In this randomized controlled trial with a sample size of 268 participants (after accounting for 20% drop out, M-CT: 107, TAU: 107) we compare an internet-supported self-directed prevention of relapse program as part of a SMS based monitoring versus treatment as usual (TAU). This M-CT program is called Depression Free
. The target population is a group at elevated risk of relapse and recurrence as identified in several guidelines (e.g. NICE) [11
] that consumes a considerable amount of health care and for whom initial benefits of antidepressants (AD) may be wane off in the long run. Relapse rates rise with increasing numbers of previous episodes up to 70% in 5 years [12
]. In our previous study, we observed up to 62% recurrences within 2 years [13
Randomization will be undertaken by an independent researcher and will be stratified by the number of previous depressive episodes and type of care (i.e. care by a general practitioner versus care in a mental health center) will then be used for stratified randomization. Thereafter our researchers receive the participant number and the automatically random generated condition in the trial by email. For a Flow diagram of the assessment methods see Figure .
We monitor the primary outcome (relapse) over a period of 24 months. Assessments by trained assessors who are blind to treatment allocation (and whose blindness is checked within each assessment session) take place directly after the start of the treatment at three months, 12 and 24 months. For the research aims focused on potential working mechanisms of M-CT we added in between self-report assessments, i.e. baseline, 1.5 months and 3 months.
: This M-CT treatment builds on a previously evaluated face to face intervention, i.e. Preventive Cognitive Therapy (PCT) [14
] and has been developed in collaboration Bockting & van Valen [15
] with the Trimbos Institute. The face to face PCT is an adapted type of cognitive therapy for acute depression [16
] and specifically developed to prevent relapse in recurrent depression in remitted patients. It consists of eight sessions. Like in regular CT, each PCT session follows a fixed structure, with agenda setting, review of homework, explanation of rationale of each session, and assignment of homework. A manual describing the structure of the treatment and interventions used is available [14
]. The intervention prevention program targets underlying cognitive vulnerability factors, such as depressogenic assumptions. Unlike CT for acutely depressed patients, PCT is not primarily directed toward modifying negative thoughts. Instead, it starts with the identification of negative thoughts and dysfunctional attitudes, aided by a self-report questionnaire with examples of attitudes and specific techniques such as the downward arrow technique. The focus of treatment is then directed on changing these attitudes using different cognitive techniques such as Socratic questioning and identification of positive attitudes. Moreover, patients are encouraged to practice with alternative attitudes in the final sessions. Remitted patients with a history of recurrences have an inability to retrieve specific memories from the past and this is associated with impaired problem-solving skills [e.g. 17], long-term course of depressive disorders [18
] and difficulties in recovering from depression. Part of the treatment is keeping a diary of positive experiences in order to enhance specific memories of positive experiences, instead of retaining overly general memories. Further specific relapse/recurrence prevention strategies are formulated in the last three sections of the M-CT resulting in a personal prevention plan.
The M-CT intervention is based on the above described face to face PCT. It is offered over the internet in a series of 8 well-structured modules with online therapist contact (with a maximum of 4 telephone sessions) and by SMS. Each module includes assignments with automatically generated feedback. In addition, films can be activated by the participant to explain more about specific topics. Each module can be completed in approximately 20 minutes, not counting the time required to complete additional assignments. If all assignments are completed, E-personal prevention book will be automatically generated that can be of help in case of relapse of lowered mood. Automated checks will be conducted to ascertain that participants have not only completed modules, but also understood them correctly, before they can move on to the next. When participants do not log in on the intervention's website they will receive friendly reminders by SMS or mail to proceed with the intervention.
Depression-related outcomes (changes in mood) will be monitored with help of SMS (or by email if requested). To this end, participants periodically receive SMS messages in which they will be asked to rate their mood. Participants have to answer by sending a message back consisting of a number with which they rate their mood. When depressed mood is present and appears to persist, then the frequency of messages is increased and other depressive symptoms are monitored also using a web-based self-report assessment (IDS-R) [19
]. In case there is indeed an indication for a depressive episode, participants will be interviewed using the SCID [20
] and the Hamilton Rating Scale for Depression (HRSD) [21
]. This allows for the early detection of possible depression onset. In that scenario, participants receive advice and are encouraged to return to the website where they can find 'prescribed' modules. Hopefully, this offers them the opportunity to better cope with lowered moods.
The intervention is designed to be easily accessible, acceptable and as non-intrusive as possible, while at the same time allowing for tele-monitoring of health related outcomes over the time-span of several years. The web-based intervention has been developed by the Trimbos Institute, the University of Groningen, CrossOver Consultancy and their partners.
Treatment As Usual
The intervention will be compared to treatment as usual (TAU). In this context TAU is fairly heterogeneous: it typically consists of antidepressant (AD) maintenance medication in primary and secondary care, counseling or face-to-face PCT in secondary care, but often there is no treatment at all. To compare the intervention with TAU is relevant from a public health perspective: it would help to demonstrate the intervention's added value over and above TAU. We will not intervene with TAU, but monitor TAU using a health service receipt interview, the TIC-P [22
]. We will assess compliance and adherence to AD use, but also the use of the M-CT program.
With 107 in M-CT versus in TAU 107 participants per condition the trial will be powered to detect a difference of 20% in the cumulative incidence rate of relapse/recurrence with a 2 year follow-up in a 2-tailed test at the conventional alpha level of 0.05 and a power of (1-beta) = 0.80, while conservatively assuming that relapse/recurrence will occur in 50% of the cases. Allowing for a drop-out of 20% we need to include 268 participants at baseline.
Referral and recruitment
Patients will be recruited by media (announcements, banners placed in various related websites, media attention in interviews), referral by general practitioners and mental health services. Patients with concurrent chronic somatic illnesses will be recruited by targeted marketing strategies (e.g. banners on website targeted on patients with chronic somatic illnesses, posters in hospitals) and specific recruitment at general practitioners.
We include recovered patients with a history of at least two previous depressive episodes in the past five years. The last episode has to be at least 2 months and no longer than 2 years ago. The last episode has to be at least 2 months and no longer than 2 years ago and a current score of ≤10 on the HRSD [21; in line with other prevention studies, e.g. 5-6]. No restriction with respect to co-morbidity on Axis II and III, i.e. a concurrent chronic somatic illness is defined as risk factor for relapse and recurrence, [APA, 10]. Consenting participants need to be fluent in Dutch and have access to the internet.
Exclusion criteria are: current mania or hypomania or a history of bipolar illness, any psychotic disorder (current and previous), alcohol or drug misuse, predominant anxiety disorder.
Assessment of Eligibility and Baseline Measures
We inform patients about the study before they come into the study in two ways. First, by informing the patient through a therapist or a general practitioner (GP). If a therapist/GP wants to inform the patient himself, the patient then receives the information via the therapist/GP and is given a letter containing all the information. If the patient is interested in participating, then the participant will contact the researcher. Subsequently, the researcher checks that the participant understands all aspects of the trial. If the participant agrees to enter the trial, she completes a copy of the consent form and sends it to the researcher.
The second procedure we use is by directly informing the patient. Participants than initiate that contact with the researcher themselves (mostly informed by media/websites or by their former therapist/GP with a letter, by advertisements or interviews). Subsequently, the researcher informs the participant, the participant receives the information in a letter with all the information in it. If the participant is still interested in participating then the researcher checks that the participant understands all aspects of the trial. If they agree to enter the trial, they complete a copy of the consent form and send it to the researcher. We remind participants that they can withdraw from the trial at any time and that this has no consequences for their treatment as usual.
We ask consenting participants to provide information about their socio-demographic background and assess their eligibility in more detail using semi-structured clinical interviews (SCID-I, by telephone) and self-completed questionnaires (web-based). The researchers assess current and past diagnostic status using the Structured Clinical Interview for DSM IV (SCID) [20
] and the Hamilton Rating Scale for Depression (HRSD) [21
]. They ask participants to describe past and current treatments for depression and use of antidepressants. If participants meet all inclusion and none of the exclusion criteria for the study, they enter the study.
Participants can withdraw from treatment or from the study at any time.
Safety monitoring and reporting
The trial protocol has been approved by an independent medical ethics committee (METIGG). Eligible people will only be included as participants in the trial after informed consent has been obtained.
We record and report suspected serious adverse events to the Multi-center Ethic Committee (METIGG) according to their individual guidelines.
For the baseline assessment we ask participants themselves to complete the web-based self-report questionnaires in two packages, i.e. explicit and implicit measures. The first part with assessments starting directly within a week, the second part containing implicit measures will be offered within 2 days after completion of self-report assessments measures. The following self-report measures will be used: the Inventory of Depressive Symptomatology, IDS-SR [19
], Nemesis Somatic illnesses list [23
], negative life events (Life events questionnaire, LGV) [24
], self-esteem (Self-esteem Questionnaire) [25
], personality pathology (Personality Diagnostic Questionnaire, PDQ-4) [26
], everyday problems (EPCL) [27
], hypomania (HCL-32) [28
], direct and indirect costs (TIC-P) [22
] and Medication Adherence Questionnaire (MAQ) [29
], a measure of general quality of life (Euro-QOL EQ-5D) [30
], and rumination (Ruminative Responses Subscale of the Response Styles Questionnaire RSQ) [31
], dysfunctional attitudes (Dysfunctional Attitudes Scale, DAS) [32
], LEIDS [33
], acceptance (Acceptance and Action Questionnaire, AAQ) [34
], coping (Utrecht Coping List, UCL) [35
], Mastery 7 [36
]. After 6 weeks this set will be repeated with the exception of the TIC-P, LGV, PDQ, MAQ EQ-5 D and Nemesis Somatic illnesses list. During follow-up every three months the following self-report assessments will be repeated: IDS-SR, HCL-32, TIC-P, EPCL, Mastery7 and EQ-5 D will be administered. For a complete overview of the assessments see Table . Participants in the M-CT group will also answer questions of the Dutch version of the Credibility and expectancy questionnaire [37
] before and after finishing M-CT.
For an overview of the assessments at baseline, in between- and post treatment and follow up assessments see Table .
Primary outcome is time until relapse or recurrence of depression in the experimental group relative to the control group over 24 months using DSM-IV-R criteria as assessed by the SCID at 3 months, 12 months and 24 months [20
]. Co-morbidity with concurrent chronic somatic illnesses will be assessed using the NEMISIS somatic illnesses list [23
]. Secondary outcome is symptom severity as measured with the Inventory of Depressive Symptomatology (IDS-R) [19
] and number of relapses as assessed by the SCID [20
]. For the economic evaluation we will use the EuroQoL (EQ-5D) to obtain a generic quality of life related outcome [30
]. Cost data related to health care uptake will be sampled using the TIC-P [22
]. Cost data stemming from production losses due to absenteeism and working less efficiently while at work will be collected with the specific modules from the TIC-P [22
Moderators and Mediators
For potential moderators (illness related, stress-related and cognitive-related) predictors and mediators the following self-report measures will be used at baseline, at 1,5 and at 3 months (internet based): Inventory of Depression Symptomatology (IDS-SR every 3 months) [19
], Dysfunctional Attitude Scale, (DAS-A) [32
], LEIDS [33
], Everyday Problem Checklist (EPCL) [27
], Negative Life Events Questionnaire [24
] and to assess nonadherence to AD with the Medication Adherence Questionnaire (MAQ) [29
], Mastery7 [36
]. To enable calculating quality adjusted life years required for the economic evaluation the EQ5 D will be administered every 3 months [30
]. To test whether M-CT affects implicit attitudes and attentional bias differentially and whether residual difficulty to disengage and residual dysfunctional implicit attitudes are related to the return of depressive symptoms, an web-based Implicit Association Test (IAT) [39
] will be used to assess implicit attitudes. A web-based rapid serial visual presentation (RSVP) [40
] task will be used to assess the difficulty to disengage from negative information. Difficulty to disengage will be indexed by the magnitude of the attentional blink when negative self descriptors are presented as the first target and neutral words as the second. For an overview of the assessments at baseline, in between- and post treatment and follow up assessments see Table .
Cox regression analysis (survival analysis) will be performed, including as covariates the stratification variables that will be used in randomization, i.e.: number of previous episodes and type of care (primary/secondary/no care). Analysis will be conducted by intention to treat, including all subjects randomized in the study, and per protocol, including only subjects satisfying protocol. Statistical significance will be set at P < .05.
Mixed-model analysis of variance (ANOVA) will be used for the other variables, including baseline values of the dependent variable as a covariate in all analyses. We shall use implicit and explicit cognitive measures and stress measures (daily hassles) to explore the extent to which they mediate relapse and recurrence during treatment and follow up.
For the economic evaluation the balance between costs and health outcomes will be compared across strategies using a societal perspective. Primary outcome measure: the number of depression-free days. Both short-term and long-term consequences will be compared. Additionally, Quality Adjusted Life Years will be used as outcome (see also Table ).