Methods

Duration: 12 weeks

Follow‐up assessment points: Post‐intervention, 28 weeks

Funded by: NHS Health Technology Assessment (HTA) Programme, Central Manchester and Manchester Children’s University Hospitals

Participants

N = 208

Adolescents only (11 to 17 years)

Depression diagnoses included: DSM‐IV; criteria for major or probable major depression (four symptoms with psychosocial impairment). Participants also had to obtain a score of 7 or more on the Health of the Nation Outcome scales for children and adolescents (HoNOSCA; Gowers 1999)

Baseline risk of suicide: Measured using the suicidality items from the Kiddie Schedule for Affective Disorders and Schizophrenia Present and Lifetime Version (K‐SADS‐PL; Kaufman 1997). Patients with active suicidal intent were included in the study

Baseline               Fluoxetine + CBT n = 105 (%)           Fluoxetine  n = 103 (%)

Thoughts                             50 (47.6)                               48 (46.6)

Ideation                               40 (38.1)                               44 (42.7)

Acts                                     13 (12.4)                               21 (20.4)

Medical lethality               3 (2.9)                                   4 (3.9)

Self harm                            30 (28.6)                               23 (22.3)

Baseline severity of depression: Children’s Depression Rating Scale‐Revised (CDRS‐R; Poznanski 1996). Mean t‐score (SD):  Fluoxetine + CBT = 75.1 (6.7)  Fluoxetine = 75.3 (6.7)

Comorbidity included:

Comorbidity                                      Fluoxetine + CBT (n = 105)               Fluoxetine (n = 103)

Social Phobia                                                 43                                                  49

Obsessive compulsive disorder                 42                                                           37

Post‐traumatic stress disorder                    42                                                           36

Agorophobia                                                 36                                                           29

Separation anxiety disorder                        31                                                           28

Specific phobia                                              25                                                           22

Conduct disorder                                          18                                                           17

Panic disorder                                                21                                                          14

Oppositional defiance disorder                  17                                                           13

Generalised anxiety disorder                     19                                                           13

Panic disorder (with agoraphobia)            20                                                           13

ADHD                                                                 5                                                             6

Bulemia Nervosa                                               8                                                             4

Alcohol abuse                                                   1                                                             4

Transient tic disorder                                     2                                                             3

Tourettes syndrome                                      2                                                             2

Alcohol dependence                                      1                                                             2

Encopresis                                                         1                                                             0

Enuresis                                                             1                                                             0

Dysthymia                                                          1                                                             0

Age: Range = 11 to 17 years  

Fluoxetine + CBT (median) = 14

Fluoxetine (median) = 14                        

Sex (M:F):

Total: 54:154

Fluoxetine + CBT = 26:79

Fluoxetine = 28:75                   

Setting: Outpatient setting

Excluded psychiatric diagnoses: Schizophrenia or bipolar disorder; global learning disability (formal testing not undertaken)

Country: UK

Interventions

Combination (Fluoxetine+CBT)

N = 105

Name: CBT with core interventions including engagement and goal setting, emotional recognition, self monitoring, self reinforcement and activity scheduling, challenging negative thinking and cognitive restructuring, social problem‐solving and communication skills

 # sessions/length: 19 sessions over 28 weeks. (1 session per week for 12 weeks, 1 session per fortnight for 12 weeks, 1 final session at 28 weeks)

Manualised (Y/N): Yes

Individual or group: Individual

Parent involvement: Encouraged at the end of each session by therapist

Fidelity check: Yes. Audiotapes of the session were rated with a modified version of the cognitive therapy scale (Vallis 1986) Inter‐rater reliability k = 0.8

Delivered by: 4 Psychiatrists who either had previous CBT training or attended a 3‐day training course on CBT for depression, and 10 CBT therapists (mostly Psychologists)

Name (class & type): SSRI (Fluoxetine). However, 26 participants were taking a different SSRI when admitted to the trial; 3 switched to fluoxetine and 11 changed from fluoxetine to another SSRI

Dose (mg/day)/length: 10 mg daily for 1 week, increasing to 20 mg for 5 weeks. If no response, increase considered to 40 mg on alternate days for one week followed by 5 weeks of 40 mg. Option to increase dose to 60 mg on alternate days for 1 week followed by 60 mg daily for 5 weeks if participant did not respond by 12 weeks. Overall, there was a mean dose of 30 mg for both groups, and 2 patients received 60 mg

Delivered by: psychiatrists in the context of ongoing clinical care. The content of contact was an explanation of depression and attention to recent family or peer group conflicts. Liaison with schools and other agencies undertaken when appropriate. Participants offered 9 outpatient sessions of usual care over 28 weeks, with the option of more if needed

Medication Only

N = 103

Medication details as above Y/N: Yes

Outcomes

Clinician reported

The Children’s Depression Rating Scale‐Revised (CDRS‐R; Poznanski 1996)

Children’s Global Assessment Scale (C‐GAS; Shaffer 1983)

Suicidality items from the Kiddie Schedule for Affective Disorders and Schizophrenia Present and Lifetime Version (K‐SADS‐PL; Kaufman 1997)

Self reported

The Mood and Feelings Questionnaire (MFQ; Wood 1995)

Parent reported

The Clinical Global Impression Improvement Scale (CGI‐I; Guy 1976)

Additional Measures

The Health of the Nation Outcome Scales for Children and Adolescents (HoNOSCA; Gowers 1999)

The Clinical Global Impression Improvement Scale (CGI‐I; Guy 1976)

Notes

Dropouts during treatment to any or at least 1 adverse reaction: 1 participant experienced a fit possibly related to SSRI and 1 had an allergic reaction (possibly secondary to medication)

Suicide‐related outcome as an adverse event of treatment: 4 required admission for suicidality or self harm and were withdrawn from the study

Authors only report median age

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

“Participants were randomised to SSRI alone or SSRI plus CBT by an equal allocation ratio using stochastic minimisation balancing for severity (Childrens Global Assessment Scale <40), centre, sex, concurrent comorbidity disorder, and age”

pg. 2/8 (Under heading Assignment)

Allocation concealment (selection bias)

Low risk

“Research staff from the clinical sites enrolled patients, and an independent telephone randomisation centre allocated treatment”

pg. 2 /8 (Under heading Assignment)

Blinding (performance bias and detection bias)
Assessors

Low risk

 “...research assistants blind to treatment assignment assessed outcome”

pg. 4/8 (Under heading Outcomes)

Blinding (performance bias and detection bias)
Participants

High risk

No placebo or control psychotherapy was used. As such, participants would be aware that the medication was active and the therapy was CBT

Incomplete outcome data (attrition bias)

Low risk

ITT analysis: “Analysis was by intention to treat subject to the availability of the data”

pg. 4/8 (Under heading Statistical Analysis)

Number randomised:

Fluoxetine + CBT: 105          Fluoxetine:  103     Total: 208

Number of dropouts during intervention:

Fluoxetine + CBT: 11             Fluoxetine:6            Total: 17

Number dropouts in follow‐up:

Fluoxetine + CBT: 7              Fluoxetine: 7           Total: 14

Number analysed post‐intervention:

Fluoxetine + CBT: 105         Fluoxetine: 103      Total:  208

Number analysed follow‐up 1:  

Fluoxetine + CBT: 105         Fluoxetine: 103      Total:  208

Reasons for dropout in each group:

12 patients were formally withdrawn from the study for the following reasons: 4 required admission to hospital for suicidality or self harm, 5 failed to improve, 1 had a fit, 1 had an allergic reaction, 1 was prescribed paroxetine by a GP

18 families withdrew participants from the study: 6 were improving and did not want further treatment, 5 did not want more treatment, 2 wanted CBT, 2 did not want CBT, 1 wanted a female therapist, 1 was getting worse, 1 moved

Selective reporting (reporting bias)

Unclear risk

Authors reported data for all outcomes specified in their methods. Do not have access to trial protocol

Other bias

Low risk

Methods

Duration: 8 weeks

Follow‐up assessment points: Post‐intervention, 12 months

Funded by: National Institute of Mental Health (NIMH)

Participants

N = 63

Adolescent only (12 to 18 years)

Depression diagnoses included: DSM‐III‐R Major Depressive Disorder (MDD). Participants also had to obtain a score of 35 or more on the CDRS‐R (Poznanski 1996)

Baseline risk of suicide: Not measured

Baseline severity of depression: Chidren’s Depression Rating Scale (CDRS‐R; Poznanski 1985). Mean score (SD):

Imipramine + CBT = 46.8 (9.5)

Placebo + CBT = 52.5 (10.8)

Comorbidity included: All 63 subjects met criteria for at least 1 anxiety disorder based on either adolescent or parental interviews

Age mean (SD):

Total = 13.9 (3.6)        

Sex (M:F): 25:38                

Setting: Unclear. Likely an outpatient setting based on information regarding medication monitoring throughout the trial.

Excluded psychiatric diagnoses: ADHD, conduct disorder, bipolar disorder, eating disorder, alcohol or drug abuse on the Diagnostic Interview for Children and Adolescents‐Revised‐Adolescent Version (DICA‐R‐A) or Parent Version (DICA‐R‐P; Reich 1990), or both, mental retardation by history, bipolar affective disorder in first degree relative

Country: USA

Interventions

Combination (Imipramine + CBT)

N = 31

Name: CBT. Based on school refusal treatment by Last 1998. Included the identification of negative thoughts surrounding school attendance and teaching adaptive coping strategies.

# sessions/length: 8 (45 to 60 minutes) sessions over 8 weeks

Manualised (Y/N): Yes

Individual or group: Individual

Parent involvement: Yes. Parents joined each session for 10 to 15 minutes at the end

Fidelity check: No formal check. Weekly discussions with all therapists and principal investigators, and a fortnightly telephone consultation with an expert on CBT for school refusal

Delivered by: 3 therapists (1 behaviorally trained Clinical Psychologist, 1 Doctoral level therapist and 1 Masters level therapist)

Name (class & type): TCA  (Imipramine)

Dose (mg/day)/length: Dose based on body weight. A gradual increase every 3 to 5 days to 3 mg/kg per day by the end of week 2 Mean dose at week 3 was 184.6mg + 33.3

Delivered how: Weekly appointments monitoring side effects, and compliance were undertaken with a psychiatrist. Blood imipramine levels were monitored at 3 and 8 weeks

Combination (Placebo medication + CBT)

N = 32

Details as above (Y/N): Yes

Outcomes

Clinician reported

Children’s Depression Rating Scale‐Revised (CDRS‐R; Poznanski 1985) with a score of ≤ 35

Children’s Depression Rating Scale‐Revised (CDRS‐R; Poznanski 1985)

Self reported

The Beck Depression Inventory (BDI; Beck 1979)

Additional Measures

Anxiety Rating for Children‐Revised (ARC‐R; Bernstein 1996)

The Revised Children’s Manifest Anxiety Scale (RCMAS; Reynolds 1978)

Weekly school attendance rates

Notes

Dropouts during treatment to any or at least 1 adverse reaction: 1 participant developed manic symptoms and 1 developed psychiatric symptoms and required hospitalisation

Suicide‐related outcome as an adverse event of treatment: No

Denominator and numerator for remission rates calculated from percentages reported in the publication

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

High risk

“Random assignment to treatment was blocked on gender and whether subjects had a school vacation that lasted 5 or more days during the 8 week treatment period”. (Under heading Procedure)

Allocation concealment (selection bias)

Unclear risk

No information contained in paper to make a judgement

Blinding (performance bias and detection bias)
Assessors

Low risk

“All project personnel...were blind to medication assignment”. (Under heading Procedure)

Blinding (performance bias and detection bias)
Participants

Low risk

"...imipramine pills and matching placebo"

“To preserve the blind, increases and decreases were also suggested for randomly selected patients on placebo”. (Under heading Medication Management)

Incomplete outcome data (attrition bias)

Low risk

ITT analysis: “All randomized subjects were included in analyses based on intent to treat”. (Under heading Statistical Analyses)

Number randomised:

Imipramine + CBT: 31        Placebo + CBT:  32          Total: 63

Number of dropouts during intervention

Imipramine + CBT: 7           Placebo + CBT: 9            Total: 16

Number analysed post‐intervention:

Imipramine + CBT:  31        Placebo + CBT:32           Total: 63

Reasons for dropout in each group: 1 missed 22 does of medication, 1 missed 2 therapy appointments, 1 developed manic symptoms on study medication, 1 required hospitalisation for psychiatric symptoms, and 12 declined further participation

Selective reporting (reporting bias)

Unclear risk

Authors report data for all outcomes post‐intervention. Do not have access to trial protocol

Other bias

High risk

Authors note that placebo group significantly more symptomatic at baseline compared with imipramine group despite randomisation

Methods

Duration: 6 weeks

Follow‐up assessment points: Post‐intervention, 12, 26 and 52 weeks

Funded by: The Agency for Healthcare Research and Quality and the Garfield Memorial Fund

Participants

N = 152

Adolescent only (12 to 18 years)

Depression diagnoses included: DSM‐IV episode of major depression

Baseline risk of suicide: 73.7% (112/152) of participants reported significant levels of suicidal behaviour; assessment tool not explicitly referenced

Baseline severity of depression: Centre for Epidemiological Studies ‐ Depression Scale (CES‐D; Radloff 1977):

TAU + CBT = 35.4 (11.8)

TAU = 33.7 (9.3)

Comorbidity included: Not reported

Age mean (SD):

Total = 15.30 (1.61)    

TAU + CBT = 15.29 (1.62)

TAU = 15.32 (1.60)                                          

Sex (M:F): 34:118               

TAU + CBT = 17:60          

TAU = 17:58                                        

Setting: Primary care health maintenance organization (HMO)

Excluded psychiatric diagnoses: Schizophrenia or a significant developmental/intellectual disability

Country: USA

Interventions

TAU (SSRI) + CBT

N = 77

Name: CBT employing cognitive restructuring, or behavioural training, or both. Participants able to choose which type to try first. After completion of first module (2 to 5 sessions), therapist and youth reviewed recovery and decided whether to proceed with the second module (sessions 6 to 9), focusing on skills training

 # sessions/length: Between 0 and 9, mean 5.3 sessions. Each session 1 hour. Weekly in frequency

Manualised (Y/N): No information 

Individual or group: Individual

Parent involvement: Clinicians organised separate parent meetings, however "parents' attendance was "sparse"

Fidelity check: Yes. All sessions audio taped. 57 sessions selected at random and rated by a senior supervisor. 87.2% adherence to protocol

Delivered by: Masters level Psychologists

Name (class & type): SSRI (varied). All trial participants were able to receive any medications provided by either the HMO or outside providers

Dose (mg/day)/length: Varied

Delivered by: No information

TAU (SSRI)

N = 75

Details as above (Y/N): Yes

Outcomes

Clinician reported

Mood disorders module of the Schedule for Affective Disorders and Schizophrenia for School‐age Children‐Present and Lifetime version (K‐SADS‐PL; Kaufman 1997) and the Longitudinal Interview Follow‐Up Evaluation (Keller 1982).

This was used to define remission i.e. those who had did not have a continuing or new mood disorder since the last interview according to the K‐SADS‐PL). It was unclear if DSM‐IV or ICD time criteria were employed

Centre for Epidemiological Studies‐Depression Scale (CES‐D; Radloff 1977), cut‐off of ≤ 15

Children’s Global Adjustment Scale (C‐GAS; Shaffer 1983)

Self reported

Centre for Epidemiological Studies‐Depression Scale (CES‐D; Radloff 1977)

Parent reported

The Child Behaviour Checklist (CBCL; Achenbach 1978)

Additional Measures

Hamilton Rating Scale for Depression (HAM‐D; Hamilton 1960)

Youth Self Report (YSR; Achenbach 1991)

Internalising and Externalising subscales and an extracted depression subscale created to match DSM criteria for major depression (Clarke 1992)

Social Adjustment Scale Self Report for Youth (Weissman 1980)

Short Form‐12 (Ware 1998)

Notes

Authors do not report reasons for dropout

Dropouts during treatment to any or at least 1 adverse reaction: Not reported

Suicide‐related outcome as an adverse event of treatment: Not reported

Numbers who reached remission by interview were calculated by review authors using percentages based on depressive episodes (Table 3). Data from Table 3 were based on observed cases not ITT following advice from statistician

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

“Youths were randomized using a blocked procedure to minimize study arm imbalance”. pg. 889

Allocation concealment (selection bias)

Unclear risk

No information contained in paper to make a judgement

Blinding (performance bias and detection bias)
Assessors

Low risk

”Blinded interviewers assessed each adolescent and a participating parent by telephone at baseline and at 6, 12, 26 and 52 weeks post‐randomization”. pg. 890

Blinding (performance bias and detection bias)
Participants

High risk

No placebo or therapy control arm. As such, participants were aware if they were receiving CBT in the trial or not

Incomplete outcome data (attrition bias)

Unclear risk

ITT analysis: "...all subjects were considered part of the study from the point of randomisation (an intent‐to‐treat design)" pg.890

“We examined continuous depression and functioning outcome measures using random effect regression analysis”. pg. 892

Number randomised:

TAU + CBT: 77                       TAU: 75                      Total: 152

Number completing post‐intervention: (dropouts)

TAU + CBT: 67 (10)                TAU: 65 (10)              Total: 132 (20)

Number completing follow‐up 12 weeks:

TAU + CBT: 61 (16)                 TAU: 61 (14)              Total: 122 (30)

Number completing follow‐up 26 weeks:

TAU + CBT: 65 (12)                  TAU: 62 (13)             Total: 127 (25)

Number completing follow‐up 52 weeks:

TAU + CBT: 56 (21)                   TAU: 58 (17)             Total: 114 (38)

*Data obtained from Fig 1. Summary of study procedures. Number analysed not clearly stated in paper

Reasons for dropout in each group: Not reported

Selective reporting (reporting bias)

Unclear risk

Remission rates only reported at 52 weeks. Do not have access to trial protocol

Other bias

High risk

Authors note that telephone administration of self report measures may have created bias

Methods

Duration: 12 weeks

Follow‐up assessment points: Post‐intervention (12 weeks)

Funded by: National Institute on Alcohol Abuse and Alcoholism

Participants

N = 50

Adolescent only (15 to 20 years)

Depression diagnoses included: DSM‐IV diagnosis of major depressive disorder (MDD) 

Baseline risk of suicide: Not measured and suicidality not stated as an exclusion criteria

Baseline severity of depression: Hamilton Rating Scale for Depression (HAM‐D; Hamilton 1960) Mean (SD):

CBT + fluoxetine = 16.88 (7.09)

CBT + placebo = 22.88 (8.79)

Comorbidity included: All participants were required to have a DSM‐IV diagnosis of an alcohol use disorder (AUD) confirmed using the Substance Use Disorders Section of the Structured Clinical Interview for the DSM (SCID)

Age mean: Not reported            

Sex (M:F):          

Total = 28:22 

CBT + fluoxetine = 12:12   

CBT + placebo = 16:10                       

Setting: Outpatient?

Psychiatric diagnoses excluded: DSM‐IV diagnosis of bipolar disorder, schizoaffective disorder, or schizophrenia, persons with and substance abuse or dependence other than nicotine dependence or cannabis use and dependence, persons with a history of intravenous drug use, persons who had received antipsychotic or antidepressant medication within 1 month prior to baseline assessment also excluded

Country: USA

Interventions

Combination: Psychotherapy + Medication

N = 24

Name (description): CBT for depressive disorder and the treatment of alcohol use disorder combined with Motivation Enhancement Therapy (MET) for the treatment of alcohol use disorder

 # sessions/length: 9 sessions over 12 weeks

Manualised (Y/N): Yes

Individual or group: Not reported

Parent involvement: Not reported

Fidelity check: No fidelity check reported

Delivered by: Not reported

Medication

Name (class & type): SSRI; fluoxetine

Dose (mg/day)/length: initiated at 10 mg, increased to 20 mg after week 2 until the end of the study, as 20 mg was target dose of the study

Delivered how: Study physicians prescribed all medication 

Combination: Psychotherapy + Placebo

N = 26

Delivered how: Pill placebo delivered in the same context as above

Outcomes

Clinician reported

Hamilton Rating Scale for Depression (HAM‐D; Hamilton 1960)

Self reported

Beck Depression Inventory (BDI; Beck 1988)

Additional Measures

Drinking behaviour measured using the Timeline Follow‐back Method (TLFB; Sobell 1988)

Notes

Dropouts during treatment to any or at least 1 adverse reaction: 0

Suicide‐related outcome as an adverse event of treatment: 0

Suicidality was not measured with a formalised tool

3 dropouts during study from placebo group due to persistent depressive symptoms

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

“Patient randomisation was conducted by urn randomisation stratified by gender” pg. 906 (Treatment and assessment)

Allocation concealment (selection bias)

Low risk

“Active medication and matching placebo were prepared by the research pharmacy” pg. 906 (Treatment and assessment)

Blinding (performance bias and detection bias)
Assessors

High risk

“The study was conducted in a double blind fashion, though one study physician remained non‐blinded in order to handle any problems which may have arisen” pg. 906 (Assessment and treatment)

Blinding (performance bias and detection bias)
Participants

Low risk

”...participants were randomly assigned to receive fluoxetine or placebo administered in identical‐looking opaque capsules” pg. 906 (Assessment and treatment)

Incomplete outcome data (attrition bias)

Unclear risk

“Statistical analyses were completed on an intent‐to‐treat study group” pg. 907 (Statistical Analyses)

Number randomised:

CBT + fluoxetine: 24 CBT + placebo: 26

Number dropped out during intervention:

CBT + fluoxetine: 0 CBT + placebo: 3

Number analysed post‐intervention:

CBT + fluoxetine: 24 CBT + placebo: 26

Selective reporting (reporting bias)

Unclear risk

Do not have access to trial protocol

Other bias

Low risk

Baseline imbalance of HAM‐D and BDI scores with fluoxetine group have significantly lower baseline depression scores

Methods

Duration:12 weeks

Follow‐up assessment points: Post‐intervention

Funded by: National Institute of Alcohol and Alcoholism (NIAAA)

Participants

N = 10

Adolescent only (15 to 18 years)

Depression diagnoses included: Not clearly stated. The Child Schedule for Affective Disorders and Schizophrenia (K‐SADS; Chambers 1985) was used to assess psychiatric disorders

Baseline risk of suicide: Not measured

Baseline severity of depression: measured using the Hamilton Rating Scale for Depression (HAM‐D; Hamilton 1960)

Sertraline + CBT = 20.40 (5.55)

Placebo + CBT = 20.80 (5.45)

Comorbidity included: All participants presented with an alcohol use disorder

Age mean (SD): Total = 16.6 (0.52)

Sertraline + CBT = 16.4 (0.55)

Placebo + CBT = 16.8 (0.45)     

Sex (M:F): 8:2 

Sertraline + CBT = 4:1   

Placebo + CBT = 4:1

Setting: Outpatient

Excluded psychiatric diagnoses: Not reported

Country: USA

Interventions

Combination (Sertraline+CBT)

N = 5

Name (description): CBT focusing on relapse prevention, coping skills, anger management, modelling and role playing

 # sessions/length: 12, average attendance was 8.2 sessions and 10.6 sessions for the placebo and sertraline groups respectively

Manualised (Y/N): No

Individual or group: Group

Parent involvement: Not reported

Fidelity check: Not reported

Delivered by: A psychiatrist, on a weekly basis

Name (class & type): SSRI (Sertraline)

Dose (mg/day)/length: 25 mg/day, increased to 25 mg weekly, to a maximum dose of 100 mg in about 4 weeks

Delivered by: A psychiatrist monitored side effects, made medication adjustments, and supplied participants with additional medication on a weekly basis

Combination (Placebo medication + CBT)

N = 5

Details as above (Y/N): Yes

Outcomes

Self reported

Outcome 4: Hamilton Rating Scale for Depression (HAM‐D; Hamilton 1960)

Additional Measures

The Time Line Follow Back (TLFB; Sobell 1988) assessed alcohol use

Notes

Dropouts during treatment to any or at least 1 adverse reaction: No. Authors note that all the side effects of sertraline were transient and did not lead to any dropouts

Suicide‐related outcome as an adverse event of treatment: No

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

“Following the baseline assessments, subjects were randomized using a computer‐generated randomisation table into sertraline or placebo groups” pg. 462

Allocation concealment (selection bias)

Low risk

All of the medication supplied by the study pharmacist were identical in appearance" pg. 462

Blinding (performance bias and detection bias)
Assessors

Unclear risk

“This study was a 12 week double blind, placebo‐controlled trial” pg. 462

Blinding (performance bias and detection bias)
Participants

Low risk

Participants in both arms were blind to medication. Both received CBT

“This study was a 12 week double blind, placebo‐controlled trial” pg. 462

Incomplete outcome data (attrition bias)

Unclear risk

ITT analysis: All subjects randomised were included in the final analysis. No information is reported regarding imputation method for missing data

Number randomised:

SSRI + CBT: 5            Placebo + CBT: 5                 Total: 10

Number of dropouts during intervention:*

SSRI + CBT: 2            Placebo + CBT: 0                 Total: 2

Number analysed post‐intervention:

SSRI + CBT: 5            Placebo + CBT:5                 Total: 10

Treatment completion was defined a priori as 8 sessions

Selective reporting (reporting bias)

Unclear risk

All outcome data specified in methods was reported. Do not have access to trial protocol

Other bias

High risk

Small study sample, and no follow‐up

Methods

Duration: 8 weeks

Follow‐up assessment points: postintervention; 4 weeks post intervention

Funded by: Korean Game Culture Foundation

Participants

N = 72

Adolescent only (13 to 18 years)

Depression diagnoses included: Major Depressive Disorder diagnosed using The Korean Child Schedule for Affective Disorders and Schizophrenia (K‐SADS; Chambers 1985); score of 19 or more on the BDI

Baseline risk of suicide: Not measured

Baseline severity of depression: measured using the Beck Depression Inventory (Beck 1961)

Buproion + CBT = 32.7 (SD 8.8)

Bupropion = 33.3 (SD 8.7)

Comorbidity included: Excluded other psychiatric disorders

Age mean (SD): Total = 16

Bupropion + CBT = 16.2 (1.4)

Bupropion = 15.9 ± 1.6

Sex (M:F): 72:0

Sertraline + CBT = 35:0

Placebo + CBT = 37:0

Setting: University clinic

Excluded psychiatric diagnoses: all other psychiatric disorders excluded

Country: South Korean

Interventions

Medication (Bupropion)

N = 37

Name (class & type): NDRI (Bupropion)

Dose (mg/day)/length: 150 mg/day for first week, increased to 300 mg daily for 7 weeks

Delivered by: A psychiatrist who meet with the participant on a weekly basis for 10 minutes to monitor progress with regard to online gaming

Combination (Bupropion+CBT)

N = 35

Name (description): CBT includes focus on negative consequences of online gaming and irrational beliefs about on‐line gamiing, problem solving and decision making, communication skills trianing, self control training, family therapy, planning for the future

# sessions/length: 8 one‐hour sessions delivered weekly, no details provided regarding actual attendance rates

Manualised (Y/N): No

Individual or group: Group

Parent involvement: One of the 8 CBT sessions was described as family therapy

Fidelity check: No

Delivered by: a multidisciplinary treatment team including a psychiatrist, nurse, psychologist, and social worker

Name (class & type): NDRI (Bupropion)

Dose (mg/day)/length: 150 mg/day for first week, increased to 300 mg daily for 7 weeks

Delivered by: A psychiatrist who meet with the participant on a weekly basis for 10 minutes to monitor progress with regard to online gaming

Outcomes

Self reported

Beck Depression Inventory (Beck 1961)

Additional Measures

Young Internet Addiction Scale (YIAS) (Young 1996)

Beck Anxiety Inventory (BAI) (Beck 1988).

Modified‐School Problematic Behavior Scale (Baker 1984).

Modified Student’s Life Satisfaction Scale (Huebner 1991)

Notes

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

No description

Allocation concealment (selection bias)

Unclear risk

No description

Blinding (performance bias and detection bias)
Assessors

Unclear risk

No assessor rated outcomes

Blinding (performance bias and detection bias)
Participants

High risk

Participants not blinded (no placebo CBT condition)

Incomplete outcome data (attrition bias)

High risk

10% missing data; observed case data used in analysis

Selective reporting (reporting bias)

Unclear risk

Authors report on all outcomes specified in their methods section. No access to trial protocol.

Other bias

Unclear risk

No rating of fidelity; very short follow‐up period

Methods

Duration: 6 weeks

Follow‐up assessment points: Post‐intervention

Funded by: Not specified

Participants

N = 33

Child and adolescent (8 to 17 years)

Depression diagnoses included: DSM‐IV Major Depression

Baseline risk of suicide: Participants who were acutely suicidal were excluded form the study. No other specific suicide measurements were administered

Baseline severity of depression: Not reported

Comorbidity included: Not reported

Age mean (SD): Total based on completed participants: 12.7 (2.88)

Sex (M:F): 25:8

Setting: Outpatient

Excluded psychiatric diagnoses: Schizophrenia, mental retardation and Gilles de la Tourette’s syndrome

Country: USA

Interventions

Combination (SNRI + Psychotherapy)

N = 20

Name: Predominantly behavioural/cognitive in nature

 # sessions/length: One weekly session over 6 weeks

Manualised (Y/N): Not reported

Individual or group: Individual sessions of 60 minutes (45 minutes plus 15 minutes "collateral")

Parent involvement: 15 minutes at the end of each session was "collateral" with parents and participants

Fidelity check: Not reported

Delivered by: Masters level therapists, trained in the procedural aspects of the study

Name (class & type): SNRI (Venlafaxine)

Dose (mg/day)/length: Children (8 to 12 yrs) began at 12.5 mg q.d for 3 days, increasing to 12.5 mg b.i.d for 3 days, and further increased to 12.5 mg t.i.d for the remainder of the study. Adolescents (13‐17yrs) began at 25mg q.d. for 3 days, increased to 25mg b.i.d for 3 days and then 25mg t.i.d. for the remainder of the study

Delivered how: Weekly clinic supplied medication/placebo and monitored vital signs and side effects

Combination (Placebo+Psychotherapy)

N = 20

Details as above (Y/N): Yes

Outcomes

Clinician reported

The Child Depression Rating Scale (CDRS; Poznanski 1979)

Self reported

Children’s Depression Inventory (CDI; Kovacs 1992)

Parent reported

The Child Behaviour Checklist (CBCL; Achenbach 1993)

Additional Measures

The Hamilton Rating Scale for Depression (Hamilton 1960)

Notes

Age and gender calculated manually from Figure 1 (pg 151 of the Mandoki 1997 publication)

Dropouts during treatment to any or at least 1 adverse reaction: One participant developed a manic episode, was hospitalised and subsequently put on lithium. Authors note in discussion that "There are specific side effects associated with venlafaxine treatment....However, these side effects were not severe enough to discontinue the medication"

Suicide‐related outcome as an adverse event of treatment: None reported

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

No information contained in paper to make a judgement

Allocation concealment (selection bias)

Unclear risk

As authors quote study as being 'double blind'

Blinding (performance bias and detection bias)
Assessors

Low risk

"After the 6‐week treatment, the study ended. The blind was broken...”. pg. 151 Under heading Procedures, Measurements, and Medication Dose

Blinding (performance bias and detection bias)
Participants

Low risk

“The patients were randomly assigned, in a double blind fashion, to either the venlafaxine and psychotherapy or the placebo and psychotherapy treatment group”. pg. 151 Under heading Procedures, Measurements, and Medication Dose

Incomplete outcome data (attrition bias)

High risk

ITT analysis: No. Figure 1 (pg 151 of the Mandoki 1997 publication) shows the age and sex composition of the final sample on which the statistical analysis was based”. Manual calculation shows the analysis was conducted only on participants completing the trial. pg. 150 Under heading Subjects

Number randomised:

SNRI + Psychotherapy: 20                Placebo + Psychotherapy: 20            Total: 40

Number of dropouts during intervention:

SNRI + Psychotherapy: 4                   Placebo + Psychotherapy:  3             Total: 7

Number analysed post‐intervention:

SNRI + Psychotherapy: 16                Placebo + Psychotherapy: 17             Total: 33

Reasons for dropouts: 6 did not continue coming to the clinic by week 2 for unknown reasons, and 1 patient (in the venlafaxine group) developed a manic episode and was hospitalised

Selective reporting (reporting bias)

High risk

No numerical outcome data was reported in the article, all data was presented in graphs only. Do not have access to trial protocol

Other bias

Low risk

Methods

Duration: 12 weeks

Follow‐up assessment points: Post‐intervention, 6 months.

Funded by: Beyond Blue, Premiers Youth Suicide Taskforce, Department of Human Services Victoria and Australian Rotary Health Research Fund.

Participants

N = 73

Adolescents only (12 to 18 years)

Depression diagnoses included: DSM‐IV major depressive disorder (MDD), dysthymic disorder (DD) and depressive disorder not otherwise specified (DDNOS)

Baseline risk of suicide: Participants who were 'actively suicidal' were excluded from the study, however 'suicidally depressed teenagers (who did not require hospitalisation) were included. Measured using the Suicidal Ideation Questionnaire‐Junior High School Version (SIQ‐JR; Reynolds 1987)

CBT = 26.05 (19.93)

Sertraline = 29.42 (27.24)

Sertraline + CBT = 30.64 (24.42)

Participants exhibiting active suicidality that required acute hospital admission were excluded from the study

Baseline severity of depression:

CBT = 83.77 (13.8)

Sertraline = 84.92 (11.20)

Sertraline + CBT = 83.96 (15.01)

Comorbidity included: 69% were diagnosed with at least 1 comorbid disorder, 22% were diagnosed with 2 or more

Comorbid disorder (n)                               CBT             Sertraline           Sertraline + CBT

Anxiety disorders                                             8                           9                             10

Dysthymic disorder                                         1                           2                              3

Conduct Disorder/ODD                                 2                           3                              1

Body dysmorphic disorder                           1                           0                              0

Adjustment disorder with anxiety            0                           1                              0

Enuresis                                                            1                           0                              0

Reading Disorder                                              0                           1                              0

Cannabis‐related disorder NOS                 0                           1                              0

Parent‐child relational problem                 5                            6                              8

Sibling relational problem                              1                             2                              3

Age mean (SD): 15.3 (1.5)    

CBT= 15.0                   

Sertraline = 15.5         

CBT + Sertraline = 15.3             

Sex (M:F): 25:48        

CBT = 7:15

Sertraline = 7:19

CBT + Sertraline = 11:14

Setting: 3 clinics collocated with public child and adolescent mental health services

Excluded psychiatric diagnoses: Bipolar disorder, psychotic disorder, primary diagnosis of substance abuse disorder, severe psychiatric disturbance that required acute hospital admission, and intellectual disability of sufficient severity to preclude participation in the study

Country: Australia

Interventions

Psychotherapy (CBT)

N = 22

Name: CBT course based in the Adolescent Coping with Depression Course (Clarke 1990). Modules included; goal setting, psycho education, affective education, self monitoring, relaxation training, social skills training, pleasant events scheduling, cognitive therapy and life goals planning

 # sessions/length: Twelve 50 minute sessions over 12 weeks. Three ‘booster’ sessions were also delivered over 3 months

Manualised (Y/N): Yes

Individual or group: Individual

Parent involvement: Parents who chose to participate received concurrent CBT sessions, with 2 family sessions

Fidelity check: No formal check. Clinicians received weekly to twice weekly supervision with an expert therapist. Peer supervision held weekly

Delivered by: 7 registered psychologists, a supervised probationary psychologist, 2 general medical practitioners, and a social worker with experience in providing CBT for adolescent depression. Training provided by chief investigators

Medication (Sertraline)

N = 26

Name (class & type): SSRI (Sertraline)

Dose (mg/day)/length: 25 mg/day for 1 week, increased to 50 mg/day at week 2 depending on response and adverse events. Maximum dose of 100 mg/day administered depending on clinical response and tolerability

Delivered how: Review sessions occurred every 2 to 3 weeks to monitor adverse effects, and included  education about depression but no CBT strategies

Combination (Sertraline + CBT)

N = 25

Details as above (Y/N): Yes

Outcomes

Clinician reported

The Schedule for Affective Disorders and Schizophrenia for School‐Age Children‐Lifetime Version (KSADS‐PL; Kaufman 1997) was used to assess for disorder or remission, which was based on DSM‐IV criteria for full remission (i.e. 8 weeks asymptomatic).

The Global Assessment of Functioning Scale (GAF; APA 1994)

Dropouts:

Post‐intervention:

CBT: 21/22 completed (1 dropout)

Sertraline: 21/26 completed treatment (5 dropouts)

Sertraline + CBT: 20/25 completed treatment (5 dropouts)

6 month follow‐up:

CBT: 19/22 completed assessment (3 dropouts)

Sertraline: 23/26 completed assessment (3 dropouts)

Sertraline + CBT: 24/25 completed assessment (1 dropout)

Self reported

Reynolds Adolescent Depression Scale (RADS; Reynolds 1986)

The Suicidal Ideation Questionnaire‐Junior High School Version (SIQ‐JR; Reynolds 1987)

Parent reported

The Child Behaviour Checklist (CBCL; Achenbach 1991)

Additional Measures

The Global Assessment of Relational Functioning Scale (APA 1994)

Revised Children’s Manifest Anxiety Scale (RCMAS; Reynolds 1978)

The Self Efficacy Questionnaire for Depressed Adolescents (SEQ‐DA; Tonge 2005)

Family Assessment Device General Functioning Scale (Epstein 1983)

Notes

Dropouts during treatment to any or at least 1 adverse reaction: 6% discontinued medication due to adverse affects. These effects included slurred speech and dizziness, feeling agitated and restless, and diarrhoea

Suicide‐related outcome as an adverse event of treatment: 11.1% (n = 45) of participants taking sertraline either alone or with CBT reported suicidal ideation. 1 participant in the sertraline + CBT received an inpatient admission for several hours, however treatment according to protocol was subsequently continued

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

“...subjects were randomly allocated by an independent statistician using a computer generated assignment to CBT, MED or COMB”. pg. 1154

Allocation concealment (selection bias)

Low risk

”...allocated by an independent statistician...Allocation for those eligible for the trial was concealed to all until after pre‐treatment assessment”. pg. 1154

Blinding (performance bias and detection bias)
Assessors

High risk

“Independent raters blind to treatment allocation were not used because of resource limitations but may have reduced the risk of experimenter bias in assessments”. pg. 1160

Blinding (performance bias and detection bias)
Participants

High risk

Psychotherapy administered in both groups and no placebo control used for medication

Incomplete outcome data (attrition bias)

Unclear risk

ITT analysis: “data were analysed using an intent‐to‐treat strategy to counter any possible overestimation of treatment outcomes, using the last observation carried forward method (Nelson 1996)”. pg. 1155

Number randomised:

CBT: 22      Sertraline: 26               CBT + Sertraline: 25        Total: 73

Number of dropouts during intervention

CBT: 1          Sertraline: 5               CBT + Sertraline: 5          Total: 11

Number dropouts in follow‐up:

CBT: 3          Sertraline: 3               CBT + Sertraline: 1          Total:7

Number analysed post‐intervention:

CBT: 22          Sertraline: 26           CBT + Sertraline: 25    Total: 73

Number analysed follow‐up 1:

CBT: 22          Sertraline: 26           CBT + Sertraline: 25     Total: 73

Reasons for dropouts:

CBT: At post‐intervention, 1 participant reported symptoms had improved. At 6‐month follow‐up, 2 refused to attend and 1 was unable to be located

Sertraline: At post‐intervention, 2 participants reported symptoms had improved, 1 reported side effects, 1 dissatisfied with programme and 1 did not pursue treatment. At 6‐month follow‐up, 1 participant refused to attend, 1 was unable to be located and 1 'trial closure'

CBT + sertraline: At post‐intervention: 2 reported side effects, 1 symptoms improved, 1 dissatisfied with programme, and 1 did not respond. At 6‐month follow‐up, 1 refused to attend

Selective reporting (reporting bias)

Unclear risk

Remission data not reported by group and functioning data not reported in a useable format. All other outcomes were reported. Do not have access to trial protocol

Other bias

Low risk

Methods

Duration:16 weeks

Follow‐up assessment points: Post‐intervention

Funded by: National Institute on Drug Abuse, National Institutes of Health

Participants

N = 126

Adolescent only (13 to 19 years)

Depression diagnoses included: DSM‐IV current MDD episode

Baseline risk of suicide:Primary measure of suicidality was question 13 on the Childhood Depression Rating Scale‐Revised (CDRS‐R; Poznanski 1996)

Baseline suicidality data: Fluoxetine + CBT = 25/63 (39.7%) CBT + placebo = 24/63 (38.1%). N = 13 displayed severe suicidal ideation (> 5 on CDRS‐R Q13)

“Adolescents with past, current or intermittent suicidal ideation (39% at baseline) were not excluded from study participation unless suicidal ideation  were severe or they were otherwise considered by the study physician and according to baseline CDRS‐R ratings (question 13) to be at high risk for a suicide attempt during the trial”

Baseline severity of depression: Children’s Depression Rating Scale‐Revised (CDRS‐R; Poznanski 1996) mean t‐score (SD):

Fluoxetine + CBT: 73.74 (8.51)

Placebo + CBT: 73.03 (7.70)

Comorbidity included: All participants had at least 1 non‐tobacco Substance Use Disorder (SUD), and lifetime Conduct Disorder (CD)

Age mean (SD): Total = 17.16 (1.66)                         

Sex (M:F):  Total = 85:41         

Setting: Outpatient

Excluded psychiatric disorders: Current or past diagnosis of a psychotic disorder or of bipolar disorder (type I or II)

Country: USA

Interventions

Combination (Fluoxetine + CBT)

N = 63

Name: CBT approach using behavioural, cognitive behavioural and motivational enhancement techniques to help adolescents reduce their drug use. The programme contains 1 session specifically on depression, helping adolescents to identify, manage and regulate mood states that often trigger substance use.

 # sessions/length: 1 hour, 16 weekly sessions

Manualised (Y/N): Yes

Individual or group: Individual

Parent involvement: Not specifically but could include up to 2 parent sessions

Fidelity check: Yes. All sessions videotaped and self rated by therapists. 32 videotapes randomly selected and independently rated for adherence and fidelity. “...neither therapist fell below present fidelity/adherence standards during any point of the study”

Delivered by: Study therapists (MD) who were trained and certified by one of the manuals developers. The developer provided ongoing supervision and quality monitoring

Name (class and type): SSRI (Fluoxetine)

Dose (mg/day)/length: 20 mg fixed daily dose

Delivered how: Monitoring of adverse effects and medication adherence was undertaken by research nurses, and occurred either immediately before or after the weekly CBT session

Combination (Placebo + CBT)

Details as above (Y/N): Yes

Outcomes

Clinician reported

Remission of depression defined as as post‐intervention Children’s Depression Rating Scale‐Revised (CDRS‐R; Poznanski 1996) score of ≤ 28

Children’s Depression Rating Scale‐Revised (CDRS‐R; Poznanski 1996)

The Clinical Global Impression Improvement rating (CGI‐I; Guy 1976)

Self reported

Question 13 on the Children’s Depression Rating Scale (CDRS‐R; Poznanski 1996)

Additional Measures

Self reported number of non‐tobacco drugs used in the past 30 days

Urine samples for substance use

Conduct Disorder: Number of self reported DSM‐IV symptoms in the past 30 days

Notes

Group means for age and gender not reported

Dropouts during treatment to any or at least 1 adverse reaction: Authors list 6 as ‘lost to follow‐up’ and 2 to ‘withdrew consent’ but do not disclose if this was due to an adverse reaction

Suicide‐related outcome as an adverse event of treatment: 5 participants (4 in the fluoxetine + CBT group and 1 in the Placebo + CBT group were evaluated in an emergency department or hospitalised for concerns of worsening suicidality during the study.

Standard error and sample size was used to calculate standard deviations for group means

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

High risk

“A non‐blinded research pharmacist assigned eligible participants to receive 20 mg of fluoxetine hydrochloride or matching placebo using a small block (6) randomisation scheme of 20 blocks to achieve balance in the treatment assignment”. pg. 1027

Allocation concealment (selection bias)

Low risk

“Active medication and matching placebo were prepared by the research pharmacy at the University of Colorado at Denver and Health Sciences Centre and then provided to clinical research staff in pre‐randomized and pre‐blinded medication bottles”. pg. 1027

Blinding (performance bias and detection bias)
Assessors

Low risk

“Research staff....remained blinded to medication status throughout the trial”. pg.1027

Blinding (performance bias and detection bias)
Participants

Low risk

: “.....participants remained blinded to medication status throughout the trial”. pg.1027

Incomplete outcome data (attrition bias)

Low risk

ITT analysis: “All analyses were intent‐to‐treat (including all randomized study participants)”. pg. 1028.

Impulation method used: “Analyses of dichotomous and continuous primary outcome measures over time used generalized estimating equation (GEE) and likelihood based methods, respectively. Both allow for estimates of changes in repeated measures in the presence of missing data, assuming those data were missing at random”. pg. 1028

Number randomised:

Combination (Fluoxetine + CBT): 63  Combination (Placebo + CBT): 63       Total: 126

Number of dropouts during intervention:

Combination (Fluoxetine+CBT): 11  Combination (Placebo + CBT):9          Total:20

Number analysed post‐intervention:

Combination (Fluoxetine + CBT): 63    Combination (Placebo + CBT): 63       Total:126

Reasons for dropouts: Fluoxetine + CBT: 4 participants went to jail/detention, 3 went to residential treatment at facility and were unable to continue the study, 3 were lost to follow‐up and 1 moved out of area.

Placebo + CBT: 1 participant went to jail/detention, 3 were lost to follow‐up, 3 moved out of area and 2 withdrew consent

Selective reporting (reporting bias)

Unclear risk

No group data on suicide outcomes reported. All other outcomes specified in methods reported. Do not have access to trial protocol

Other bias

Low risk

Methods

Duration: 12 weeks acute treatment, 6 weeks continuation treatment and 18 weeks maintenance treatment

Follow‐up assessment points: Post‐intervention (12 weeks), 18 weeks (after continuation), 36 weeks (after maintenance)

Funded by: National Institution of Mental Health to Duke University Medical Centre

Participants

N = 439

Adolescent only (12 to 17 years)

Depression diagnoses included: DSM‐IV Major Depressive Disorder (MDD) and a score of 45 or more on the CDRS‐R (Poznanski 1996)

Baseline risk of suicide:

*data obtained from Table 2, 2004 paper

Measured using the Suicidal ideation Questionnaire‐Junior High School Version (SIQ‐JR; Reynolds 1987)

Adjusted mean (SD):

CBT: 21.91 (16.28)

Fluoxetine: 21.81 (15.68)

Fluoxetine + CBT: 27.33 (18.51)

Placebo: 24.20 (16.46)

Analysed according to a cut‐off score of ≤ 31

CBT: 27/107 (25.2%)

Fluoxetine: 28/107 (26.2%)

Fluoxetine + CBT: 42/106 (39.6%)

Participants excluded if deemed ‘high risk’ because of a suicide attempt requiring medical attention within 6 months. Also excluded on the basis of having a clear intent or active plan to attempt suicide, or suicidal ideation accompanied by a disorganised family unable to guarantee adequate safety monitoring

 Baseline severity of depression:

*data obtained from Table 1, 2004 paper, t‐scores presented

Children’s Depression Rating Score (CDRS‐R; Poznanski 1996):

CBT: 75.37 (6.32)

Fluoxetine: 74.73 (6.74)

Fluoxetine + CBT: 75.67 (6.53)

Placebo: 76.14 (6.11)

Comorbidity included:

Comorbidity (%)                              CBT        Fluoxetine          Fluoxetine+CBT       Placebo

Any psychiatric comorbidity         58.18                  43.12                     55.66                     51.35

Dysthymia                                         15.45                    5.5                       10.28                     10.71

Anxiety                                               32.43                  23.85                     28.4                       25.23

Disruptive behaviour                       24.32                  22.94                    21.50                     25.00

Obsessive compulsive/tic              1.80                     1.83                      3.74                       3.57

Substance use                                    0.90                     2.75                     2.80                        0

Attention‐deficit/hyperactivity  12.61                   11.93                   13.08                     16.96

Taking medications                          3.60                     2.75                      3.74                      8.93

Age mean (SD): Total = 14.6 (1.54)     

CBT = 14.62 (1.50)                 

Fluoxetine = 14.50 (1.57)                    

CBT + Fluoxetine = 14.6 (1.48)

Placebo = 14.51 (1.62)            

Sex (M:F): 200:239                 

CBT= 50:61                

Fluoxetine = 50:59                  

CBT + Fluoxetine =  47:60

Placebo = 53:59                                  

Setting: Outpatient

Excluded psychiatric disorders: Current or past diagnosis of bipolar disorder, severe conduct disorder, current substance abuse or dependence, pervasive developmental disorder(s), thought disorder or psychiatric disorders requiring out of protocol treatments

Country: USA

Interventions

Psychotherapy (CBT)

N = 111

Name: CBT modules included psycho education about depression and it’s causes, goal‐setting, mood monitoring, increasing pleasant activities, social problem‐solving, and cognitive restructuring

 # sessions/length: Fifteen 1 hour sessions during stage 1, 6 additional sessions for partial responders and bi‐weekly sessions for full responders in stage 2, and 3 sessions (1 every 6 weeks) in stage 3

Manualised (Y/N): Yes

Individual or group: Individual

Parent involvement: 1 to 3 conjoint parent and adolescent sessions took place

Fidelity check: Not reported

Delivered by: Not reported

Medication (Fluoxetine)

N = 109

Name (class and type): Fluoxetine (SSRI)

Dose (mg/day)/length: 10 mg/day and increased up to 40 mg/day by week 8. At week 12, dose raised to 50 to 60mg/day for ‘partial responders’ and ‘full responders’ remained on same fluoxetine dose

Delivered how: Monitoring and status and medication effects occurred during 20 to 30 minute visits. Clinician also offered general encouragement about the effectiveness of pharmacotherapy for MDD

Combination (Fluoxetine+CBT)

N = 107

Details as above (Y/N): Yes

Placebo

N = 112

Outcomes

Clinician reported

Schedule for Affective Disorder and Schizophrenia for School‐Age Children‐Present and Lifetime Version (K‐SADS‐P‐L; Kaufman 1997)

This was used to define remission i.e. those who had did not have a continuing or new mood disorder since the last interview according to the K‐SADS‐PL).It was unclear if DSM‐IV or ICD time criteria were employed

Children’s Depression Rating Scale‐Revised (CDRS‐R; Poznanski 1996)

Self reported

Reynolds Adolescent Depression Scale (RADS; Reynolds 1986).

Suicidal ideation Questionnaire‐Junior High School Version (SIQ‐JR; Reynolds 1987)

Additional Measures

Clinical Global Impression Improvement (CGI‐I; Guy 1976)

Child and Adolescent Impact Assessment (Angold 1998)

Columbia University classification scheme of the US Food and Drug Administration analyses of antidepressant‐associated suicidal events

Notes

Dropouts during treatment to any or at least 1 adverse reaction:

Suicide‐related outcome as an adverse event of treatment: 24 (5.5%) of participants experienced a suicide‐related adverse event Total number (%): *data obtained from 2006 paper

CBT: 5 (4.5)

Fluoxetine: 10 (9.2)

CBT + Fluoxetine: 5 (4.7)

Placebo: 3 (2.7)

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

“Eligible participants were randomly assigned...using a computerized stratified randomisation, a 1:1:1:1 treatment allocation ratio, permuted blocking (first block size = 4, with subsequent random block sizes of 4 and 8) within each striatum, and site and sex stratification variables”. pg. 808

Allocation concealment (selection bias)

Low risk

“Participants were randomly assigned...at the coordinating centre”. pg. 449

Blinding (performance bias and detection bias)
Assessors

Low risk

“TADS used 2 primary measures of depression status assessed...by an independent evaluator blind to condition”. pg. 448

Blinding (performance bias and detection bias)
Participants

High risk

“Participants and all study staff remained masked in the ‘pills only’ condition (fluoxetine therapy and placebo) until the end of stage 1 (week 12). Patients and treatment providers in the combination and CBT conditions were aware of treatment assignment”. pg. 1133

Incomplete outcome data (attrition bias)

Unclear risk

“The primary analyses of remission rates...were conducted using an “intention to treat” (ITT) approach in which the analysis included all participants randomized to treatment regardless of protocol adherence and/or treatment completion”. pg. Under heading Data Analysis, 2009

Imputation method: LOCF

Number randomised:

CBT: 111      Fluoxetine: 109      Fluoxetine + CBT: 107     Placebo:  112     Total: 439

Number of dropouts during intervention

CBT: 41       Fluoxetine: 38        Fluoxetine + CBT: 23       Placebo: 14         Total: 116

Number of dropouts in follow‐up (18 weeks):

CBT: 21        Fluoxetine: 37      Fluoxetine + CBT: 15      Placebo: 8           Total: 81

Number of dropouts in follow‐up (36 weeks):

CBT: 25        Fluoxetine: 21         Fluoxetine+CBT:23    Placebo: 15        Total: 84

Number analysed post‐intervention:

CBT: 111     Fluoxetine: 109      Fluoxetine + CBT:107     Placebo: 112     Total: 439

Number analysed follow‐up 1 (18 weeks):

CBT: 111      Fluoxetine: 109      Fluoxetine + CBT:107    Total: 327

Number analysed follow‐up 2 (36 weeks):

CBT: 111      Fluoxetine: 109      Fluoxetine + CBT: 107     Total: 327

For active treatment arms: 84/327 exited the study because of loss of follow‐up or withdrawal of consent (n = 21 for CBT + fluoxetine, n = 32 for fluoxetine, n = 31 for CBT). 96/327 discontinued treatment before week 36 due to premature termination or non‐response at the end of stage 1 (n = 25 for CBT + fluoxetine, n = 39 for fluoxetine, n = 32 for CBT), and this discontinuation was decided by the study physician

For placebo: 13/112 participants were terminated prematurely from the study by week 12 due to clinical worsening

Selective reporting (reporting bias)

Unclear risk

Trial protocol located

Other bias

High risk

Combination therapy group had an excess of suicidal ideation at baseline relative to fluoxetine or CBT

Methods

Duration: 6 months

Follow‐up assessment points: Post‐intervention (24 weeks)

Funded by: National Institute of Mental Health (NIMH)

Participants

N = 124

Adolescent only (12 to 18 years)

Depression diagnoses included: DSM‐IV Major Depressive Disorder (MDD), Dysthymic Disorder (DD) or Depressive Disorder not otherwise specified (DD‐NOS). Participants also had to obtain a score of 36 or more on the CDRS‐R (Poznanski 1996)

Baseline risk of suicide: Participants were only eligible for participation if they had made a suicide attempt in the last 90 days. Beck Scale for Suicidal Ideation (SSI; Beck et al 1979). Mean (SD):

Total = 6.3 (7.7)

CBT‐SP = 5.0 (6.0)

SSRI = 3.9 (6.0)

SSRI+CBT‐SP = 6.9 (8.2)

Baseline severity of depression: 96% met criteria for MDD and 10.5% had DD and DD. Children’s Depression Rating Scale‐Revised (CDRS‐R; Poznanski 1996). Mean (SD):

Total =  50.4 (12.6)

CBT‐SP =  46.9 (14.7)

SSRI = 43.4 (11.1)

SSRI+CBT‐SP= 52.1 (12.0)

Comorbidity included:

Comorbidity (%)               CBT‐SP  SSRI       SSRI+CBT‐SP

Anxiety                                   23.5        28.6        63.4

ADHD                                    11.8        14.3        23.7

ODD/CD                                0.0          35.7        15.1

Age mean (SD):

Total =  15.7 (1.5)

CBT‐SP = 15.7 (1.5)

SSRI = 15.6 (1.4)

SSRI + CBT‐SP = 15.7 (1.6)

Sex (M:F): 28:96            

CBT‐SP = 1:16

SSRI = 1:13

SSRI + CBT‐SP = 26:67   

Setting: Academic sites

Excluded psychiatric diagnoses: Substance dependence, bipolar disorder, psychosis and pervasive developmental disorders (PDD)

Country: USA

Interventions

Psychotherapy (CBT‐SP)

N = 17

Name: CBT + Suicide prevention (CBT‐SP). Modules included; chain analysis of the suicide attempt, safety planning, formulation of the participants cognitive, behavioural, affective and contextual problems, behavioural activation, cognitive restructuring, problem‐solving, and relapse prevention

# sessions/length: Up to 22 sessions, length not specified

Manualised (Y/N): Yes

Individual or group: Individual

Parent involvement: Parent‐youth sessions were included

Fidelity check: No formal check. Weekly telephone conferences were held to review cases

Delivered by: Trained psychotherapists under the supervision of senior experts

Medication (SSRI)

N = 14

Name (class and type): SSRI. Step 1: Monotherapy with an SSRI. Step 2: In the case of non‐response changed to a different SSRI. Stage 3: Medication changed to an alternative class (venlafaxine, duloxetine, mirtazapine, or bupropion) with option of augmenting with lithium or another SSRI

Dose (mg/day)/length: Not specified

Delivered how: By psychopharmacologists

Combination (SSRI + CBT‐SP)

N = 93

Details as above (Y/N): Yes

Outcomes

Clinician reported

Children’s Depression Rating Scale‐Revised (CDRS‐R; Poznanski 1996)

Children’s Global Assessment Scale (C‐GAS; Shaffer 1983)

Self reported

Beck Depression Inventory (BDI; Beck 1988)

Beck Scale for Suicidal Ideation (SSI; Beck 1979a)

Recurrence of a suicidal event

Additional Measures

The Montgomery‐Asberg Depression Rating Scale (MADRS; Montgomery 1979)

The Multidimensional Anxiety Scale for Children (MASC; March 1997)

The Clinical Global Impressions‐Severity (CGI‐S) and Improvement scales (CGI‐I; Guy 1976)

Notes

Dropouts during treatment to any or at least 1 adverse reaction:

Suicide‐related outcome as an adverse event of treatment: 19.5% of participants experienced a suicide event and 12% made a suicide attempt. 1 participant died of suicide 20 days after completing the 24 week SSRI + CBT ‐ SP treatment

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

High risk

“The study started as a randomized controlled trial; however; after approximately 9 months of low enrolment despite intensive recruitment efforts, the design was changed so that patients and their families could accept randomisation or choose which treatment to receive”. pg. 998

Allocation concealment (selection bias)

High risk

As participants chose their treatment condition, allocation concealment is not applicable

Blinding (performance bias and detection bias)
Assessors

Low risk

“Independent evaluators were trained to ensure interrater reliability and remained blind to patient treatment assignment”. pg. 999

Blinding (performance bias and detection bias)
Participants

High risk

“Most (n = 104) chose their treatment rather than being randomized”. pg. 1000

Incomplete outcome data (attrition bias)

Unclear risk

“The data were analysed with an intent‐to‐treat approach”. pg. 999

Imputation method: LOCF

Number enrolled (included non‐randomised participants):

CBT‐SI: 17                  SSRI: 14              SSRI + CBT‐SP: 93                Total: 124

Number of dropouts during intervention

CBT‐SI: 6                     SSRI: 6                 SSRI + CBT‐SP: 26               Total: 36

Number analysed post‐intervention:

CBT‐SI: 17                  SSRI: 14               SSRI+ CBT‐SP: 93                Total: 124

Reasons for dropout: 2 participants reported suicidal intent with inability to commit to safety plan, 5 showed lack of adherence to treatment, 9 had a need for different treatments and services and 23 withdrew consent or failed to return for visits for unspecified reasons

Selective reporting (reporting bias)

High risk

Only total and combination treatment outcomes reported. Do not have access to trial protocol

Other bias

High risk

The SSRI + CBT ‐ SP group had higher levels of depression severity at baseline compared with the SSRI or CBT alone group. The SSRI + CBT ‐ SP group also had a higher prevalence of comorbid anxiety and more functional impairment than the other 2 groups

Only 20/124 participants were randomised, the remaining 104/124 chose their treatment option