Study type and design
This study is a two-parallel group open RCT. An open RCT means that those who administer the treatment and the study subjects are open to treatment code, but the statistician stays blinded to it until a blind review is done for the primary and secondary outcome variables. Blinding is an established approach in clinical trials to minimize the risk of performance and detection bias. Clinical trial guidelines recommend that statisticians remain blinded to allocation prior to the final analysis (Chan et al., 2013). In this RCT, pregnant women with depressive symptoms are randomized 1:1, by study site, into the Stronger Together iCBT intervention with telephone coaching or the educational control group. Both groups also receive standard treatment, according to established regional standard care guidelines, and this may involve therapy and medication. However, ongoing psychotherapy at recruitment is an exclusion criterion (Table 1). The study outline is presented in Figure 1.
Figure 1. Flowchart of the RCT study
Recruitment
Finnish maternity clinics provide regular check-ups 11-15 times during pregnancy to monitor the wellbeing of the fetus, pregnant woman and the whole family. Up to 99.8% of pregnant women in Finland use these services each year (Ministry of Social Affairs and Health, 2023). This RCT consists of all the women attending regular maternity check-ups in the Varha, Siun sote and Soite regions and the cities of Espoo, Tampere, Pori and Oulu. These areas account for approximately 14 000 live births per year, which is nearly a third of the infants born in Finland each year (Finnish Institute for Health and Welfare, 2021). The women are screened with the EPDS, which is one of the most widely studied and validated self-report instruments for depressive symptoms during the perinatal period (Levis et al., 2020). Finnish national care guidelines state that the EPDS should be routinely used during health check-ups at 13-18 weeks of gestation.
All the women attending maternity clinics in the study areas are given an envelope during the first face-to-face health visit at 8–10 weeks of gestation. This includes an information letter, informed consent document and the EPDS questionnaire. Healthcare professionals also give the women verbal information about the ongoing study. The women are asked to complete the forms by the next clinic check-up at 13–18 weeks and provide informed consent if the research team is allowed to contact them about taking part in the RCT. The women are also provided with contact details to the study team in case they have any questions.
The informed consent forms and EPDS questionnaires are collected during the routine check-up at 13–18 weeks. The women can fill in the forms during their visit if they have forgotten them. Scores of 10 points or more on the EPDS scale of 0 to 30 indicates possible depression. Those who have reached these scores, and agreed to be contacted by the study group, are invited to undergo an eligibility assessment for the RCT. As the EPDS form is routinely used by maternity health clinics, any symptoms are also discussed with the public health nurses working in the clinics. The public health nurses can contact the local study nurses if they have any questions.
Eligibility criteria
The study coaches call the women to introduce the study in more detail and screen them for eligibility, according to the inclusion and exclusion criteria in Table 1. The coaches are specially trained to deliver the intervention and they receive weekly supervision from two senior CBT therapists. Phone calls are recorded for quality control.
Table 1. Inclusion and exclusion criteria
If the woman is eligible, she is asked to provide informed consent on a secure website. At the start of the baseline assessments, the women fill in the EPDS scale again to assess how stable their depressive symptoms are. If they score at least nine points, they proceed to the full digital baseline assessment. If they score eight points or less, they are excluded. A separate phone call is made after the baseline questionnaires. In this phone interview, the women are assessed using selected sections of the structured diagnostic interview, Mini-International Neuropsychiatric Interview (MINI) (Sheehan et al., 1998). The included sections cover depression, suicidality, bipolar disorder and anxiety disorders. If the woman reports current suicidality she is excluded from the study and referred to appropriate services. Other exclusion criteria include psychotic and bipolar disorders, substance abuse, active ongoing psychotherapy, participating in another intervention and multiple pregnancy (Table 1). The women can be referred to local health care services if more individual treatment is considered necessary and the coaches can also consult a psychiatry specialist at any time.
Intervention
Stronger Together is a targeted iCBT intervention for pregnant women with depressive symptoms who score ≥9 points on the EPDS. It is based on the Coping with Depression course, which has been effective in preventing and treating depression in several RCTs and has been implemented in different target populations and settings (Cuijpers et al., 2009). The highly structured Stronger Together intervention includes basic CBT components: psychoeducation, behavioral activation, coping with social relationships, cognitive restructuring and preventing setbacks (Table 2). It comprises seven weekly themes on a digital platform and weekly phone coaching. There will be also two booster sessions approximately two and seven months after the birth, including a phone call and a digital component. No physical visits or face-to-face communication are included. The aim is to strengthen the expecting mother’s coping skills. The phone coaches are specially trained healthcare professionals who support and motivate the women. The weekly phone sessions are scheduled when they are most convenient for the women. To ensure the fidelity and quality of the intervention, the phone coaches receive weekly supervision by two experienced CBT therapists, including case consultations, intervention management and support on the use of CBT principles during the coaching process.
Table 2. Core components of the Stronger Together iCBT intervention
Educational control group
The control group receives access to digitally delivered psychoeducational material, which includes information about pregnancy-related life changes, social relationships, well-being and healthy habits such as regular exercise, sleep and coping with stress. The controls do not receive weekly phone coaching or access to the core CBT components. The controls receive standard prenatal care at the maternity clinics.
Adverse events
Stronger Together is a minimal risk trial based on CBT, which is considered the first-line treatment for depression. CBT has no significant adverse effects and we do not anticipate that the intervention will have negative side effects. Any adverse effects that do occur are handled during the weekly phone calls and the women can also contact the study team at any time.
Primary outcome
The primary outcome variable is any change in the total EPDS score (Cox et al., 1987) from baseline at 13-18 weeks of gestation to 11 weeks after randomization. The EPDS is a standard 10-item self-report questionnaire, which was originally developed to screen for postpartum depression. It has been widely validated to detect depression both during (Cronbach’s α 0.82 – 0.84) (Bergink et al., 2011) and after pregnancy (Cronbach’s α 0.87) (Cox et al., 1987). However, it has not been validated specifically in Finland. The respondents are asked about symptoms of depression in the past week. The score ranges from 0 to 30 and 13 points or above is normally used as a cutoff for depression during pregnancy. Because this study focuses on early intervention, we use a cutoff of 10 points on the total EPDS score to ensure that we include pregnant women with subclinical depressive symptoms (Khanlari et al., 2019; Levis et al., 2020).
Secondary outcomes
Depression. The Beck Depression Inventory-II (BDI-II) (Beck et al., 1996) is also used to examine depressive symptoms. It is a widely used 21-item self-report instrument measuring the severity of depression with high validity and reliability: Cronbach’s α 0.88 when used during pregnancy and 0.89 among postpartum women (Bos et al., 2009). In Finland, the scale has been validated for adult populations (Seppänen et al., 2022). Each item is scored from 0 to 3 and the total score, which we will evaluate, ranges from 0 to 63. The BDI can be used to assess the nine symptoms of depression included in the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (American Psychiatric Association, 2013).
General anxiety. The total score of the General Anxiety Disorder 7-Item Scale (GAD-7) is used to assess changes in anxiety symptoms from baseline to follow-up. GAD-7 is a brief screening measure for generalized anxiety (Spitzer et al., 2006) with good reliability (Cronbach’s α=0.89). It includes seven questions about anxiety symptoms in the last two weeks and the frequency is rated from not at all to nearly every day. A total score of nine indicates potentially clinically significant anxiety symptoms. The GAD-7 has been validated for pregnant women (Zhong et al., 2015) and for use in Finnish primary care settings (Kujanpää et al., 2014).
Pregnancy-related anxiety. Changes in anxiety symptoms related to pregnancy are assessed using the second revision of the Pregnancy-Related Anxiety Questionnaire (PRAQ-R2) (Huizink et al., 2016). This version can be administered to pregnant women regardless of parity. The scale has been implemented and validated in Finland, and shown sufficient reliability (Cronbach’s alpha 0.71–0.85 depending on parity and pregnancy week) (Huizink et al., 2016). It consists of 10 items with three subscales: fear of giving birth, worries about delivering a physically or mentally handicapped child and concerns about the woman’s appearance. We will evaluate the total score and the three subscales separately.
Social anxiety. Changes in social anxiety symptoms are assessed using the total score of the Social Phobia Inventory (SPIN), which contains 17 items about avoidance and fear of embarrassment. Each item is rated on a scale of 0 to 4, with higher scores indicating greater distress. A result of 19 points or above on the SPIN is considered indicative for social phobia. The SPIN has demonstrated good test-retest reliability, internal consistency (Cronbach’s alpha 0.82–0.94) and validity (Connor et al., 2000). In Finland, it has been validated for adolescents (Ranta et al., 2007), but not pregnant women.
Stress. The Perceived Stress Scale (PSS) (Cohen et al., 1983) examines changes in personal stress levels. The 10-item scale covers stressful feelings and thoughts during the last month, ranging from 0 for never to 4 for very often. The total score, which we will evaluate, ranges from 0 to 40, with a higher score indicating greater perceived stress. The Cronbach’s α for this instrument is between 0.84 and 0.86 and it has been validated for pregnant women (Solivan et al., 2015). In Finland, it has been validated for measuring stress symptoms of undergraduate students (Bhurtun et al., 2021).
Maternal attachment. The total score of a shorter 12-item version of the Maternal Antenatal Attachment Scale (MAAS) (Condon, 1993) is used to evaluate antenatal attachment. This scale focuses on maternal attitudes, thoughts, feelings and behaviors towards the growing fetus and higher scores indicate a more adaptive mother-fetal bonding style. This scale has an internal consistency of 0.73 (Navarro-Aresti et al., 2016) and has not been validated in Finland.
Sleep and quality of life. Single questions regarding quality of life, loneliness, satisfaction with social relationships and sleep are administered. Each question is analyzed separately.
Background measures
Sociodemographic and health-related data including age, family structure, level of education, employment and chronic diseases are collected at baseline. This includes detailed information about previous psychiatric diagnoses and their treatment, smoking, substance abuse and previous pregnancies and childbirth. Further, the women are asked about meaningful life events twelve months before the beginning of the study.
Treatment-related measures
Treatment-related measures are only administered to the intervention group.
Therapeutic alliance. The Working Alliance Inventory - Short Revised (WAI-SR) is used after the iCBT treatment. This instrument measures the therapeutic and collaborative relationship between the coach and the patient. It has shown good psychometric properties specifically for guided internet interventions (Cronbach’s α: 0.93) (Gómez Penedo et al., 2020).
The use of the digital content. The digital platform automatically collects user data on how long the woman spends on the website, which material she uses and how often she reads it. The frequencies are daily/almost daily, a couple of times a week/once a week or never. We also examine the proportion of completed coaching calls, the time interval between calls and the call durations. The woman’s location online and during coaching calls is also recorded, namely at home, at work or in another place.
Treatment satisfaction. Women who receive the intervention are asked about the general usability and perceived benefit of the intervention after they finish the treatment. In addition, the structured Client Satisfaction Questionnaire for internet-based interventions (CSQ-I) is used. This tool has shown good psychometric qualities and reliability (McDonald’s Ω 0.93–0.95) for assessing general user satisfaction with Web-based interventions for depression (Boß et al., 2016).
Table 3. Instruments and assessment points for self-rating questionnaires
Randomization and masking
Randomization is stratified by the study site, and it takes place after the baseline assessment and MINI interview are completed. The study statistician creates the randomization codes separately for each study site using the letters A and B for the intervention and control groups. The randomization sequences are generated with a 1:1 ratio using the SAS 9.4 computerized random permuted block sequence generator (SAS Institute, Cary, NC, USA) with concealed block sizes. The statistician stays blinded to the treatment groups until the randomization expert opens the coding after the blind review. After each subject is randomized, the platform unlocks the appropriate user interface and the subjects receive an email telling them what group they are in. The statistician and randomization expert are part of the study team, but they are not directly involved in the conduct of the study.
Data management
All the data gathered by the digital platform are stored in a PostgreSQ database (The PostgreSQL Global Development Group). After the collection period has been completed, the data manager will ensure the accuracy of the data and then import the SAS datasets for statistical analysis and reporting. Metadata are stored in spreadsheets and a data dictionary provides guiding principles for the dataset and variable construction, as well as study-specific information. All datasets from different data sources will be imported to SAS with programs that follow the guidelines in the data dictionary. Code books will be created to document the contents of the SAS datasets: variable names, labels, types and formats. Quality control procedures such as database auditing, data reviewing and reconciliation will be implemented to ensure that the statistical analyses use acceptable quality data.
All the baseline questionnaires are filled in online, and participants have to complete all the required fields. The platform does not allow them to select multiple responses to Likert-type questions. During the follow-up phase, participants are given the choice to complete the questionnaires online or on paper versions, which are manually digitalized using Microsoft Access (Microsoft Corp, Redmond, WA, USA). To ensure consistency, the online templates follow a similar layout to the paper questionnaires. Personnel digitalizing these follow guidelines on how to handle unclear answers, such as multiple answers selected. The collected data are stored on the University of Turku fileservers and access is limited to appointed researchers, statisticians and the data manager. The files are automatically backed up each day and kept for 26 weeks.
Power
The aim of the power calculations was to determine how many subjects should be enrolled in the intervention and control groups to achieve 80% power to reject the null hypothesis of equal means. The power calculations were based on a population mean difference of μ1 - μ2 = 9,0 – 10.5 = -1.5 in EPDS scores, with a standard deviation of 6.30 for both groups and a significance level of alpha equals 0.05. This resulted in 278 subjects in both groups. Because we anticipate 30% attrition in both groups, we aim to enroll 400 dyads in both groups.
Analysis plan
Statistical analyses will be carried out on the intention-to-treat dataset and repeated for the per-protocol dataset that consists of all evaluable subjects. Categorical variables will be presented as numbers and percentages and continuous variables as means and standard deviations. Pearson’s chi-square tests or Fisher’s exact tests will be conducted to explore differences in categorical variables between the intervention and control groups at baseline. Two-tailed, 2-sample t tests will be used to explore differences in continuous variables between the groups at baseline. The primary outcome measure is the total EPDS score, which is collected at baseline and 11 weeks after randomization. The follow-up EPDS total scores and changes in EPDS scores will be analyzed using linear mixed-effects model where the baseline EPDS total score and the stratifying factor study site are included as covariates. The follow-up EPDS total scores and changes in EPDS scores will be analyzed using analysis of covariance (ANCOVA), adjusted for the baseline EPDS total score and stratifying factor study site as a covariate. The secondary outcome measures will be analyzed using the same modelling approach as the primary outcome measure. If feasible, the study subjects will be included as a random effect, to generalize the results beyond our study sample. The primary outcome will be analyzed blinded, as the treatment code will only be opened after the study statistician has analyzed the primary outcome. Two-sided significance levels of 0.05 and 95% confidence intervals will be used in statistical testing. The statistical analyses will be carried out with SAS statistical software (SAS 9.4, SAS Institute, Cary, NC USA).
Ethics
This protocol conforms to the Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) statement (Chan et al., 2013). The intervention is described in accordance with the Template for Intervention Description and Replication (TIDieR) checklist (Hoffmann et al., 2014). We have registered the study with ClinicalTrials.gov (NCT04223115) and the study has been approved by the research ethics board of the wellbeing county of Southwest Finland (115/1801/2019). The relevant bodies have, and will, be informed of any important protocol modifications after the original registration and approval. Permission to conduct the study has been obtained from the local authorities in charge of maternity health care.
Participation is voluntary and refusing to participate will not affect the women’s use of other services. Both written and oral information about the study is given to the women during the screening phase and before the intervention. All collected data are strictly confidential.
The Research Centre for Child Psychiatry at the University of Turku has initiated and planned the study and is responsible for its management. The funders have played no role in planning the study. The anonymized results will be published in scientific journals and disseminated to healthcare professionals and the general public. The study participants will not receive any payments or reimbursements for participating.