10-months.
The characteristics of the final sample are described in Table 1 below.
The sample were on average 46.5 years old, male (n = 19, 65.5%), white
2.12. Secondary outcomes Scottish (n = 27, 93.1%) with on average 13.2 years of education.
During the trial we observed 7 Serious Adverse Events (SAEs). All
The Positive and Negative Syndrome Scale (PANSS; Kay et al., 1987) SAEs were related to Hospitalisation, 2 (28.6%) were in the
was used to measure psychiatric symptoms. We adopted the 5-factor ACTdp + SC group and 5 (71.4%) in the SC group. There were 14
model that yields scores for positive, negative, disorganisation, excite- unblindings during the trial. Of these 11 (78.6%) were in the
ment and emotional distress symptoms (van der Gaag et al., 2006). ACTdp + SC group and 3 (21.4%) were in SC. All unblindings were
Inter-rater reliability for the PANSS in this trial was 83% (range 63– rated by another Researcher masked to treatment allocation and subse-
97%), and the Intraclass Correlation Coefficient was ICC = 0.88 (95%CI, quent follow-ups were blind rated.
0.83–0.91).
The Process of Recovery Questionnaire (QPR; Neil et al., 2009) was 3.2. Intervention
used to measure service user rated recovery. The QPR has two subscales
measuring intrapersonal tasks and interpersonal factors relevant to per- On average 17.4 (s.d. = 5.9) ACTdp sessions were scheduled, and
sonal recovery and has excellent reliability and validity. In this study 15.4 (s.d. = 6.2) were attended by participants. Of the remaining ses-
Cronbach's Alpha was α = 0.93 (95%CI, 0.91–0.95). sions, 0.7 (s.d. = 1.4) were cancelled and 1.2 (s.d. = 1.5) not attended.

Please cite this article as: Gumley, A., et al., A parallel group randomised open blinded evaluation of Acceptance and Commitment Therapy for
depression after psychosis: Pilot tri..., Schizophr. Res. (2016), http://dx.doi.org/10.1016/j.schres.2016.11.026
4 A. Gumley et al. / Schizophrenia Research xxx (2016) xxx–xxx

TOTAL REFERRALS
N = 55

Declined Consent n = 12
Unable to contact n = 4
Not eligible n = 1

Not eligible n = 7
CONSENTED
Declined to proceed n = 2 N = 38

RANDOMISED
N = 29

ACTdp Standard Care

Baseline Baseline
N = 15 N = 14

ACTdp Standard Care

5 Months 5 Months
N = 13 N = 14
N = 2 declined

ACTdp Standard Care

10 Months 10 Months
N = 12 N = 13
N =1 lost to follow up N = 1 declined

Fig. 1. Adapt consort diagram.

Table 1
Demographic characteristics of sample.

ACTdp + SC
Variable Statistic Full ITT (N = 29) Standard Care (N = 14) (N = 15)

Age (Years) Nobs (Nmiss) 29 (0) 14 (0) 15 (0)

Mean (SD) 46.5 (9.0) 46.2 (8.9) 46.8 (9.3)
Gender Nobs (Nmiss) 29 (0) 14 (0) 15 (0)
Male N (%) 19 (65.5%) 9 (64.3%) 10 (66.7%)
Ethnicity Nobs (Nmiss) 29 (0) 14 (0) 15 (0)
White N (%) 27 (93.1%) 12 (85.7%) 15 (100.0%)
Black minority ethnic N (%) 2 (6.9%) 2 (14.3%) 0 (0.0%)
Years of education Nobs (Nmiss) 28 (1) 13 (1) 15 (0)
Mean (SD) 13.2 (4.2) 14.0 (4.9) 12.5 (3.4)
Highest education Nobs (Nmiss) 27 (2) 13 (1) 14 (1)
Primary or less N (%) 4 (14.8%) 2 (15.4%) 2 (14.3%)
Secondary N (%) 9 (33.3%) 3 (23.1%) 6 (42.9%)
Tertiary/further N (%) 13 (48.1%) 7 (53.8%) 6 (42.9%)
Other general N (%) 1 (3.7%) 1 (7.7%) 0 (0.0%)
Not known N (%) 0 (0.0%) 0 (0.0%) 0 (0.0%)

Please cite this article as: Gumley, A., et al., A parallel group randomised open blinded evaluation of Acceptance and Commitment Therapy for
depression after psychosis: Pilot tri..., Schizophr. Res. (2016), http://dx.doi.org/10.1016/j.schres.2016.11.026
 A. Gumley et al. / Schizophrenia Research xxx (2016) xxx–xxx 5

Table 2
Primary outcomes at baseline, 5-months and 10-months.

ACTdp + SC
Variable Statistic ITT (N = 29) SC (N = 14) (N = 15)

Calgary Depression Scale - Baseline Nobs (Nmiss) 29 (0) 14 (0) 15 (0)

Mean (SD) 14.0 (4.9) 15.0 (5.8) 13.1 (4.0)
Calgary Depression Scale - 5-months Nobs (Nmiss) 27 (2) 14 (0) 13 (2)
Mean (SD) 10.2 (6.1) 12.0 (6.6) 8.3 (5.0)
Calgary Depression Scale - 10-months Nobs (Nmiss) 25 (4) 13 (1) 12 (3)
Mean (SD) 9.5 (5.9) 10.0 (6.5) 9.0 (5.3)
Beck Depression Inventory - baseline Nobs (Nmiss) 29 (0) 14 (0) 15 (0)
Mean (SD) 31.7 (12.6) 29.9 (14.3) 33.3 (11.1)
Beck Depression Inventory - 5-months Nobs (Nmiss) 27 (2) 14 (0) 13 (2)
Mean (SD) 21.9 (10.5) 25.2 (11.3) 18.4 (8.7)
Beck Depression Inventory - 10-months Nobs (Nmiss) 25 (4) 13 (1) 12 (3)
Mean (SD) 23.5 (14.3) 25.7 (14.3) 21.2 (14.6)

The number of participants attending 10 or more sessions (our a priori In terms of KIMS subscales we observed the following outcome sig-
definition of an adequate number of sessions) was 10 (71.4%). nals. At 10-months significant outcomes in favour of ACTdp + SC were
noted for KIMS Observing (Coeff = 4.94, 95%CI 0.47, 9.41, p = 0.03) and
KIMS Total (Coeff = 10.91, 95%CI 0.22, 21.60, p = 0.05). In the PPS anal-
3.3. Outcomes ysis significant effects in favour of ACTdp + SC were noted for KIMS
Accepting at 5-months (Coeff = 4.78, 95%CI 0.68, 8.88, p = 0.02) and
The outcomes of depression measured by CDSS and BDI are de- 10-months (Coeff = 5.39, 95%CI 1.29, 9.40, p = 0.01) and KIMS Total
scribed in Table 2 below. We did not observe significant differences be- at 5-months (Coeff = 9.44, 95%CI 0.07, 18.82, p = 0.05) and 10-months
tween groups on the CDSS total score at 5-months (Coeff = − 1.43, (Coeff = 12.63, 95%CI 3.25, 22.00, p = 0.01).
95%CI − 5.17, 2.32, p = 0.45) or at 10-months (Coeff = 1.8, 95%CI
−2.10, 5.69, p = 0.36). Outcomes Per Protocol Sample (PPS) did not dif-
fer from ITT analysis. In terms of BDI, we noted a statistically significant 3.4. Secondary outcomes
effect in favour of ACTdp + SC at 5-months (Coeff = −8.38, 95%CI −
15.49, −1.27, p = 0.02) but not at 10-months (Coeff = −4.85, 95%CI We did not observe any other effects on outcomes using the Ques-
−12.10, 2.39, p = 0.18). In the PPS analyses we noted sustained out- tionnaire for Personal Recovery and the Positive and Negative Syn-
comes in favour of ACTdp + SC at both 5-months (Coeff = − 10.18, drome Scale with one exception which was that there was a
95%CI − 17.69, − 2.68, p = 0.01) and 10-months (Coeff = − 8.11, significant effect in favour of ACTdp + SC on PANSS Cognitive
95%CI −15.62, −0.60, p = 0.04). Disorganisation (Coeff = − 4.92, 95%CI − 8.50, −1.35, p = 0.01) and
The outcomes for AAQ and KIMS are described in Table 3. We ob- PANSS Excitement at 5-months (Coeff = 1.98, 95%CI − 3.60, − 0.35,
served a significant effect in favour of ACTdp + SC in terms of improved p = 0.019). In the PPS analyses we noted significant effects in favour
psychological flexibility (reduced AAQ score) at 5-months follow-up of ACTdp + SC for PANSS Negative (Coeff = − 4.03, 95%CI − 7.77,
(Coeff = −0.883, 95%CI −14.94, −2.71, p b 0.01) but not at 10-months − 0.28 p = 0.04), PANSS Cognitive Disorganisation (Coeff = − 6.16,
(Coeff = −4.92, 95%CI −11.09, 1.25, p = 0.11). In the PPS analysis out- 95%CI − 10.01, − 2.30, p = 0.003), PANSS Emotional Distress
comes at 5-months (Coeff = −10.71, 95%CI −16.16, −5.26, p b 0.01) (Coeff = − 2.46, 95%CI − 4.74, − 0.18, p = 0.04) and PANSS Total
and 10-months (Coeff = −6.27, 95%CI −11.72, −0.82, p = 0.03). (Coeff = −12.45, 95%CI −23.11, −1.80, p = 0.02) at 5-months.

Table 3
Therapy specific measures at baseline, 5-months and 10-months.

ACTdp + SC
Variable Statistic ITT (N = 29) SC (N = 14) (N = 15)

AAQ - baseline Nobs (Nmiss) 29 (0) 14 (0) 15 (0)

Mean (SD) 34.8 (7.4) 34.4 (9.5) 35.2 (5.1)
AAQ - 5-months Nobs (Nmiss) 27 (2) 14 (0) 13 (2)
Mean (SD) 32.5 (10.6) 36.4 (10.2) 28.3 (9.7)
AAQ - 10-months Nobs (Nmiss) 25 (4) 13 (1) 12 (3)
Mean (SD) 32.1 (10.3) 33.8 (9.4) 30.2 (11.3)
KIMS Observing Baseline Nobs (Nmiss) 29 (0) 14 (0) 15 (0)
Mean (SD) 36.0 (8.2) 39.4 (7.7) 32.8 (7.5)
KIMS Observing 5-months Nobs (Nmiss) 27 (2) 14 (0) 13 (2)
Mean (SD) 36.7 (8.1) 37.1 (8.8) 36.4 (7.8)
KIMS Observing 10-months Nobs (Nmiss) 25 (4) 13 (1) 12 (3)
Mean (SD) 39.9 (7.3) 39.3 (8.4) 40.6 (6.3)
KIMS Describing Baseline Mean (SD) 21.6 (6.0) 22.8 (7.3) 20.5 (4.6)
KIMS Describing 5-months Mean (SD) 22.9 (5.7) 22.8 (6.4) 23.0 (5.1)
KIMS Describing 10-months Mean (SD) 24.6 (5.7) 24.1 (7.0) 25.2 (4.0)
KIMS Awareness Baseline Mean (SD) 26.1 (6.0) 27.2 (6.0) 25.0 (5.9)
KIMS Awareness 5-months Mean (SD) 27.1 (6.3) 27.9 (7.1) 26.4 (5.5)
KIMS Awareness 10-months Mean (SD) 26.6 (6.1) 26.5 (5.8) 26.7 (6.7)
KIMS Accepting Baseline Mean (SD) 23.8 (7.1) 23.6 (8.4) 24.0 (6.0)
KIMS Accepting 5-months Mean (SD) 25.0 (6.9) 23.8 (7.3) 26.3 (6.4)
KIMS Accepting 10-months Mean (SD) 24.7 (7.2) 22.6 (7.1) 27.0 (6.8)

Please cite this article as: Gumley, A., et al., A parallel group randomised open blinded evaluation of Acceptance and Commitment Therapy for
depression after psychosis: Pilot tri..., Schizophr. Res. (2016), http://dx.doi.org/10.1016/j.schres.2016.11.026
6 A. Gumley et al. / Schizophrenia Research xxx (2016) xxx–xxx

3.5. Mechanisms of change Although the study was not powered or designed as an efficacy or ef-
fectiveness trial we did wish to explore treatment signals and we had
In order to investigate hypothesized mechanisms of change we cal- established a priori primary outcomes, therapy specific outcomes, and
culated change scores at 5-months and 10-months for the CDSS, BDI, secondary outcomes. We did not observe any change signals on the
KIMS and AAQ. We then investigated associations between changes in CDSS. We noted a statistically significant improvement in BDI-II for
depression and changes in mindfulness and psychological flexibility. the ACTdp + SC group at 5-months but not at 10-months. We noted a
Table 4 below summarizes the observed correlations at 5-months and between group Cohen's Effect Size of r = 0.67 at 5-months end of treat-
10-months. Consistent with the lack of observed effects on CDSS we ob- ment and r = 0.31 at 10-months. These effects compare well with those
served no significant associations between depression and hypothe- identified in a recent meta-analysis (Hacker et al., 2016) who found that
sized mechanisms of change at either 5 or 10-months. With respect to studies investigating ACT for depression as a primary treatment target
changes in BDI we observed significant correlations with changes in (n = 12 studies comprising n = 674 participants) suggested also
psychological flexibility at 5-months (r = 0.54, p b 0.01) and 10-months large significant between group effect size (d = 0.73, p b 0.001). For
(r = 0.41, p = 0.04). Furthermore, we observed significant associations ACT versus active control conditions (n = 15 studies comprising n =
between KIMS Observing (r = − 0.58, p b 0.01), KIMS Describing 755 participants) findings suggested a small non-significant between
(r = − 0.63, p b 0.01), KIMS Awareness (r = − 0.41, p = 0.03) and groups effect size (d = 0.26, n.s.).
change in BDI at 5-months. Finally we observed as significant associa- The different patterns of depression change scores on the observer-
tion between KIMS Describing (r = −0.49, p = 0.01), KIMS Awareness rated versus self-report measures have potential implications for future
(r = − 0.53, p b 0.01), KIMS Acceptance (r = − 0.49, p = 0.01) and trials. First, the effect size for depression reduction on the CDSS for the
change in BDI at 10-months. ACT + SC care group was d = 0.30 at the 5 month end of treatment
time-point and d = −0.15 at the 10 month post treatment follow-up
4. Discussion suggesting that the CDSS detected smaller effects. Second, we noted a
lower than expected ICC = 0.68 for the CDSS and this may have contrib-
This is the first randomised controlled trial of any psychological ther- uted to some inconsistency of outcomes. Given, the pilot nature of this
apy in psychosis that specified depression as a primary outcome. We study, it is difficult to draw firm conclusions regarding the inconsistency
sought to map the parameters for a larger scale definitive trial of ACT of these outcomes and therefore the CDSS should be retained in a future
for depression after psychosis. Our primary concern was the identifica- main trial. However, since service users with psychosis express a prefer-
tion of outcome signals in relation to depression (CDSS, BDI-II), and ence for self-report measures as primary trial outcomes (Crawford et al.,
established therapy mechanisms (KIMS and AAQ) as a basis to estimate 2011), any future RCT following from this pilot could justify being
the power and sample size requirements for a future trial. Over a period powered to detect an effect on the BDI-II.
of 12-months recruitment, we received 92 potential enquiries, Consistent with the model of therapeutic change in ACT and with
converting to 55 participant referrals (referrals 4.6 per month) leading earlier observations (White et al., 2011, 2013) we observed statistically
to 29 (2.4 per month) randomisations. Over 10-months we retained significant changes in psychological flexibility and mindfulness. We
25 (86.3%) of participants in follow-up. noted significant associations between psychological flexibility at both
One of the key challenges for the study was the identification and re- 5 and 10-month follow-up using the BDI and associations between
cruitment of potential participants. Depression co-occurring with psy- changes in mindfulness and depression at 5 and 10-month follow-up.
chosis was not routinely identified by mental health staff. Positive Given this replication of earlier findings we propose that the strongest
symptoms tended to be prioritised for monitoring and assessment, signals in relation to ACTdp relate to measures of self reported depres-
there was poor discrimination between negative symptoms and depres- sion, psychological flexibility and mindfulness. Although lack of change
sion, potential participants rarely reported depressed mood (despite se- in objective measures of symptom severity have been reported in previ-
vere symptoms on the BDI-II), and conversations about depressed mood ous trials of ACTp (e.g. Bach and Hayes, 2002) lack of any signal on any
and hopelessness were frequently avoided. In a future large-scale trial, observer based measures in our study is an important consideration.
supporting mental health staff with knowledge and skills to detect We would also argue for the additional inclusion of behaviourally
and respond to depression in this group would increase recruitment. based measures of functioning. This would be in keeping with the
Our results also point to the need to develop pathways to enhance goals of ACTdp to increase behavioural activation through engagement
self-referral for the assessment and treatment of depression. This is par- in valued activities. A future large-scale trial could include the measure-
ticularly important given the role of depression in reduced quality of life ment of daily activities as a key outcome. In their trial of Social Recovery
and serious adverse outcomes such as death by suicide (Foster, 2015; CBT (CBTsr) in first episode psychosis, Fowler et al. (2009) used the
Pompili et al., 2008). Time Use Survey (Short, 2006) as an interview based measure of
The ACTdp intervention was well tolerated. On average participants hours spent in ‘Constructive Economic Activity’ and in ‘Structured Activ-
received 15 sessions over 5-months, rates of cancellation and non-at- ity’. The pilot study data showed signals suggesting that compared to
tendance were low and 70% received at least 10-sessions of therapy. TAU CBTsr was associated with improved economic activity (increase
of 4.4 h versus 3.2 h) and improved structured activity (12.0 h versus
Table 4 4.1 h). These measures can also be supplemented by use of objective
Associations between primary outcomes (change scores CDSS and BDI) and therapy measures of physical activity such as ActiGraphy. For example, using
mechanisms (and respective change scores KIMS and AAQ). actigraphic measurement of daytime routines Wichniak et al. (2011)
KIMS KIMS KIMS KIMS AAQ found that in people with schizophrenia, higher rates of depression
Variable Observing Describing Awareness Acceptance total were associated with lower rates of physical activity and longer time
CDSS −0.11 −0.34 −0.14 −0.31 0.30 spent in bed. This is a highly relevant outcome as a lack of physical
5-months and leisure time activity is linked to poorer quality of life in schizophre-
CDSS 0.06 −0.33 −0.26 −0.33 0.28 nia (Vancampfort et al., 2011).
10-months
Our study had some noteworthy limitations. First, the sample size of
BDI −0.58⁎⁎ −0.63⁎⁎ −0.41⁎⁎ −0.21 0.55⁎⁎
5-months 29 was small. But, the 11 randomised controlled trials included in the
BDI −0.17 −0.49⁎⁎ −0.53⁎⁎ −0.49⁎⁎ 0.41⁎ Whitehead et al. (2002) Cochrane Review of antidepressants in schizo-
10-months phrenia all randomised n b 30 participants. Of the 5 RCTs of ACTp (Bach
⁎ p b 0.05. and Hayes, 2002; Gaudiano and Herbert, 2006; White et al., 2011;
⁎⁎ p b 0.01. Shawyer et al., 2007; Gaudiano et al., 2015) the mean sample size is

Please cite this article as: Gumley, A., et al., A parallel group randomised open blinded evaluation of Acceptance and Commitment Therapy for
depression after psychosis: Pilot tri..., Schizophr. Res. (2016), http://dx.doi.org/10.1016/j.schres.2016.11.026
 A. Gumley et al. / Schizophrenia Research xxx (2016) xxx–xxx 7

40.6 (range 13–80). Gaudiano et al. (2015) conducted a randomised Acknowledgements

controlled trial of an ACT based treatment package for people with af- We would like to acknowledge the support of participants, their
fective psychoses (mostly major depression with psychotic features) families and health service staff in supporting the ADAPT trial.
and found medium to large effects for of ACT on depression and psycho-
social functioning. The study sample size was n = 13, randomization References
was not independent of trial team and assessment of outcome was un- Addington, D., Addington, J., Schissel, B., 1990. A depression rating scale for schizo-
blinded. Taken together these findings indicate a need for large scale, phrenics. Schizophr. Res. 3 (4):247–251. http://dx.doi.org/10.1016/j.schres.2013.12.
well-designed methodologically rigorous trials targeting depression in 014.
Addington, D., Addington, J., Maticka-Tyndale, E., 1993. Rating depression in schizophre-
the context of psychosis. A second limitation is that although our trial
nia: a comparison of a self-report and an observer-report scale. J. Nerv. Ment. Dis. 181
was blinded, our rates of unblinding were high in the ACTdp + SC (9), 561–565.
group. Our RAs were instructed to remind both staff and participants Angermeyer, M.C., Matshinger, H., Corrigan, P.W., 2004. Familiarity with mental illness
and social distance from people with schizophrenia and major depression: testing a
about the importance of blinding. In addition, separate office facilities,
model using data from a representative population survey. Schizophr. Res. 69 (1–
an independent randomization procedure that protected the blind, ban- 2), 175–182.
ning discussions regarding participants between RAs and Trial Thera- Bach, P., Hayes, S.C., 2002. The use of acceptance and commitment therapy to prevent the
pists were instituted throughout the trial. Despite these efforts, the rehospitalization of psychotic patients: a randomized controlled trial. J. Consult. Clin.
Psychol. 70 (5):1129–1139. http://dx.doi.org/10.1037/0022-006X.70.5.1129.
main source of unblindings was participants and less frequently health Baer, R.A., Smith, G.T., Lykins, E., Button, D., Krietemeyer, J., Sauer, S., Walsh, E., Duggan, D.,
service staff. Where an unblinding occurred we were able to allocate an Williams, J.M.G., 2004. Construct validity of the five facet mindfulness questionnaire
alternative blinded researcher to undertake subsequent follow-up as- in meditating and nonmeditating samples. Assessment 15 (3):329. http://dx.doi.org/
10.1177/1073191107313003.
sessments. In addition, since all assessments were audio recorded Beck, A.T., Steer, R.A., Brown, G.K., 1996. Manual for the Beck Depression Inventory-II. Psy-
(with consent) we were able to subject unblinded assessments to inde- chological Corporation, San Antonio.
pendent verification of symptoms severity. Finally, our approach to re- Birchwood, M.J., Iqbal, Z., Chadwick, P., Trower, P., 2000. Cognitive approach to depression
and suicidal thinking in psychosis. Br. J. Psychiatry 177:516–528. http://dx.doi.org/10.
cruitment was informed by an earlier feasibility trial of ACT for 1192/bjp.177.6.522.
emotional distress in psychosis (White et al., 2011). A clear signal Bond, F.W., Hayes, S.C., Baer, R.A., Carpenter, K.M., Guenole, N., Orcutt, H.K., Waltz, T.,
emerging from the trial was a change in depression amongst those par- Zettle, R.D., 2011. Preliminary psychometric properties of the Acceptance and Action
Questionnaire - II: a revised measure of psychological flexibility and experiential
ticipants receiving ACT (White et al., 2013, 2015). All of these steps sug-
avoidance. Behav. Ther. 42 (4):676–688. http://dx.doi.org/10.1016/j.beth.2011.03.
gest that the observed treatment effect on the self-report BDI-II cannot 007.
be easily dismissed as a spurious consequence of unblinding. Similarly, Conley, R.R., 2009. The burden of depressive symptoms in people with schizophrenia.
Psychiatr. Clin. North. Am. 32 (4):853–861. http://dx.doi.org/10.1016/j.psc.2009.09.
the probability that the insignificant effect on the CDSS was due to inad-
001.
vertent contamination of observer ratings is reduced by the steps taken Craig, P., Dieppe, P., Macintyre, S., Michie, S., Nazareth, I., Petticrew, M., 2008. Developing
to deal with unblindings. The main conclusion to draw about this relates and evaluating complex interventions: the new Medical Research Council guidance.
to the importance of future trials having a very robust a priori plan for Br. Med. J. 337:a1655. http://dx.doi.org/10.1136/bmj.a1655.
Crawford, M.J., Robotham, D., Thana, L., Patterson, S., Weaver, T., Barber, R., Wykes, T.,
minimizing the likelihood of penetration of blind treatment allocation. Rose, D., 2011. Selecting outcome measures in mental health: the views of service
This is a particular challenge in psychological treatment studies where users. J. Ment. Health 20 (4):336–346. http://dx.doi.org/10.3109/09638237.2011.
therapy content may “leak” through in the accounts given by patients 577114.
Drake, R., Whitaker, A., Gates, C., Cotton, P., 1985. Suicide among schizophrenics: a review.
to research assessors. For example, ACTdp treated patients may inadver- Compr. Psychiatry 26 (1):90–100. http://dx.doi.org/10.1016/0010-440X(85)90053-
tently report values consistent behavior that has arisen as a conse- 7.
quence of receiving treatment. Drake, R.J., Pickles, A., Nentall, R., Kinderman, P., Haddock, G., Tarrier, N., Lewis, S.W., 2003.
The evolution of insight, paranoia and depression during early schizophrenia.
Depression is a major contributor to loss of quality of life in schizo- Psychol. Med. 34 (2):285–292. http://dx.doi.org/10.1016/j.psychres.2005.10.002.
phrenia and other psychotic disorders (Saarni et al., 2010) and is a Foster, T., 2015. Schizophrenia and bipolar disorder: no recovery without suicide preven-
clear risk factor for suicide (Drake et al., 1985). Depression has been as- tion. Br. J. Psychiatry 207 (5):371–372. http://dx.doi.org/10.1192/bjp.bp.115.165944.
Fowler, D., Hodgekins, J., Painter, M., Reilly, T., Crane, C., Macmillan, I., Mugford, M.,
sociated with poorer adherence to treatment (Conley, 2009), longer du-
Croudace, T., Jones, P.B., 2009. Cognitive behaviour therapy for improving social re-
ration of untreated psychosis (Drake et al., 2003), greater problems in covery in psychosis: a report from the ISREP MRC trial platform study improving so-
family relationships (Rocca et al., 2005), poorer interpersonal function- cial recovery in early psychosis. Psychol. Med. 39:1627–1636. http://dx.doi.org/10.
1017/S0033291709005467.
ing (Rocca et al., 2005), and greater experiences of stigma (Angermeyer
Gaudiano, B.A., Herbert, J.D., 2006. Acute treatment of inpatients with psychotic symp-
et al., 2004). The development of future large scale generalizable toms using Acceptance and Commitment Therapy: pilot results. Behav. Res. Ther.
randomised controlled trials to improve depression and functioning in 44 (3):415–437. http://dx.doi.org/10.1016/j.brat.2005.02.007.
people with schizophrenia and other psychoses is clearly merited. ACT Gaudiano, B.A., Busch, A.M., Wenze, S.J., Nowlan, K., Eptein-Lubow, G., Miller, I.W., 2015.
Acceptance-based Behaviour Therapy for depression with psychosis: results from a
is one clear candidate treatment that has the potential to improve en- feasibility randomized controlled trial. J. Psychiatr. Pract. 21 (5), 320–333.
gagement with valued life goals and activities alongside feelings of de- Gumley, A.I., Braehler, C., Laithwaite, H., Macbeth, A., Gilbert, P.A., 2010. Compassion fo-
pression and demoralization. cused model of recovery after psychosis. Int. J. Cogn. Psychotherapy 3 (2):186–201.
http://dx.doi.org/10.1521/ijct.2010.3.2.186.
Gumley, A.I., White, R.G., Briggs, A., Ford, I., Barry, S., Stewart, C., Beedie, S., Clarke, C.,
Macleod, R., Lidstone, E., Nam, J., McLeod, H., 2015. A parallel group randomized
Contributors
open blinded evaluation of Accept and and Commitment Therapy for Depression
AG was Chief Investigator, RW supervised the therapists, RW, AB, IF and HMcL were
after psychosis: a pilot trial protocol (ADAPT). Psychosis 8 (2):143–155. http://dx.
Principal Investigators, CS, SB, JMcT, and CC recruited participants and undertook assess- doi.org/10.1080/17522439.2015.1100669.
ments, RMacL & EL delivered the experimental interventions, SB, BSR and RY undertook Hacker, T., Stone, P., Macbeth, A., 2016. Acceptance and Commitment Therapy – Do we
statistical analyses of the data. All authors contributed to the writing of the manuscript. know enough? A sequential meta-analysis of randomized controlled trials. J. Affect.
Disord. 15:551–565. http://dx.doi.org/10.1016/j.jad.2015.10.053.
Jauhar, S., McKenna, P.J., Radua, J., Fung, E., Salvador, R., Laws, K.R., 2014. Cognitive-behav-
Role of funding source ioural therapy for the symptoms of schizophrenia: systematic review and meta-anal-
The authors acknowledge the financial support of NHS Research Scotland (NRS), ysis with examination of potential bias. Br. J. Psychiatry 204 (1):20–29. http://dx.doi.
through The Chief Scientist Office (CZH/4/743), NHS Greater Glasgow & Clyde org/10.1192/bjp.bp.112.116285.
(GN11CP470) and of the Scottish Mental Health Research Network. The research project Kay, S., Fiszbein, A., Opler, L., 1987. The Positive and Negative Syndrome Scale (PANSS) for
was adopted and supported by the Scottish Mental Health Research Network. schizophrenia. Schizophr. Bull. 13 (2):262–276. http://dx.doi.org/10.1093/schbul/13.
2.261.
Kim, S.W., Kim, S.-J., Yoon, B.-H., Kim, J.-M., Hwang, M., Yoon, J.-S., 2006. Diagnostic valid-
ity of assessment scales for depression in patients with schizophrenia. Psychiatry Res.
Conflict of interests 144 (1):57–63. http://dx.doi.org/10.1016/j.psychres.2005.10.002.
Mintz, A.R., Dobson, K.S., Romney, D.M., 2003. Insight in schizophrenia: a meta-analysis.
None. Schizophr. Res. 61 (1):75–88. http://dx.doi.org/10.1016/S0920-9964(02)00316-X.

Please cite this article as: Gumley, A., et al., A parallel group randomised open blinded evaluation of Acceptance and Commitment Therapy for
depression after psychosis: Pilot tri..., Schizophr. Res. (2016), http://dx.doi.org/10.1016/j.schres.2016.11.026
8 A. Gumley et al. / Schizophrenia Research xxx (2016) xxx–xxx

Neil, S., Pitt, L., Kilbride, M., Welford, M., Nothard, S., Sellwood, W., Morrison, T., 2009. The White, R.G., 2015. Treating depression in psychosis: self-compassion as a valued life direc-
questionnaire about the process of recovery (QPR): a measurement tool developed in tion. In: Gaudiano, B.A. (Ed.), Incorporating Acceptance and Mindfulness Into the
collaboration with service users. Psychosis 1 (2):145–155. http://dx.doi.org/10.1080/ Treatment of Psychosis: Current Trends and Future Directions. Oxford University
17522430902913450. Press, Oxford. ISBN: 9780199997213, pp. 81–107.
Pompili, M., Lester, D., Innamorati, M., Tatarelli, R., Girardi, P., 2008. Assessment and treat- White, R.G., Gumley, A.I., McTaggart, J., Rattrie, L., McConville, D., Cleare, S., Mitchell, G.,
ment of suicide risk in schizophrenia. Expert. Rev. Neurother. 8 (1):51–74. http://dx. 2011. A feasibility study of Acceptance and Commitment Therapy for emotional dys-
doi.org/10.1586/14737175.8.1.51. function following psychosis. Behav. Res. Ther. 49 (12):901–907. http://dx.doi.org/10.
Rocca, P., Bellino, S., Calvarese, P., Marchiaro, L., Patria, L., Rasetti, R., Bogetto, F., 2005. De- 1016/j.brat.2011.09.003.
pressive and negative symptoms in schizophrenia: different effects on clinical fea- White, R.G., Gumley, A.I., McTaggart, J., Rattrie, L., McConville, D., Cleare, S., Mitchell, G.,
tures. Compr. Psychiatry 46 (4):304–310. http://dx.doi.org/10.1016/j.comppsych. 2013. Depression and anxiety following psychosis: associations with mindfulness
2004.09.001. and psychological flexibility. Behav. Cogn. Psychother. 41 (1):34–51. http://dx.doi.
Saarni, S.I., Viertiö, S., Perälä, J., Koskinen, S., Lönnqvist, J., Suvisaari, J., 2010. Quality of life org/10.1017/S1352465812000239.
of people with schizophrenia, bipolar disorder and other psychotic disorders. Br. White, R.G., Gumley, A.I., McTaggart, J., Rattrie, L., McConville, D., Cleare, S., McLeod, H.J.,
J. Psychiatry 197 (5):386–394. http://dx.doi.org/10.1192/bjp.bp.109.076489. Mitchell, G., 2015. Acceptance and commitment therapy for depression following
Shawyer, F., Ratcliffe, K., Mackinnon, A., Farhall, J., Hayes, S.C., Copolov, D., 2007. The psychosis: an examination of clinically significant change. J Contextual Behav. Sci. 4
voices acceptance and action scale (VAAS): pilot data. J. Clin. Psychol. 63 (6): (3):203–209. http://dx.doi.org/10.1016/j.jcbs.2015.06.004.
593–606. http://dx.doi.org/10.1002/jclp.20366. Whitehead, C., Moss, S., Cardno, A., Lewis, G., Furtado, V.A., 2002. Antidepressants for peo-
Short, S., 2006. Review of the UK 2000 Time Use Survey. Office for National Statistics, ple with both schizophrenia and depression. Cochrane Database Syst. Rev. 2
London. (CD002305). http://dx.doi.org/10.1002/14651858.CD002305.
Trial Managers Network, 2014. A Guide to Efficient Trial Management. National Institute Wichniak, A., Skowerska, A., Chonjnacka-Wojtowicz, J., Taflinski, T., Wierzbicka, A.,
for Health Research, pp. 1–76. Jernajczyk, W., Jerema, M., 2011. Actigraphic monitoring of activity and rest in schizo-
van der Gaag, M., Hoffman, T., Remijsen, M., Hijman, R., de Haan, L., van Meijel, B., van phrenic patients treated with olanzepine or risperidone. J. Psychiatr. Res. 45 (10):
Harten, P.N., Valmaggia, L., de Hart, M., Cuipers, A., Wiersma, D., 2006. The five-factor 1381–1386. http://dx.doi.org/10.1016/j.jpsychires.2011.05.009.
model of the Positive and Negative Syndrome Scale II: a ten-fold cross-validation of a Wykes, T., Steel, C., Everitt, B., Tarrier, N., 2008. Cognitive behavior therapy for schizophre-
revised model. Schizophr. Res. 85 (1–3):280–287. http://dx.doi.org/10.1016/j.schres. nia: effect sizes, clinical models, and methodological rigor. Schizophr. Bull. 34 (3):
2006.03.021. 523–537. http://dx.doi.org/10.1093/schbul/sbm114.
Vancampfort, D., Probst, M., Schweewe, T., Maurissen, K., Sweers, K., Knapen, J., De Hart,
M., 2011. Lack of physical activity during leisure time contributes to an impaired
health related quality of life in patients with schizophrenia. Schizophr. Res. 129 (2–
3):122–127. http://dx.doi.org/10.1016/j.schres.2011.03.018.

Please cite this article as: Gumley, A., et al., A parallel group randomised open blinded evaluation of Acceptance and Commitment Therapy for
depression after psychosis: Pilot tri..., Schizophr. Res. (2016), http://dx.doi.org/10.1016/j.schres.2016.11.026