'Third wave' cognitive and behavioural therapies versus treatment as usual for depression

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims. The quote only partly states this claim.

Sixty-four studies were excluded from the whole MAP programme; five had a third wave CBT arm, four compared third wave CBT with treatment as usual or a non-active control, and these four were excluded for inpatient, cognitive-impairment, or non-depressed samples.

“A total of 64 studies were excluded from the whole Meta-Analysis of Psychotherapies (MAP) programme of reviews”

• cites Hopko 2003aNot matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• cites Bohlmeijer 2011 sourceNot checked — no full text of the source
• cites Reynolds 2011 sourcePartly supported

  Split into 2 atoms — 0 confirmed, 1 partially supported, 0 disputed, 0 misattributed, 1 unverifiable. Expand the claim to see per-atom verdicts.

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

The March 2013 update excluded four additional studies because their samples did not meet the review's diagnostic inclusion criteria.

“In the updated search conducted in March 2013 (CCDANCTR to 01/02/2013), four additional studies were excluded because the diagnoses did not fully meet the inclusion criteria of the review programme.The samples in each study consisted of subclinical depression (Kaviani 2012), partial remission (Korrelboom 2012), mixed population (Pinniger 2012) and chronic depression (Folke 2012).”

• cites Kaviani 2012Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• cites Korrelboom 2012 sourcePartly supported

  The Patients section identifies Korrelboom 2012 as a study that included participants in partial remission: 21 of its 61 patients had that diagnosis. It does not support describing the whole sample as consisting of patients in partial remission; the remaining participants had other diagnoses or disease states. The March 2013 search and exclusion decision belong to the citing review, not this trial. (LLM, conf=0.88)

  Experimental — accuracy not yet measured
• cites Pinniger 2012 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

The importance of an observed I-squared value depends on the size and direction of treatment effects and on the strength of evidence for heterogeneity.

“However, the importance of the observed I 2 depended on the magnitude and direction of treatment effects and the strength of evidence for heterogeneity (Higgins 2003;Deeks 2008).”

• cites Higgins 2003 sourceSupported
• cites Deeks 2008Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S3 (load-bearing).

Barth 2013's systematic review and network analysis assessed seven psychological therapies for depression.

“A recently published systematic review and network analysis examined the efficacy of seven psychological therapies for individuals with depression (Barth 2013); however, third wave CBT interventions were not managed in a separate category, therefore no comparable findings were reported.”

• cites Barth 2013 sourceSupported

May combine two claims.

Barth 2013 included both internet-delivered and face-to-face therapies, and studies whose participants had a primary medical condition.

“It is notable, too, that the authors included Internet-delivered therapies and those delivered face to face, together with studies of participants whose primary disorder was a medical condition, resulting in a more heterogeneous set of populations and interventions than those included in the current review.”

• cites Barth 2013 sourceSupported

Probably not a substantial claim: may not pass S1 (asserted), S3 (load-bearing), S4 (based).

Four small studies with 224 participants were included in the review.

“Four small studies (224 participants) were included in the review.”

Probably not a substantial claim: may not pass S3 (load-bearing).

The fourth study compared ACT with a psychological placebo.

“The fourth study compared ACT with a psychological placebo condition.”

May combine two claims.

At post-treatment, mean depression levels were 1.12 standard deviations lower in intervention groups (95% CI -1.53 to -0.71), based on 4 studies and 211 participants, with very low quality evidence.

“Mean depression levels at post-treatment in the intervention groups were 1.12 standard deviations lower (1.53 to 0.71 lower) 211 (4 studies) ⊕⊝⊝⊝ very low a,d,f,g SMD -1.12 (-1.53 to -0.71)”

May combine two claims.

The review included four studies with a total of 224 participants.

“The review included four studies, involving a total of 224 people.”

May combine two claims.

The included studies evaluated three third wave CBT forms: extended behavioural activation (two studies), ACT (one study), and competitive mind training (one study).

“The studies examined three different forms of third wave CBT, consisting of extended behavioural activation (two studies), acceptance and commitment therapy (ACT) (one study) and another form of third wave CBT called competitive mind training (one study).”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims. The quote only partly states this claim.

Among the 122 eligible studies, seven had third wave CBT arms; three were excluded because they compared third wave CBT with other psychological therapies, and four were included in this review.

“Of those 122 studies, seven studies had third wave CBT arms.Three of these studies were not eligible for inclusion in the current review because they compared third wave CBT approaches with other psychological therapy approaches (Zettle 1984;Zettle 1989;Dimidjian 2004).”

• Also stated in: § Results of the search · plain

  “The remaining four studies were included in the current review (Pellowe 2006;Gawrysiak 2009;Ekers 2011;Ekkers 2011).”

• cites Zettle 1984Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• cites Zettle 1989 sourceNot checked — no full text of the source
• cites Dimidjian 2004Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S3 (load-bearing).

All four studies reported baseline depression severity, using the Geriatric Depression Scale or the BDI-II.

“The baseline severity of depression was reported in each of the four studies, using the Geriatric Depression Scale (Ekkers 2011) and the BDI-II (Pellowe 2006;Gawrysiak 2009;Ekers 2011).”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims. The quote only partly states this claim.

Two studies took place in a non-medical university setting, one in primary care, and one in an outpatient psychiatric institute; two were in the USA, one in the Netherlands, and one in the UK.

“Two studies were conducted in a non-medical university setting (Pellowe 2006; Gawrysiak 2009), one was conducted in primary care (Ekers 2011) and one was conducted in an outpatient psychiatric institute (Ekkers 2011).”

• Also stated in: § Setting · plain

  “Two studies were conducted in the USA (Pellowe 2006; Gawrysiak 2009), one was conducted in the Netherlands (Ekkers 2011) and one in the UK (Ekers 2011).”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

All four studies reported dropout rates, and the single-session Gawrysiak 2009 intervention had no dropouts.

“Dropout rates were reported by all four studies.The intervention by Gawrysiak 2009 was a single session with no dropouts.”

May combine two claims.

Two studies compared third wave CBT with usual care or treatment as usual, one used a no-treatment condition, and one used a psychological placebo supportive therapy condition.

“Two studies compared a third wave CBT intervention against usual care/treatment as usual”

• Also stated in: § Comparisons · plain

  “One study used a no treatment condition (Gawrysiak 2009), and one study used a psychological placebo 'supportive therapy' condition, in which therapists facilitated an unstructured group discussion (Pellowe 2006).”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Two studies included university-based young adults, one included adults over 18, and one included adults aged 65 or older with a mean age of 72.7 years.

“Two studies included young adults based at a university (Pellowe 2006; Gawrysiak 2009), one study included adults over the age of 18 (Ekers 2011) and one study involved adults of 65 years or older, with a mean age of 72.7 years (Ekkers 2011).”

Probably not a substantial claim: may not pass S3 (load-bearing).

The four included studies had sample sizes ranging from 30 to 93 participants.

“The overall sample size in each of the four studies ranged from 30 participants (Gawrysiak 2009) to 93 participants (Ekkers 2011).”

Probably not a substantial claim: may not pass S3 (load-bearing).

One study was limited to a single-session intervention with college students reporting mild depression.

“One study limited to single-session intervention with college students reporting mild depression.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Two studies used allocation concealment independent of the research team, while the other two did not report their allocation methods.

“Two of the studies conducted the allocation concealment process independent of the research team (Ekers 2011).The other two studies did not report on the methods used for allocating participants to groups.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Depression diagnosis methods varied: one study used CIS-R with ICD-10 criteria, one used a clinically established MDD diagnosis, and two used BDI-II thresholds.

“One study used a computerised version of the revised Clinical Interview Schedule (CIS-R) to make an assessment of depression according to ICD-10 criteria (Ekers 2011).One study used a clinically established diagnosis of major depressive disorder (MDD), which was agreed for each potential participant through discussion at team meetings (Ekkers 2011).Both of the other studies used a BDI-II score of”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

After duplicate removal, the search found 6710 relevant records; 6524 were excluded by title and abstract, 186 full texts were read, and 122 studies were eligible for the review programme.

“A er removing duplicates, we identified 6710 records that were relevant for the programme of reviews on all psychological therapies for depression.We excluded 6524 records on the basis of information provided in the titles and abstracts.We read the full text of 186 studies to assess their eligibility.A total of 122 studies were judged as eligible for inclusion in the programme of reviews (Figure 1”

Probably not a substantial claim: may not pass S3 (load-bearing).

The proportion of female participants ranged from 62% to 80% across the studies.

“The proportion of female participants in the studies ranged from to 62% (Ekers 2011) to 80% (Gawrysiak 2009).”

May combine two claims.

Three categories of third wave CBT were examined: two studies used extended behavioural activation, one used acceptance and commitment therapy, and one used competitive memory training.

“Three categories of third wave CBT were examined in the four included studies.Two studies examined forms of extended behavioural activation (Gawrysiak 2009; Ekers 2011), one study examined acceptance and commitment therapy (ACT) (Pellowe 2006) and one study (Ekkers 2011) examined competitive memory training (COMET)”

May combine two claims.

Only one single-session study did not permit naturalistic antidepressant prescribing; in the other three studies, 8% to 79% of participants received naturalistic antidepressants.

“Only one study, a single-session intervention, did not allow naturalistic prescribing of antidepressants (Gawrysiak 2009).”

• Also stated in: § Comparisons · plain

  “In the other three studies, the number of participants prescribed antidepressants on a naturalistic basis ranged from 8% (Pellowe 2006) to 79% of the sample (Ekkers 2011).”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Two studies were awaiting classification: a Farsi-language mindfulness trial awaiting translation and a single-session extended behavioural activation trial focused on religious behaviours.

“Two studies are awaiting classification.The study by Azargoon 2010, written in Farsi, compares mindfulness-based therapy with a control group and is awaiting translation.The study by Armento 2012 compares single-session extended behavioural activation focused on religious behaviours with a non-active 'supportive therapy' control.”

Probably not a substantial claim: may not pass S3 (load-bearing).

None of the four studies provided information on sequence generation.

“No information was provided on sequence generation across the four studies.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims. The quote only partly states this claim.

The March 2013 update found 151 new references; 142 were ineligible, three were ongoing-study protocols, six full texts were read, all compared third wave CBT with a non-active control, and two studies were awaiting classification.

“A total of 151 new references were identified.On the basis of the information provided in abstracts, 142 references were found not to be eligible.Three references reported on protocols for ongoing studies that appear to meet the criteria for the third wave CBT reviews”

• Also stated in: § Results of the search · plain

  “All of the studies compared a third wave CBT with a non-active control condition.Two studies were assessed as potentially eligible for this review and are awaiting classification.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Therapists included mental health nurses, registered CBT therapists, doctoral students, and advanced clinical psychology graduate students.

“The therapists were mental health nurses (Ekers 2011), registered CBT therapists (Ekkers 2011), doctoral students (Gawrysiak 2009) and advanced clinical psychology graduate students (Pellowe 2006).”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Three studies were excluded from this review because they compared third wave CBT with another psychological therapy; they were included in the companion third wave CBT review.

“In addition, three studies (Zettle 1984;Zettle 1989;Dimidjian 2004) were not included in the current review because they compared third wave CBT with another psychological therapy approach.These three studies are included in the HIRED companion third wave CBT review (Hunot 2012).”

• cites Zettle 1984Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• cites Zettle 1989 sourceNot checked — no full text of the source
• cites Dimidjian 2004Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S3 (load-bearing).

Treatment length ranged from a single session to 12 sessions over 3 months.

“Treatment length varied from a single session to 12 sessions over 3 months.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Of three ongoing studies meeting inclusion criteria, one compares dialectical behaviour therapy with a waiting-list control in people with major depressive disorder and suicidal ideation.

“Of three ongoing studies that meet the inclusion criteria for the third wave CBT reviews, one compares a third wave CBT approach with a non-active treatment control.This study compares the effect of dialectical behaviour therapy with a waiting list control in a sample of participants with major depressive disorder and suicidal ideation (NCT01441258).”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims. The quote only partly states this claim.

Three studies reported post-treatment clinical response; all three used the Jacobson and Truax reliable-change procedure, and Ekers 2011 also used additional response measures.

“Three of the studies provided clinical response figures at posttreatment.All three studies (Gawrysiak 2009; Ekers 2011; Ekkers 2011) used the Jacobson and Truax procedure to calculate reliable change.”

• Also stated in: § Comparisons · plain

  “Ekers 2011 used two additional measures of response”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Three studies used the BDI-II for continuous depression change, while Ekkers 2011 used the QIDS-SR and the Geriatric Depression Scale.

“Three studies used the BDI-II to measure continuous change in depression scores (Pellowe 2006; Gawrysiak 2009; Ekers 2011).Ekkers 2011 used the Quick Inventory of Depression Symptoms Self Report (QIDS-SR) and the Geriatric Depression Scale (GDS).”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

All four included studies used a parallel design; two were single-centre and two were multicentre.

“The four studies all used a parallel design.Two of the studies were single centre (Pellowe 2006; Gawrysiak 2009), and two were multicentre (Ekers 2011;Ekkers 2011).”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims. The quote only partly states this claim.

All studies used treatment manuals; sessions ranged from one 90-minute session to 12 sessions over three months, and no study conducted follow-up assessments.

“All of the studies used manuals to guide the intervention.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

One study used a single-session intervention, resulting in no treatment dropouts.

“One study used a single-session intervention; therefore no dropouts from treatment.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Because only one subgroup was present for this outcome, subgroup difference testing was not possible.

“Only one subgroup was included in this outcome; therefore a test for subgroup differences could not be conducted.”

May combine two claims.

All analyses included at most four studies, three with small samples, producing wide confidence intervals and insufficient statistical power.

“All analyses included no more than four studies, three of which had small sample sizes, resulting in wide confidence intervals and lack of statistical power to asses the effects of third wave CBT approaches.”

The quote only partly states this claim.

Anxiety levels at post-treatment significantly favoured third wave CBT over TAU in one study with 30 participants (MD -5.50, 95% CI -10.01 to -0.99), with very low evidence quality.

“A significant difference in anxiety levels favoured third wave CBT when compared with TAU conditions (30 participants, MD -5.50, 95% CI -10.01 to -0.99).”

The quote only partly states this claim.

Social adjustment levels at post-treatment significantly favoured third wave CBT over TAU in one study with 38 participants (MD -11.56, 95% CI -17.89 to -5.23), with very low evidence quality.

“A significant difference in social adjustment levels favoured third wave CBT compared with TAU conditions (38 participants, MD -11.56, 95% CI -17.89 to -5.23).”

May combine two claims.

Clinical response rates significantly favoured third wave CBT over TAU/usual care/no treatment in three studies with 170 participants (RR 0.51, 95% CI 0.27 to 0.95), with substantial heterogeneity (I2=63%).

“A significant difference in clinical response rates favoured third wave CBT compared with TAU/usual care/no treatment (3 studies, 170 participants, RR 0.51, 95% CI 0.27 to 0.95).Substantial statistical heterogeneity was indicated (I 2 = 63%; see Analysis 1.1).”

May combine two claims. The quote only partly states this claim.

Clinical response at post-treatment was significantly better for third wave CBT than for TAU conditions, based on three studies with 170 participants (RR 0.51, 95% CI 0.27 to 0.95); heterogeneity was substantial (I2=63%) and evidence quality very low.

“A significant difference in clinical response rates favoured third wave CBT compared with TAU conditions (3 studies, 170 participants, RR 0.51, 95% CI 0.27 to 0.95).Substantial statistical heterogeneity was indicated (I 2 = 63%; see Analysis 1.1).”

• Also stated in: § Additional information · hedged

  “Results showed evidence of a difference in favour of third wave CBT compared with TAU for efficacy at posttreatment (RR of clinical response 0.51, 95% CI 0.27 to 0.95) based on 3 studies involving a total of 170 participants.”

May combine two claims.

The four studies differed in participant age, baseline depression severity, type and duration of third wave CBT, and comparators, so substantial statistical heterogeneity after combining data was not surprising.

“It is clear that the four studies included in the review were diverse in terms of participant age (ranging from a student to an older population), level of depression severity at baseline, type and duration of third wave CBT under examination and the control comparators used.Therefore, it is not surprising that substantial statistical heterogeneity was observed when the study data were combined.”

May combine two claims.

Dropout-based treatment acceptability did not differ significantly between third wave CBT and TAU/usual care/no treatment in three studies with 170 participants (RR 0.71, 95% CI 0.03 to 16.07), with substantial heterogeneity (I2=92%).

“No significant difference in treatment acceptability based on dropout rates was found between third wave CBT and TAU/usual care/no treatment (3 studies, 170 participants, RR 0.71, 95% CI 0.03 to 16.07).Substantial statistical heterogeneity was indicated (I 2 = 92%; see Analysis 2.2).”

May combine two claims. The quote only partly states this claim.

Remission rates at post-treatment significantly favoured third wave CBT over TAU in two studies with 140 participants (RR 0.77, 95% CI 0.67 to 0.88), with very low evidence quality.

“A significant difference in clinical response rates favoured third wave CBT compared with TAU (140 participants, RR 0.77, 95% CI 0.67 to 0.88; see Analysis 1.3).”

• Also stated in: § Additional information · plain

  “Analysis of secondary outcomes revealed an effect in favour of third wave CBT approaches compared with TAU conditions for remission (2 studies, 140 participants, RR 0.77, 95% CI 0.67 to 0.88)”

May combine two claims. The quote only partly states this claim.

Depression levels at post-treatment significantly favoured third wave CBT over TAU across four studies with 211 participants (SMD -1.12, 95% CI -1.53 to -0.71), with moderate heterogeneity (I2=44%) and very low evidence quality.

“A significant difference in depression levels favoured third wave CBT compared with TAU conditions (211 participants, SMD -1.12, 95% CI -1.53 to -0.71) based on completers' data (Ekers 2011), full dataset (Gawrysiak 2009) and LOCF data (Pellowe 2006;Ekkers 2011).Moderate statistical heterogeneity was indicated (I 2 = 44%; see Analysis 1.4).”

• Also stated in: § Additional information · plain

  “Analysis of secondary outcomes revealed an effect in favour of third wave CBT approaches compared with TAU conditions for remission (2 studies, 140 participants, RR 0.77, 95% CI 0.67 to 0.88) and for depression levels (4 studies, 211 participants, SMD -1.12, 95% CI -1.53 to -0.71) at post-treatment.”

May combine two claims.

Moderate to substantial statistical heterogeneity appeared in all analyses and remained unexplained because few studies were included.

“Moderate to substantial statistical heterogeneity was observed in all analyses and remained unexplained because of the small number of studies included in the review.”

May combine two claims. The quote only partly states this claim.

Dropout-based treatment acceptability at post-treatment did not differ significantly between third wave CBT and TAU conditions across four studies with 224 participants (RR 1.01, 95% CI 0.08 to 12.30); heterogeneity was substantial (I2=86%) and evidence quality very low.

“No significant difference in treatment acceptability based on dropout rates was found between third wave CBT approaches and TAU conditions (RR 1.01, 95% CI 0.08 to 12.30).Substantial statistical heterogeneity was indicated (I 2 = 86%; see Analysis 1.2).”

• Also stated in: § Additional information · hedged

  “No evidence indicated a difference between third wave CBT and TAU control conditions for acceptability at post-treatment based on all 4 studies, involving 224 participants (RR 1.01, 95% CI 0.08 to 12.30).”

Probably not a substantial claim: may not pass S1 (asserted), S3 (load-bearing), S4 (based).

All four studies were at high risk of performance bias because participants and therapists could not be blinded.

“As it is not possible to blind participants and therapists in psychological therapy trials, the studies were all at high risk of performance bias.”

• AssumptionRests on an assumption: Participants and therapists cannot be blinded in psychological therapy trials.

May combine two claims.

Risk of bias was judged high for researcher allegiance and unclear for therapist qualifications.

“The risk of bias was assessed as high for researcher allegiance and as unclear for therapist qualifications.”

May combine two claims.

The evidence quality was very low due to few studies/participants, diverse client groups, interventions, control conditions, and possible researcher allegiance, making confident conclusions difficult.

“However, the quality of evidence was very low because of the small number of studies/participants included in the review, together with the diverse client groups, interventions and control conditions used and possible allegiance of researchers towards the active treatments, making it difficult to draw conclusions with any confidence.”

May combine two claims.

The limited number of studies, varied interventions and comparators, and methodological weaknesses prevent firm conclusions about short- or long-term effects.

“The very small number of available studies and the diverse types of interventions and control comparators, together with methodological limitations, limit the ability to draw any conclusions on their effect in the short term or over a longer term.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Eligibility disagreements were discussed with a third review author, and study authors were consulted if needed, until consensus was reached.

“Conflicts of opinion regarding eligibility of a study were discussed with a third review author a er the full paper had been retrieved and consultation with the study authors sought, if necessary, until consensus was reached.”

Probably not a substantial claim: may not pass S3 (load-bearing).

The review quantified heterogeneity using the I-squared statistic.

“Heterogeneity was also quantified using the I 2 statistic”

Classified as probably not a substantial claim.

The P-value threshold for the Chi-squared heterogeneity test was set at 0.1 because the test has low power with few participants or trials.

“Because the Chi 2 test has low power to assess heterogeneity when a small number of participants or trials are included, the P value was conservatively set at 0.1.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

The review also searched CINAHL in May 2010 and PSYNDEX in June 2010.

“In addition to CCDANCTR, we also searched CINAHL in May 2010 and PSYNDEX in June 2010 (see Appendix 2).”

Probably not a substantial claim: may not pass S3 (load-bearing).

Two review authors independently extracted data from each study.

“Data from each study were extracted independently by two review authors.”

Classified as probably not a substantial claim.

Full articles of all studies identified by either review author were obtained and inspected by the same two authors against the inclusion criteria.

“Full articles of all studies identified by either of the review authors were then obtained and inspected by the same two review authors to identify trials meeting the following criteria.”

Classified as probably not a substantial claim.

NICE guidelines recommend MBCT for preventing depression in patients with three or more prior episodes.

“the updated National Institute for Health and Care Excellence treatment guidelines for depression (NICE 2009) have already recommended MBCT specifically for preventing depression in patients who have had three or more episodes of depression.”

• cites NICE 2009Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

May combine two claims.

Using follow-up data from waiting-list-controlled studies could introduce bias because the intervention was not delivered as originally allocated at randomisation and participants might no longer be comparable.

“Inclusion of follow-up data could introduce bias, as the intervention was not provided as originally allocated at the point of randomisation, and participants might no longer be comparable.”

May combine two claims.

Gawrysiak 2009 had 0% dropout and Pellowe 2006 had very low dropout with reasons; both were assessed as low risk of attrition bias.

“Gawrysiak 2009 examined a single-session intervention with a 0% dropout.Pellowe 2006 reported a very low dropout rate and provided reasons for dropout.These two studies were assessed as being at low risk of attrition bias.”

Probably not a substantial claim: may not pass S3 (load-bearing).

I-squared values from 50% to 90% were pre-specified as substantial statistical heterogeneity.

“I 2 values in the range of 50% to 90% were considered to represent substantial statistical heterogeneity”

Probably not a substantial claim: may not pass S3 (load-bearing).

When missing outcomes were found, the review authors attempted to obtain available data directly from the trial authors.

“When we found evidence of missing outcomes, we attempted to obtain any available data direct from the authors.”

Probably not a substantial claim: may not pass S1 (asserted), S3 (load-bearing).

Funnel plots were planned when enough trials were available to assess reporting biases and small-study effects.

“funnel plots were to be constructed to establish the potential influence of reporting biases and small-study effects.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Two review authors screened abstracts of all publications found by the search.

“Two review authors (RC and VH) examined the abstracts of all publications obtained through the search strategy.”

Probably not a substantial claim: may not pass S3 (load-bearing).

The searches had no date, language or publication-status restrictions.

“No restriction on date, language or publication status was applied to the searches.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Subject experts were contacted to verify that all relevant published and unpublished studies had been considered.

“Subject experts were contacted to check that all relevant studies, published and unpublished, had been considered for inclusion.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

The review searched two CCDAN clinical trial registers in June 2010, with updates in April 2011 and February 2012, using a broad list of search terms.

“We searched two clinical trials registers created and maintained by the Cochrane Depression, Anxiety and Neurosis Group (CCDAN) -the CCDANCTR-Studies Register and the CCDANCTR-References Register-in June 2010, and updated searches were carried out in April 2011 and February 2012”

Probably not a substantial claim: may not pass S3 (load-bearing).

A mindfulness-therapy-specific website was searched.

“A website related specifically to mindfulness-based therapies (http://www.mindfulexperience.org/)was searched.”

May combine two claims.

All four studies used self-report outcome measures, so detection bias was assessed as high.

“Each of the four studies used self-report outcome measures; therefore the risk of detection bias was assessed as high.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Outcome reporting bias was checked by recording all planned and reported trial outcomes and noting missing outcomes.

“We tried to identify outcome reporting bias in trials by recording all trial outcomes, planned and reported, and noting where outcomes were missing.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Allocation concealment was used as a trial-quality marker for sensitivity analyses.

“Allocation concealment was used as a marker of trial quality for the purposes of undertaking sensitivity analyses.”

Probably not a substantial claim: may not pass S3 (load-bearing).

The review formally tested statistical heterogeneity with the Chi-squared test.

“Statistical heterogeneity was formally tested using the Chi 2 test”

Probably not a substantial claim: may not pass S3 (load-bearing).

The tau-squared estimate was used to derive an approximate range of intervention effects for the primary outcomes.

“we also used the tau 2 estimate to determine an approximate range of intervention effects using the method outlined in Section 9.5.4 of the Cochrane Handbook for Systematic Reviews of Interventions (Deeks 2008).”

• cites Deeks 2008Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Data-extraction disagreements were discussed with an additional review author, and study authors were contacted when needed.

“Any disagreement was discussed with an additional review author, and, when necessary, the authors of the studies were contacted for further information.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Findings from Jacobson 1996 suggested that behavioural components alone might be as effective as CBT.

“The findings from Jacobson 1996, a component analysis trial of CBT, suggested that behavioural components alone might work just as well as CBT.”

• cites Jacobson 1996 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

Blinding of clinicians and participants was not possible in psychological therapy trials.

“As with all psychological therapy trials, blinding of clinicians/participants was not achievable.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Two review authors independently assessed risk of bias in selected studies.

“Two review authors independently assessed the risk of bias in selected studies.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Risk-of-bias disagreements were discussed with a third review author, and study authors were contacted when needed.

“Any disagreement was discussed with a third review author.When necessary, study authors were contacted for further information.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Relevant review papers were also checked.

“Relevant review papers were checked.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Reference lists of all selected studies were searched for additional published reports and unpublished-study citations.

“The references of all selected studies were searched for additional published reports and citations of unpublished studies.”

Classified as probably not a substantial claim.

Control condition descriptions in each included study were checked to confirm they were not an active psychological therapy.

“In each study, descriptions of the control conditions were scrutinised to ensure that they did not comprise an active psychological therapy treatment.”

May combine two claims.

Sample sizes were small to very small with wide confidence intervals.

“Small to very small sample sizes with wide confidence intervals.”

Probably not a substantial claim: may not pass S3 (load-bearing).

CBT is recommended as the first-line psychological therapy for depression.

“An extensive evidence base is available on the effectiveness of CBT, which is recommended as the first-line psychological therapy approach for depression (NICE 2009).”

• cites NICE 2009Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Classified as probably not a substantial claim. May combine two claims.

NICE guidelines recommend behavioural activation for moderate to major depressive disorder but note that its evidence base is less robust.

“Although these guidelines also recommend the use of behavioural activation (BA) for moderate to major depressive disorder, they acknowledge that the evidence for BA is currently less robust.”

• cites NICE 2009Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S3 (load-bearing).

Small-study effects and potential publication bias were examined with funnel plots when appropriate.

“when appropriate, small-study effects and potential publication bias were examined using funnel plots.”

Classified as probably not a substantial claim.

A random-effects model was used in all analyses because of potential heterogeneity across therapy approaches and likely differences in comorbid mental disorders.

“Given the potential heterogeneity of psychological therapy approaches for inclusion, together with the likelihood of differing secondary comorbid mental disorders in the population of interest, a random-effects model was used in all analyses.”

• AssumptionRests on an assumption: Potential heterogeneity across psychological therapy approaches and differing comorbid mental disorders justify using a random-effects model.

Probably not a substantial claim: may not pass S3 (load-bearing).

External subject or methodological experts were consulted when necessary.

“External subject or methodological experts were consulted as necessary.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

A systematic review of 17 RCTs found CBT and behavioural therapy equivalent in depression recovery rates, symptom levels, and dropout.

“Prompted by continuing debate in this area, a recent systematic review of 17 randomised controlled trials (RCTs) demonstrated equivalence between CBT and behavioural therapy in terms of depression recovery rates, symptom levels and participant dropout (Ekers 2008).”

• cites Ekers 2008 sourceNot checked — no full text of the source

Classified as probably not a substantial claim.

Risk of bias was assessed for each included study with The Cochrane Collaboration's 'Risk of bias' tool.

“Risk of bias was assessed for each included study using The Cochrane Collaboration's 'Risk of bias' tool (Higgins 2008a).”

• cites Higgins 2008aNot matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Classified as probably not a substantial claim.

The effectiveness and acceptability of third wave CBT for acute depression remain unclear.

“However, the effectiveness and acceptability of third-wave cognitive and behavioural therapy (CBT) approaches as treatment for acute depression remain unclear.”

• cites Ost 2008 sourceNot checked — no full text of the source

Classified as probably not a substantial claim. May combine two claims.

Extended behavioural activation was categorised as third wave because its added rumination-focused strategies distinguished it from earlier behavioural activation and aligned it with third wave approaches.

“we decided that the addition of behavioural strategies for targeting rumination, including 'an emphasis on the function of ruminative thinking and on moving attention away from the content of ruminative thoughts towards direct, immediate experience' (Dimidjian 2004) in extended BA set it apart from earlier BA approaches and placed it more in line with third wave approaches.”

• cites Dimidjian 2004Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

May combine two claims.

Therapists in the four studies had diverse qualifications and inadequate experience information, leading to an overall unclear risk of bias.

“The therapists used in the four included studies consisted of a diverse group of clinicians in terms of background qualifications, and information provided on their level of experience was inadequate; hence an overall assessment of unclear risk of bias was made.”

• AssumptionRests on an assumption: Therapist qualifications and experience are a potential source of bias in psychological therapy studies.

Probably not a substantial claim: may not pass S3 (load-bearing).

Little information was available about how participants were allocated to groups in the included studies.

“Little information was provided about the process of allocating participants to groups”

May combine two claims.

All four studies described random assignment, but none fully reported sequence generation and allocation concealment; missing information in two studies created uncertainty about allocation bias.

“Each of the four studies included in this review described its assignment procedure as 'randomised'; however, none of the studies provided a full account of the sequence generation and allocation concealment methods used.The complete lack of information on these procedures as followed in two studies introduces considerable uncertainty as to whether bias may have been introduced during the”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

The review team held regular meetings to reach full consensus on categorising approaches and contacted authors when information was unclear.

“We have held regular team meetings to ensure 100% consensus regarding the categorisation of approaches, and when uncertainties continued because of lack of information provided in the articles, we have made contact with the authors to request a fuller description.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Competitive memory training was classified within the Other category to adhere to the review's a priori classifications.

“we classified competitive memory training, a newly developed third Cochrane Database of Systematic Reviews wave approach, within the 'Other' category to ensure that we adhered fully to our a priori classifications.”

Not yet classified.

Ekkers 2011 reported that an independent rater observed a sample of sessions but gave no further details.

“One study stated that a sample of sessions was observed by an independent rater but gave no further information (Ekkers 2011).”

Probably not a substantial claim: may not pass S3 (load-bearing).

A recent study indicates that people with residual symptoms may benefit regardless of how many previous depression episodes they have had.

“It is notable that a recently published study indicates that those with residual symptoms may also benefit, irrespective of the number of previous depression episodes (Geschwind 2012), and highlights the importance of conducting trials involving participants with acute depression symptoms.”

• cites Geschwind 2012Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

May combine two claims.

Supportive therapy was re-categorised as a psychological placebo when participants saw it as active but investigators treated it as a control, leading to Pellowe 2006 being included in this review.

“However a er careful discussion by the review team, it was decided that when supportive therapy interventions were regarded as an active treatment by participants but as a control treatment by study investigators, they would be re-categorised as a 'psychological placebo' for inclusion in the TAU reviews of the MAP programme.Hence, Pellowe 2006 was included in the current review.”

May combine two claims.

Pellowe 2006 and Ekers 2011 were thorough on fidelity, Ekkers 2011 lacked detail on observation methods, and Gawrysiak 2009 used therapist self-assessment; overall fidelity bias was unclear.

“Two studies were thorough in their consideration of fidelity and applied appropriate criteria (Pellowe 2006;Ekers 2011).The study by Ekkers 2011 used several appropriate methods in measuring fidelity but lacked information on the method used in observing sessions and on how many sessions were assessed.The study by Gawrysiak 2009 relied upon the therapists' own assessment of their treatment”

• AssumptionRests on an assumption: Independent monitoring of therapist fidelity to treatment manuals is a key methodological requirement for attributing observed treatment effects to specific therapy components.

May combine two claims.

The evidence base is of very low quality, which limits conclusions about the efficacy of third wave CBT as individual or collective approaches.

“However, whilst the findings from this review appear to suggest that third wave CBT approaches are more effective than TAU conditions in treating the acute symptoms of depression, the very low quality of the evidence base limits the ability to draw conclusions on their efficacy, either as individual approaches or as a collective approach.”

May combine two claims.

All four studies were exposed to researcher allegiance bias, and the potential for such allegiance was considered high across them.

“Each of the studies included in the review were exposed to this form of bias, as authors were involved in development of the therapy (Ekkers 2011), had developed a variation of an already established treatment (Gawrysiak 2009; Ekers 2011) or expressly favoured the active therapy against a psychological placebo approach (Pellowe 2006).Overall, therefore, the potential for researcher allegiance”

• AssumptionRests on an assumption: Researcher allegiance is a common source of bias in psychological therapy trials.

May combine two claims.

The other two studies had differential dropout, including 50% in the TAU arm of Ekkers 2011, and neither gave reasons; both were assessed as high risk of attrition bias.

“The other two studies had differential dropout rates between experimental and control conditions.One study reported a dropout rate of 50% in the TAU arm (Ekkers 2011), and neither study provided reasons for dropout; therefore both were assessed as being at high risk of attrition bias.”

May combine two claims.

Neither Cuijpers 2008 nor Ekers 2008 distinguished the extended BA model from the pure BT approach.

“Two systematic reviews have examined the efficacy of behavioural activation treatments (Cuijpers 2008;Ekers 2008), but neither attempted to differentiate between the extended BA model developed by Martell 2001 and the 'pure' BT approach originally evaluated in the seminal trial by Jacobson 1996; therefore these findings are not comparable with those of the current review.”

• cites Cuijpers 2008 sourceNot checked — no full text of the source
• cites Ekers 2008 sourceNot checked — no full text of the source
• cites Martell 2001Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S3 (load-bearing).

The four studies provided insufficient information to identify other consistent sources of bias.

“Insufficient information was provided across the four studies to allow identification of other consistent sources of bias.”

May combine two claims.

Pellowe 2006 showed potential therapist allegiance bias, Ekers 2011 therapists were deemed not to have allegiance, and the other two studies gave insufficient information.

“In the study by Pellowe 2006, the researcher was also the primary therapist, and potential bias towards the active ACT treatment was indicated”

• Also stated in: § Therapist allegiance/conflict of interest · plain

  “In the study by Ekers 2011, the therapists had not received primary psychotherapy training and therefore were deemed not to have allegiance to the experimental condition.The other two studies gave insufficient information to permit a decision to be made.”

May combine two claims.

In three studies the manualiser was a team member, and Pellowe 2006 implied allegiance to the experimental condition; overall researcher allegiance bias was assessed as high.

“In three of the four studies, the person responsible for manualising the experimental condition was a member of the research team (Gawrysiak 2009;Ekers 2011;Ekkers 2011).In the study by Pellowe 2006, the supportive therapy condition was designed 'as an attention control group rather than an active treatment such as CBT', thereby implying an allegiance towards the experimental condition.The overall”

• AssumptionRests on an assumption: Researcher allegiance is a common source of bias in psychological therapy trials.

May combine two claims.

Gawrysiak 2009 used therapist self-checklists without independent checks and was assessed as high risk of fidelity bias.

“The fourth study required therapists to check off a list of therapy components as they conducted the session; no independent checks were carried out (Gawrysiak 2009), and the study was deemed to be at high risk of bias.”

• AssumptionRests on an assumption: Independent monitoring of therapist fidelity to treatment manuals is a key methodological requirement for attributing observed treatment effects to specific therapy components.

May combine two claims.

Pellowe 2006 and Ekers 2011 monitored sessions with audiotapes or independent raters and a rating scale, and were assessed as low risk of fidelity bias.

“Two of the studies included in the review monitored therapy sessions using audiotapes/independent raters together with a rating scale to assess fidelity (Pellowe 2006; Ekers 2011); they were assessed as being at low risk of bias.”

• AssumptionRests on an assumption: Independent monitoring of therapist fidelity to treatment manuals is a key methodological requirement for attributing observed treatment effects to specific therapy components.