IMPACT
Randomised Controlled Trial of Brief Psychodynamic Psychotherapy (BPP), Cognitive Behaviour Therapy (CBT) and Treatment as Usual (TAU) in adolescents with moderate to severe depression attending routine child and adolescent mental health clinics
Funded by the NHS Health Technology Assessment (HTA) programme
Principal Investigator: Ian Goodyer (Cambridge)
Co-PIs: Peter Fonagy, Jonathan Hill, Shirley Reynolds, Chris Roberts, Sarah Byford, Raphael Kelvin, Mary Target, Rob Senior
Background
Aim of the study
Design
Setting
Target population
Health technologies being assessed
Measures of cost and outcomes
Sample size
Project timetable including recruitment rate
Background
Depression in adolescents is a very serious problem. Over half of those diagnosed retain their problem into young adulthood. Current best evidence based ethical treatment involves psychosocial interventions together with an anti-depressant (fluoxetine). However, after 6 months only 20% show full recovery, 30% improve but are not fully recovered, a further 30% are left with a high number of residual depressive symptoms and 20% do not respond at all.
The reasons for the poor responses are unclear. Some possible explanations are:
- the duration of psychological treatment in routine practice may be shorter than necessary
- the treatment may not be administered to a sufficiently high standard
- Patients may not have been followed along for long enough
Thus many patients remain at considerable risk of relapse in the weeks and months following the cessation of treatment.
Given that depression in adulthood is one of the three most important health burdens on UK society and emerging evidence suggesting that adolescent brain functioning may have formative influence for later development, finding ways to decrease the risk adolescent recurrent depressions through adequate treatment of early episodes would be a major public health advance.
Aim of the study
This investigation seeks to establish if increasing the amount and quality of psychological treatments offered in one or both of two ways will increase treatment response, reduce the level of residual symptoms and decrease the proportion of patients at risk for depressive relapse. To achieve our objectives we will compare 3 different treatments as delivered in adolescent mental health services across the UK and over a time period of 12 months , twice as long as any previous study.
One of the treatments will be ‘treatment as usual’ ( TAU generally addressing the life situation of the adolescent and the family plus fluoxetine). The other two will also receive TAU but will also have either a brief psychodynamic psychotherapy or cognitive behaviour therapy. These are the treatments with strongest evidence base and most commonly offered routinely in the NHS but often inadequately or for too short a period.
We do not expect that the specialist therapies on their own would effectively treat depression but we predict that, in conjunction with TAU, one or both the therapies will generate better outcomes than TAU alone over a 12 month period.
To our knowledge this will be the first RCT comparing two manualised psychological treatments with TAU in major adolescent depression. This will allow us to determine if there are treatment specific and non-specific effects and yield preliminary data for predictors of different responses to particular treatments.
Design
This randomized controlled trial will compare two psychological treatments with treatment as usual (TAU) over a 52 week period. Placebo conditions longer then 12 weeks for potentially suicidal adolescents are neither feasible nor ethical. After diagnosis and collection of baseline characteristics patients will be offered TAU and reassessed 2 weeks later and if eligible then randomly assigned to BPP, CBT or continuing TAU. The primary endpoint will be imapirment at 28 weeks on the HoNOSCA. Secondary endpoints will be clinical diagnosis and depressive symptoms with follow-up assessments at 32 and 52 weeks to obtain residual symptoms and relapse rates. To assess the course of treatment response, repeated measures of symptoms and function will be taken at 6 timepoints (weeks 0, 6, 12, 28, 36 and 52). Patient recruitment and treatment will take 36 months, treatments will be offered over a period of 28 weeks for individual patients.
Setting
The study will be based in 14 Tier 3 child and adolescent mental health clinics to maximize generalizability. Participating clinicians will be trained to criteria of adherence. All other services will be offered as described in the recently published NICE guidelines.
Target population
Selection criteria have been set to be representative of the population: Criteria for inclusion (i) age 11-17, (ii) at least 4 DSM IV depressive symptoms, (iii) HoNOSCA score greater than 7, (iv) failure to remit on either of latter two criteria in 2-4 sessions of TAU. Criteria for exclusion (i) history/current diagnosis of psychosis (schizophrenia and bipolar disorder), substance dependence (not abuse) (ii) English too poor to benefit from treatment, (iii) evidence of severe disabling neurological disorder, (iv) IQ below 75, (v) characteristics interfering with treatment completion eg impending custody, life-threatening illnessv vi) prepubertal depressed children who are rare and have different causes and treatments.
Health technologies being assessed
The specialist psychological treatments reflect real-life best practice in NHS CAMHS. This is a pragmatic trial and treatment manuals will act as guides to treatment that could be easily incorporated into NHS practice. The manuals for both treatments have already been used as part of RCTs, The CBT programme is a 28-week programme targeting specific domains of functioning, aimed at reducing dysfunctional behaviour and cognitions associated with depression. The BPP programme is a 30-week programme offering weekly 50-minute sessions aimed at exploring the young person’s conscious and unconscious feelings and thoughts about their relationships. All participants will additionally receive TAU (regular monitoring of mental states, support and encouragement, liaison with parents and professionals and fluoxetine as required).
Measures of cost and outcomes
Outcome measures have been chosen with regard to comparability with major studies in US and UK and to incorporate measures suggested in user and carer consultations in the first phase of the project. As well as symptom oriented measures assessments will include measures of personality functioning (ADAPFA) and treatment modality specific moderator variables assessed at baseline (ruminative thinking style, attachment status) as well as factors hypothesized to moderate treatment response more generally (emergent Axis-Il disorders, homozygosity for the s/s allele of the serotonin promoter gene, cortisol hypersecretion). A key secondary measure of outcome will be costing support for young peolpe using the Child and Adolescent Service Use Schedule. This yields information concerning health, social, education,voluntary, private and family costs including travel, chiuldcare treatment and productivity losses pre-trial, during treatment and follow-up. Total costs for each arm will be computed using locally applicable estimates of long-run marginal opportunity costs supplemented by national unit costs. Cost-effectiveness analysis will be explored for both lry and 2ry outcomes and QALYS.
Sample size
The expected difference in proportion of our trial is derived from the ADAPT triall where 20% recover, 30% improve, 30% residual symptoms, 20%no resposne. Assuming 50% of patients in the therapy arms improved by 1 category by 52 week power calculaitons estimate between 420 and 510 patients (140-170 in each arm) are required. Lower figure bonferroni adjusted but no account of clustering due to therapist which was non-sig. in ADAPT. Higher figure assumes anintra-therapist correlationof 0.0 15 from other published studie.s All patients will receive a brief initial intervention involving psychoeducation, problem-solving for acute conflicts, and parent support. Only patients non-responsive to this in two weeks will be randomized to prevent inclusion of patients who are close to remission. On the basis of the ADAPT trial we anticipate that 612 patients will need to be recruited to achieve 510 entering the trial.
Project timetable including recruitment rate
The first 6 months of the project will include consultation with users, training of therapists, and beginning recruitment. 6-36 months will be recruitment, randomization, treatment and follow-up. 36-42 months is follow-up and data analysis. 42-48 months is data analysis and reports. Statistics collected as part of the ADAPT study indicate that approximately 20 young people with MDD who meet our criteria present to a CAMHS service per year. This means 18 CAMHS clinics, six from each region, North-West, East Anglia and North London. The lead researchers from each of the three regions have already succeeded in obtaining initial agreement to collaborate from these sites. It is hoped that the trial will be adopted by the relevant MHRN networks to facilitate further recruitment. Measures taken to reduce bias These will include: consecutive participants being randomized within referral setting, a given therapist only providing one form of treatment, therapist will be trained to criteria in either CBT or BPP and adherence to the treatment manual will be checked by audio tape-recording, random adherence checks and monthly supervision from therapy experts. Stochastic minimization will be used to ensure balance on severity (prescribed medication), institution, gender, comorbid behavioural disorder and age. Parental treatment expectancies (questionnaire) before and after randomization will be controlled. Analysis will be intent to treat of all randomized patients. Blinding of patients and therapists is not possible; however, assessments will be carried out by independent evaluators blind to treatment or controlled condition. External quality management of the study will be provided by a panel of international experts and randomization will be provided by the Department of Medical Statistics at the Christie Hospital.
