Randomised trials on the effects of different psychotherapy formats showed that a more intense dose of treatment can lead to a faster response, less drop-out and better or non-inferior outcomes compared with less-intense treatments.Reference Bohni, Spindler, Arendt, Hougaard and Rosenberg1–Reference Foa, McLean, Zang, Rosenfield, Yadin and Yarvis4 In the field of depression, a meta-regression analysis showed that the number of sessions per week, and not the total number of sessions or duration of therapy, was correlated with outcome.Reference Cuijpers, Huibers, Daniel Ebert, Koole and Andersson5 In an observational study, a higher session frequency was associated with improvement and recovery within the first three months of treatment.Reference Tiemens, Kloos, Spijker, Ingenhoven, Kampman and Hendriks6 However, a randomised trial on the direct effect of psychotherapy session frequency on depression is lacking. The original Beck manual for cognitive–behavioural therapy (CBT) for depressionReference Beck, Rush, Shaw and Emery7 prescribes that CBT should start twice weekly in the beginning of treatment, whereas interpersonal psychotherapy (IPT) should be delivered once a week, according to Klerman's manual.Reference Klerman, Weissman, Rounsaville and Chevron8 In both treatments, however, a higher session frequency might lead to better depressive outcomes. Our hypothesis is that a higher session frequency leads to better recall of the content of the sessions, which will lead to a better development of therapy-specific skills and consequently better treatment outcomes.Reference Bruijniks, Bosmans, Peeters, Hollon, van Oppen and van den Boogaard9 An alternative hypothesis is that a higher session frequency leads to a better working alliance, which then leads to better patient adherence and motivation, and consequently better treatment outcomes.
This is the first randomised trial to investigate the effects of once weekly versus twice weekly sessions of CBT and IPT for depression. The primary hypothesis was that twice weekly sessions in the first 16 sessions of treatment leads to greater reduction of depressive symptoms and other outcomes in the acute phase of treatment. We also assessed the effects on attrition and time to response, and potential differences between CBT and IPT session frequency.
This study was registered with the Netherlands Trial Register (registration number NTR4856). Details about the study design have been described elsewhere.Reference Bruijniks, Bosmans, Peeters, Hollon, van Oppen and van den Boogaard9 Data are from a multicentre, randomised controlled trial (2 × 2 factorial design) on the effects of session frequency in CBT and IPT for depression, with a 24-month follow-up. Patients were randomly assigned to four conditions: 16 sessions of CBT twice a week, 16 weekly sessions of CBT, 16 sessions of IPT twice a week and 16 weekly sessions of IPT. In each condition, the 16th session was followed by four biweekly sessions. Data presented here concern post-treatment outcomes 6 months after the start of treatment; follow-up data will be reported in a subsequent report. Clinical outcomes on the long-term effects (Longitudinal Interview Follow-up Evaluation) and cost-effectiveness (quality of life, EuroQol-5D and the Trimbos and Institute for Medical Technology Assessment Questionnaire on Costs associated with Psychiatric Illness) outcomes will be reported in follow-up papers.
Participants and recruitment
Patients were adult out-patients referred to one of nine Dutch specialised mental healthcare centres located across the Netherlands. Inclusion criteria were as follows: (a) a primary diagnosis of a DSM-IV or DSM-5 major depressive disorder (including chronic depression) or DSM-5 persistent depressive disorder as confirmed by the Structured Clinical Interview for DSM-IV Axis I DisordersReference First, Spitzer, Gibbon and Williams10 or the Mini International Neuropsychiatric Interview PlusReference Van Vliet and De Beurs11, (b) aged 18 to <65 years, (c) sufficient knowledge of the Dutch language, (d) pre-treatment score ≥20 on the Beck Depression Inventory-II (BDI-II)Reference Beck, Steer and Brown12 and (e) access to internet facilities (because some of the assessments were online). Exclusion criteria were as follows: (a) starting antidepressants or dosage change in the past 3 months, (b) acute suicide risk, (c) DSM-IV or DSM-5 diagnosis of drug or alcohol dependence, (d) presence of a DSM-IV or DSM-5 diagnosis of a cluster A or B personality disorder as evaluated by a clinician during the intake with or without help of a structured interview and (e) received more than five sessions of adequate CBT or IPT in the previous year (clinician-evaluated at intake).
Patients were recruited during regular intakes at the clinical sites. Eligible patients were approached for participation in the study and received a general information folder. A week after the intake, an independent research assistant contacted eligible patients to give them additional study information and to check whether the patient was motivated to participate in the study. Subsequently, patients were invited for a diagnostic interview (if not completed during intake), an online baseline assessment and signing of informed consent. Informed consent for videotaping the sessions was signed separately; patients who were included in the trial were allowed to refuse videotaping of their therapy sessions. After the baseline assessment, block randomisation took place. Randomisation codes were generated by an independent computer scientist to ensure allocation concealment. Patients were randomised into one of the four conditions, using a computer script (allocation ratio: 1:1:1:1). Randomisation was stratified according to severity (high, BDI-II score ≥29; low, BDI-II score ≤28) and treatment site, using different blocks for high versus low severity and each treatment site. After baseline, online assessments were conducted 2 weeks after the start of treatment and every month until month 6. In addition, participants filled out the BDI-II and single questions on mood and happiness online in the therapist's office before the start of each therapy session. Post-treatment effect size for the difference in depressive symptoms between the session frequencies was estimated to be around 0.45.Reference Cuijpers, Huibers, Daniel Ebert, Koole and Andersson5 Therefore, 200 patients were needed for adequate power (alpha = 0.05, power (1−ß) = 0.80, two-tailed) to detect a difference between session frequencies, taking 25% drop-out into account.
Treatment manuals were used for both interventions regardless of frequency and both treatments consisted of a minimum of 12 and maximum of 20 45-minute face-to-face sessions, depending on patient progress. CBT was based on the manual by Beck et al Reference Beck, Rush, Shaw and Emery7 and IPT was based on the manual by Klerman et al.Reference Klerman, Weissman, Rounsaville and Chevron8 Except for the structuring of session frequency, no changes were made to the original treatment manuals after the seeking advice from various CBT and IPT experts. Participants who were randomised to the condition with twice weekly sessions received 16 sessions during the first 8 weeks of treatment, and 4 sessions during the last 8 weeks (up to 20 sessions over a period of 16 weeks). Patients who were randomised to the condition with once weekly sessions received 16 sessions during the first 16 weeks and 4 sessions during the last 8 weeks (up to 20 sessions over a period of 24 weeks).
Before the study, therapists received 2 days of training by S.D.H. (CBT) or Dr Holly Swartz (IPT). During the study, therapists received supervision and consulted each other on current cases in on-site consultation sessions. The format and the amount and frequency of supervision and consultation sessions differed per centre and therapist, varying from 30 min to 2 h biweekly (i.e. in most centres, these meetings also included discussion of patients that did not participate in the study). Therapists that were still in postgraduate clinical training received additional weekly to monthly supervision from an experienced supervisor in the context of the study. Therapists that entered the study after the training received additional training and elaborate instructions about the study and protocol. To prevent contamination, therapists were uniquely assigned to one of the treatment conditions. However, because of problems with capacity, 7 therapists treated 37 patients in both conditions.
The modified version of the Collaborative Study Psychotherapy Rating Scale version 6 (CSPRS-6Reference Lemmens, Arntz, Peeters, Hollon, Roefs and Huibers13,Reference Hollon, Waskow, Evans and Lowery14 ) was used as a measure of adherence (i.e. whether the therapy protocol was followed). Treatment adherence was measured by independent expert raters (3 psychotherapists and 1 psychologist) that assessed 51 randomly selected videos of different patients drawn between sessions 6 and 12. To examine therapy competence (i.e. how well the therapy was performed), master-level clinical graduate students rated a total of 116 unique CBT videos and 131 unique IPT videos. Total scores on the Cognitive Therapy ScaleReference Dobson, Shaw and Vallis15 (CTS; range 11–77), the IPT Adherence and Quality Scale (short version)Reference Stuart16 (IPT-AQS; range 9–125) and the score on the single item ‘How would you rate the clinician overall in this session, as a CBT/IPT therapist?’ measured on a Likert scale ranging from 1 (poor) to 7 (excellent) were used. Details about treatment integrity can be found in Data Supplement 1 available at https://doi.org/10.1192/bjp.2019.265.
Primary outcome was depression severity as measured with the BDI-II.Reference Beck, Steer and Brown12 The BDI-II is a 21-item self-report instrument assessing depressive symptoms over the past 2 weeks. A score of 0–13 indicated minimal depression, 14–19 indicated mild depression, 20–28 indicated moderate depression and 29–63 indicated severe depression. Reliability and validity of the BDI-II has been previously reported.Reference Beck, Steer and Garbin17
Secondary outcomes included a single item rating of ‘current mood’ (0, ‘worst mood ever’ to 100, ‘best mood ever’); two single items rating happiness, ‘general life happiness’ and ‘happiness today’ (1, ‘completely unhappy’ to 7, ‘completely happy’); the general health perception domain of the RAND 36-Item Health Survey (RAND-36);Reference Ware and Sherbourn18,Reference Zee and Sanderman19 and the Remission of Depression QuestionnaireReference Zimmerman, Martinez, Attiullah, Friedman, Toba and Boerescu20 (RDQ). The RDQ measures remission as a broad construct including reduction in depressive symptoms and other depression-related domains, with lower scores representing better functioning on each subscale. The RDQ and RAND-36 were measured at baseline and at 3 and 6 months after the start of treatment. The mood and happiness questions were both included before each session and in the online assessments. For a detailed description of these instruments and their psychometric properties, we refer to our protocol paper.Reference Bruijniks, Bosmans, Peeters, Hollon, van Oppen and van den Boogaard9
The flow chart, pre-treatment variables, treatment and study adherence, therapists characteristics and treatment integrity were described in the result section. Differences in days between sessions between frequency conditions were tested with univariate analysis of variance. Differences in attrition between session frequencies were tested with a χ2 test.
Multilevel regression analyses with maximum likelihood estimation were conducted to investigate the effect of frequency condition on depression (BDI-II scores). Intervention was represented by two dichotomous variables: CBT (−1) versus IPT (1) and once weekly (−1) versus twice weekly (1) sessions. Because of the long waiting time (mean 35.20 days, s.d. 27.05) between baseline and session 1 (see also Data Supplement 3), we decided to add the BDI-II score measured at the start of session 1 to the outcome (i.e. dependent variable was BDI-II score measured before session 1, and 2 weeks and monthly after start of treatment up to month 6, leading to a total of eight time points). The model corrected for baseline values by adding BDI-II baseline as a covariate in the model.Reference de Boer, Waterlander, Kuijper, Steenhuis and Twisk21 The initial basic model was a two-level model with repeated measurements (level 1) nested within patients (level 2) with two two-way interactions testing the difference in change of BDI-II scores over time in days for the different session frequencies (time×frequency) and treatments (time×treatment). Time in days was centred 90 days after the start of treatment and it was tested whether a quadratic or cubic function of time led to better model fit. Subsequently, the addition of random levels on therapist and treatment centre were tested, followed by a test of random slopes for time on all fitted random levels. Fit of different (co)variance structures for both the random effects and residuals errors within the lowest level (repeated measurements) were compared. To explore whether the effect of session frequency was different for CBT versus IPT, a model with a three-way interaction (frequency × treatment × time) was fitted. Analyses were intention to treat, and significance levels were set at P < 0.05. The same method was used to investigate effects on the secondary outcomes.
Effect sizes from session 1 to month 6 (Cohen's d)Reference Cohen22 for the primary outcomes were computed: (estimated session 1 mean – estimated mean at time i) / estimated s.d. at session 1). Response (reliable change) was defined as a decrease of at least 9 BDI-II points from session 1 to month 6 and remission (clinically significant change) as an absolute value ≤9 on the BDI-II at month 6. Cox regression was conducted to examine differences in the time in days to response and remission for weekly versus twice weekly sessions, including treatment modality and the BDI-II baseline score as covariates.
Three sensitivity analyses were conducted. First, it was investigated whether the total number of sessions, the presence of a comorbid anxiety disorder or the use of antidepressants were covariates that influenced the main results. Second, because a number of the time points (week 2, month 1, month 2, month 3 and month 4) are not comparing the same number of treatment sessions between frequency conditions and this could potentially skew the estimates in the analysis, we additionally tested the change in BDI-II scores from baseline to month 6, ignoring symptom levels across the other time points. Third, a completers analysis was conducted, reporting estimated difference in BDI-II scores at month 6 and the difference in effect size between frequency conditions. Non-completers were defined as patients who discontinued therapy, deviated to other protocols or changed antidepressant use before session 12. All other patients were considered completers.
Results are reported according to the Consolidated Standards Of Reporting Trials guidelines.Reference Boutron, Altman, Moher, Schulz and Ravaud23 For more details about the data analyses we refer to Data Supplement 2.
The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008. All procedures involving human patients were approved by the Medical Ethical Committee of VU Medical Centre Amsterdam (registration number 2014.337).
Description of the sample
The flow of patients is shown in Fig. 1. A total of 3179 patients were initially screened for eligibility, of whom 2380 did not meet the inclusion criteria and 235 patients met the inclusion criteria but declined to participate. This resulted in a total sample of 200 patients (CBT weekly, n = 49; CBT twice weekly, n = 49; IPT weekly, n = 55, IPT twice weekly, n = 47).
Mean age at baseline was 37.85 years (s.d. 12.26) and 61.5% of participants were female. Mean baseline BDI-II score was 34.70 (s.d. 9.96), 43.3% of the sample experienced recurrent depression and self-reported duration of symptoms was 40.87 months (s.d. 68.31; median 20). The majority (68%) of the total sample was still actively employed. Pre-treatment characteristics per condition are presented in Table 1. Observed means (s.d.) on all outcome measures at each time point for each condition are given in Table 2.
Note that severe depression is defined as a BDI-II score >28. Low educational level has been defined as no former education or special lower education or primary school or practical training school; middle educational level has been defined as completing lower general secondary education or higher general secondary education or intermediate vocational education; and higher education level has been defined as completing higher vocational education or pre-university education or university. Note that students were also considered as active employment. The RAND-36 reports the total score of the subscale ‘general experience of health’, range of 5 (low perception of general health) to 25 (high perception of general health). Comorbid Axis I disorders were established with use of the MINI-Plus or SCID-I. In case of missing data, the number of available data per condition is given. Data were missing for first depressive episode and number of episodes: CBT weekly (missing n = 5), CBT twice weekly (missing n = 10), IPT weekly (missing n = 8) and IPT twice weekly (missing n = 6); and for MINI-Plus or SCID-I: CBT weekly (missing n = 5), CBT twice weekly (missing n = 1), IPT weekly (missing n = 3) and IPT twice weekly (missing n = 4). CBT, cognitive–behavioural therapy; IPT, interpersonal psychotherapy; RAND-36, RAND 36-Item Health Survey; BDI-II, Beck Depression Inventory-II; PTSD, post-traumatic stress disorder; MINI-Plus, Mini International Neuropsychiatric Interview Plus; SCID-I, Structured Clinical Interview for DSM-IV Axis I Disorders.
Note that not all patients completed all measurements, total lost to follow-up per time point on the BDI-II scores can be found in Fig. 1. The RAND-36 reports the total score of the subscale ‘general experience of health’, range of 5 (low perception of general health) to 25 (high perception of general health). Note that for the RDQ, lower scores represent better functioning on each subscale. Missing data on the secondary outcomes differed from missing data on the BDI-II (also see flow chart) on the following time points (number of missing data is given): RAND-36 week 2, CBT weekly n = 23 and CBT twice weekly n = 15; RAND-36 month 3, IPT weekly n = 21 and IPT twice weekly n = 8; RAND-36 month 6, CBT twice weekly n = 10 and IPT weekly n = 19; current mood week 2, CBT twice weekly n = 14; current mood month 1, IPT weekly n = 13; current mood month 3, IPT weekly n = 21; current mood month 5, CBT twice weekly n = 20, IPT weekly n = 24 and IPT twice weekly n = 12; current mood month 6, CBT weekly n = 11, CBT twice weekly n = 11 and IPT weekly n = 19; happiness items week 2, CBT twice weekly = 14; happiness items month 1, IPT weekly n = 13; happiness items month 3, IPT weekly n = 21 and IPT twice weekly n = 8; happiness items month 5, IPT weekly n = 24 and IPT twice weekly n = 12; happiness items month 6, CBT weekly = 11, CBT twice weekly = 10 and IPT weekly = 20; RDQ depressive symptoms month 3, IPT weekly n = 21 and IPT twice weekly n = 10; RDQ depressive symptoms month 6, IPT twice weekly n = 9; RDQ coping month 3, IPT weekly n = 21; IPT twice weekly n = 10; RDQ other symptoms month 3, IPT weekly n = 21 and IPT twice weekly n = 10; RDQ positive health month 3, IPT weekly n = 21 and IPT twice weekly n = 10; RDQ functioning month 3, IPT weekly n = 21 and IPT twice weekly n = 11; RDQ life satisfaction month 3, IPT weekly n = 22 and IPT twice weekly n = 10; RDQ well-being month 3, IPT weekly n = 21 and IPT twice weekly n = 10. CBT, cognitive–behavioural therapy; IPT, interpersonal psychotherapy; RAND-36, RAND 36-Item Health Survey; BDI-II, Beck Depression Inventory-II; RDQ, Remission of Depression Questionnaire.
Treatment and study adherence
Recruitment took place from November 2014 to January 2018. The patients that started treatment (n = 191) received a mean number of 16.54 (s.d. 4.73) sessions. There was a significant difference in days between the sessions and treatment duration between patients who received the weekly (mean 8.98 days, s.d. 3.06; mean 170 days, s.d. 67.23) versus the twice weekly (mean 4.68 days, s.d. 2.28; mean 113 days, s.d. 43.62) frequency condition (t(177) = 10.92, P = <.001 and t(165.68) = 6.98, P = <0.001 for session frequency and treatment duration, respectively). There was no difference between frequency conditions in the total number of sessions. Of the 48 patients that were registered as non-completers, 14 withdrew or were lost to follow-up, 9 never started with treatment, 20 switched to another intervention or started or switched antidepressants before the 12th session, 2 received a session frequency different from the one to which they were randomised and 3 could not complete treatment because of a lack of therapists at the treatment centre. Four patients who completed treatment before session 12 were considered to be completers because their therapist indicated that therapy was successful. More patients dropped out of weekly (n = 32) than twice weekly (n = 16) treatment (χ 2 = 5.44, P = 0.02). With regard to study adherence, the percentages of participants completing assessments on each time point were 58% for week 2, 78.5% for month 1, 76.5% for month 2, 71.5% for month 3, 68% for month 4, 65.5% for month 5 and 72.5% for month 6. Study adherence was not different between frequency conditions, except for month 3, in which there were more missing data in the weekly condition (n = 37 compared with n = 20; χ2 = 5.32, P = 0.02). Treatment characteristics per condition are given in Data Supplement 3.
The 76 therapists (25–61 years of age, 81.6% female) who participated in the study were licensed mental health psychologists (n = 38), master-level (MSc) psychologists in post-master training (n = 32), psychiatrists (n = 2) and mental health nurses (n = 4). Therapists had an average of 9.55 years (s.d. 8.29; n = 64) of clinical experience (range 0.5–29 years). The number of patients per therapist ranged from 1 to 12 (mean 2.67, mode 1). IPT therapists had fewer years of experience with IPT than CBT therapists had with CBT (IPT: mean 4.01, s.d. 6.50; CBT: mean 6.46, s.d. 6.13). There were no significant differences in treatment outcome (t(143) = 1.19, P = 0.23) or attrition (t(198) = 0.79, P = 0.42) between patients that were treated by single versus multiple modality therapists.
A total of 1771 videos of 129 patients were available. CBT-specific behaviour (adherence) as rated with the CSPRS-6 was significantly more evident in CBT than in IPT (CBT: mean 89.92, s.d. 33.84, n = 27 versus IPT: mean 57.56, s.d. 10.19, n = 23, t(31.41) = 4.72, P < 0.001). IPT-specific behaviour as rated with the CSPRS-6 was more present in IPT than in CBT (IPT: mean 95.50, s.d. 20.26, n = 24 versus CBT: mean 32.59, s.d. 5.38, n = 27; t(25.88) = −14.75, P < 0.001). CBT competence as rated with the CTS varied from poor to very good (Likert scale: mean 3.69, s.d. 1.08, n = 61; CTS total: mean 45.55, s.d. 9.82, n = 60), and 16.3% of the CBT sessions were rated good to excellent. IPT competence as rated with the IPT-AQS varied from poor to very good (Likert scale: mean 3.64, s.d. 1.27, n = 66; IPT-AQS: mean 66.48, s.d. 14.02, n = 66) and 12.1% of the IPT sessions were rated good to excellent. There were no differences in treatment adherence or treatment competence between the once weekly versus the twice weekly conditions.
Effect of session frequency over time
Results show a significant effect of session frequency on BDI-II scores over time in favour of twice weekly sessions (estimated mean difference between weekly versus twice weekly sessions at month 6: 3.85 points on the BDI-II, difference in effect size d = 0.55; estimated mean scores: 31.05 (95% CI 29.29–32.81) to 23.01 (95% CI 19.45–26.57) for once weekly CBT; 32.58 (95% CI 30.84–34.32) to 20.69 (95% CI 17.18–24.21) for twice weekly CBT; 32.49 (95% CI 30.77–34.21) to 24.30 (95% CI 20.77–27.83) for once weekly IPT; and 34.02 (95% CI 32.25–35.79) to 21.98 (95% CI 18.44–25.52) for twice weekly IPT; also see Data Supplement 4). There were no differences in change over time on the BDI-II between CBT and IPT. There was no different effect of session frequency in CBT versus IPT. Change of BDI-II scores over time between conditions is given in Fig. 2. Corrected mixed-model estimated means (95% confidence intervals) on all outcome measures at each time point for each condition are given in Data Supplement 4. The mixed regression model on the BDI-II scores can be found in Data Supplement 5. Within- and between-group effect sizes can be found in Data Supplement 9.
Sensitivity and completers analyses
The total number of sessions, the presence of a comorbid anxiety disorder or the use of antidepressants were not significant covariates and did not influence the results. In addition, the effect of session frequency remained significant when testing change in BDI-II scores from baseline to month 6, without including the other time points. In congruence with the intention-to-treat analyses, results show a significant effect of session frequency on BDI-II scores over time in favour of twice weekly sessions when running the analysis in completers only (n = 152; estimated BDI-II mean difference between weekly versus twice weekly sessions at month 6, 3.80; difference in effect size d = 0.53) and there were no differences in change over time on the BDI-II between CBT and IPT.
Analysis of secondary outcome measures
There was a significant difference between session frequencies in current mood, happiness in general and happiness today over time in favour of twice weekly sessions (current mood estimated mean difference at month 6, 3.98; happiness in general estimated mean difference at month 6, 0.25; happiness today estimated mean difference at month 6, 0.42). Results show trends toward significance on the RAND-36 (in favour of the twice weekly sessions; estimated mean difference at month 6, 2.45) and RDQ positive health (twice weekly session estimated mean difference at month 6, 1.49; IPT estimated mean difference at month 6, 1.25). There were no differences between frequencies or treatments on change over time on the other secondary outcomes. Model descriptions can be found in Data Supplement 6. Corrected mixed-model estimated means (95% confidence intervals) on all outcome measures at each time point for each condition are displayed in Data Supplement 4.
Reliable change and clinically significant change
At month 6, 44.4% (CBT) and 32.3% (IPT) of the patients showed reliable change in the weekly conditions, whereas 60% (CBT) and 52.9% (IPT) of the patients showed reliable change in the twice weekly conditions (χ 2 = 4.27, P = 0.039). At month 6, 21.6% (CBT) and 17.6% (IPT) of the patients showed clinically significant change in the weekly conditions, whereas 25.7% (CBT) and 33.3% (IPT) of the patients showed clinically significant change in the twice weekly conditions (χ 2 = 1.94, P = 0.16). Reliable and clinically significant change rates are presented in Data Supplements 7 and 8.
Time to response and remission
There was a significant difference between frequency conditions in time to response, in favour of twice weekly sessions (hazard ratio 1.48, 95% CI 1.00–2.18, P = 0.049), when controlling for treatment modality and BDI-II baseline scores. There was no significant difference between session frequencies in time to remission (hazard ratio 1.22, 95% CI 0.68–2.18, P = 0.49).
To our knowledge, this was the first randomised trial to investigate the effects of once weekly versus twice weekly sessions of CBT and IPT in patients with depression. We found that twice weekly sessions lead to a greater reduction in depressed symptoms, less time to response and lower attrition compared with weekly sessions. The effect of session frequency did not differ between CBT and IPT and there was no significant difference between frequency conditions in time to remission.
The finding that only a small proportion of the patients in the present study showed clinically significant change (17.6–33.3%) and most patients still showed moderate levels of depression post-treatment contrasts the findings of three other recent, randomised CBT-IPT studies for depression that report end scores of minimal to mild levels of depressionReference Lemmens, Arntz, Peeters, Hollon, Roefs and Huibers13,Reference Luty, Carter, McKenzie, Rae, Frampton and Mulder24,Reference Quilty, McBride and Bagby25 and higher rates of clinically significant change.Reference Lemmens, Arntz, Peeters, Hollon, Roefs and Huibers13 However, our findings are in line with a fourth CBT-IPT trialReference Ekeblad, Falkenström, Andersson, Vestberg and Holmqvist26 that also reported moderate post-treatment levels of depression. This latter study was similar to ours in that they also included a large number of therapists, whereas the trials that report lower levels of depression post-treatment had less variation in settings and therapists, and provided extensive on-site supervision.Reference Lemmens, Arntz, Peeters, Hollon, Roefs and Huibers13,Reference Luty, Carter, McKenzie, Rae, Frampton and Mulder24,Reference Quilty, McBride and Bagby25 One explanation for our findings might be that the large variation in therapists and treatment settings resulted in suboptimal levels of treatment quality and subsequent high average levels of depression and low clinically significant change at the end of treatment.
A major strength of the present study is its external validity. Although the present study has certain characteristics of an efficacy study, the participation of multiple treatment centres and a large number of therapists with varying educational backgrounds (i.e. from less than a year of experience to 29 years of experience) was necessary to reach the inclusion rate of 200 patients. Therefore, the present study can best be qualified as an ‘effectiveness’ trial that generalises to clinical practices outside academic research settings. Moreover, because the quality of the delivered treatments in the present study was only moderate on average, the effects of session frequency might be even larger in therapy settings with a higher level of treatment quality. Another strength of the study is the fine-grained assessment of session frequency: because patients completed online questionnaires before each session, we were able to record the exact difference in session frequency. In addition, although the implementation of a higher session frequency is a challenge that will require an organisational change, the finding that the delivery of twice weekly sessions was entirely possible in this effectiveness trial is a promising finding in regard to future implementation.
Several limitations should be considered. First, the ‘pure’ effect of frequency may have been confounded by the difference in treatment duration (i.e. total time being in therapy) between the frequency conditions. However, this difference was deliberate in the design of this trial, and relates to cost-effectiveness (to be reported in a follow-up paper), i.e. the assumption that more frequent sessions leads to less indirect costs (i.e. a reduction of work-days lost). Second, there were only a few patients with a low educational background and the number of patients with recurrent depression was low compared with earlier randomised controlled trials.Reference Ehlers, Ph, Hackmann, Psy, Grey and Wild2,Reference Cuijpers, Huibers, Daniel Ebert, Koole and Andersson5 These groups should therefore be considered as underrepresented in the present study. Third, the more frequent assessment of BDI-II at the beginning of each session might have contributed to the lower scores in the twice weekly condition, as two preliminary studies suggest that more frequent BDI assessments can lead to lower BDI-II scores.Reference Atkeson, Calhoun, Resick and Ellis27,Reference Longwell and Truax28 However, such an effect could have only have occurred in the first 3 months and would have disappeared in the last 3 months, when BDI-II assessments in the twice weekly condition were no longer more frequent and equalled out with the total number of BDI-II assessments in the weekly condition. Fourth, the present study was not powered to detect small effects or a difference between the four treatment conditions. The lack of an interaction between session frequency and treatment modality should therefore be considered as a preliminary result. Fifth, all outcome measures were self-reported and the clinical status of the participants at post-treatment could not be confirmed in the absence of an observer-rated outcome measure. Sixth, no psychometric properties are available for the IPT-AQS yet, and therefore conclusions about the quality of IPT should be considered preliminary.
In summary, the present study demonstrates that in clinical practice settings both CBT and IPT are best provided twice weekly. This finding implies that reorganisation of the specialised mental healthcare services for depression (going from once a week to twice a week) can lead to less attrition, quicker response and better outcomes across the course of treatment.
Supplementary material is available online at https://doi.org/10.1192/bjp.2019.265.
This study was funded by ZonMw and Stichting tot Steun VCVGZ. The funders were not involved in the study design; the collection, analysis and interpretation of data; the writing of the report or the decision to submit the article for publication.
We acknowledge the contribution of participating patients and therapists at Altrecht, GGZ inGeest, GGZ Oost-Brabant, Pro Persona, PsyQ Amsterdam, PsyQ Den Haag, PsyQ Haarlem, PsyQ Leiden and Riagg Maastricht. Without their cooperation the trial would have been impossible to conduct. Furthermore, we gratefully thank Danielle Tilburgs, Nicole Billingy, Kris Wijma, Sofie Jansen and Annie Hendriks for their assistance during the data collection. In addition, we also acknowledge Carolien Christ, Kosse Jonker, Ina Leeuw and Dina Snippe for their contributions to the expert treatment integrity check and thank all students that have been involved in the ratings of treatment competence.
Data will be available upon request.
S.D.H., J.B., F.P.M.L.P., P.v.O., P.C., A.A. and M.J.H.H. obtained funding for this study. All authors contributed to the design of the study. F.P.M.L.P., P.D., L.W., P.v.O., M.v.d.B. and J.S. were involved in the recruitment of patients, therapists and coordination of the treatments. S.J.E.B. coordinated the recruitment of patients, data collection and integrity checks. S.J.E.B. conducted the analyses. S.J.E.B., M.J.H.H., J.T. and A.A. were involved in the interpretation of the analyses. S.J.E.B. wrote the manuscript with contributions from L.H.J.M.L. and M.J.H.H. All authors read, contributed to and approved the final manuscript.