The Reproductive Mental health of Ontario Virtual Intervention Network (MOVIN): a pilot randomized controlled trial

Abstract

Background

The Reproductive Mental health of Ontario Virtual Intervention Network (MOVIN) aims to improve perinatal depression care across a large Canadian health jurisdiction. It involves a web-based platform, care coordinator for personalized treatment planning, and psychiatrist to liaise with primary care clinicians, and provide direct consultation as needed. This was a pilot randomized controlled trial (RCT) of MOVIN.

Methods

Participants aged ≥18 years, pregnant or within 12 months postpartum and with Edinburgh Postnatal Depression Scale (EPDS) score >12 recruited from across Ontario, Canada, were randomized 1:1 to MOVIN for 24 weeks or a control condition. The primary outcome was feasibility, inclusive of recruitment, acceptability, and research protocol follow-up. Depression symptom remission was a main secondary outcome for the pilot trial. EPDS scores and remission (EPDS ≤ 12) were compared between groups.

Results

Of 101 participants (n = 48 MOVIN; n = 53 control), 80% completed 24-week follow-up. Participant views of MOVIN were very positive (high acceptability) and multiple opportunities for protocol adjustment in a larger future study were identified. At 24 weeks post-randomization, EPDS scores were lower in MOVIN vs. controls, with a mean difference adjusted for baseline score of −2.32 (95% Confidence Interval, CI −4.23 to −0.42). At the same time point, 75.0% of the MOVIN group vs. 51.1% of controls were in remission (chi-square=4.83, p = 0.03).

Conclusions

With high feasibility, including in recruitment, acceptability, and research protocol adherence, and preliminary suggestion of efficacy, the results of this study support proceeding to a large-scale RCT of MOVIN to definitively evaluate its efficacy at a larger scale.

Plain language summary

Depression affects up to 20% of people during and after pregnancy, but as few as 1 in 5 are treated to remission. In the Reproductive Mental health of Ontario Virtual Intervention Network (MOVIN), participants can view educational resources about depression during and after pregnancy, and receive personalized evidence-based treatment planning with a care coordinator with oversight by a psychiatrist. Automated symptom check-ins help identify when to iterate treatment plans over time. This pilot randomized controlled trial (RCT) showed that MOVIN was feasible to implement and highly acceptable to participants. After 6 months, more MOVIN participants were in remission of their depression (~75%) than were participants randomized to a usual care control condition (~51%). These results support proceeding to a larger trial to definitely evaluate the MOVIN program.

Vigod et al. present a pilot randomized trial of a virtual collaborative mental healthcare model for perinatal depression that aims to optimize care in a large Canadian province. They find that the intervention is feasible to implement and highly acceptable to participants, with preliminary evidence suggesting it is superior to a control condition at improving depressive symptoms.

Introduction

Depression is common around the time of pregnancy, affecting 15–20% of childbearing people during the perinatal period^1,2. It contributes to poor quality of life for the affected person and is linked to problems with child development^3–5. Guidelines recommend that treatment occur within a stepped-care approach, where intensity of intervention matches the level of need⁶. For mild symptoms, psychosocial interventions such as peer support, support groups and self-guided tools may be sufficient. More significant symptoms may require formal psychological treatments (e.g., cognitive behaviour therapy), medications, or, in more rare cases, other higher-intensity medical treatments (e.g., electroconvulsive therapy).

Perinatal depression is undertreated, with key barriers to care that include shame, stigma, logistics such as transportation, childcare, and finances, and lack of specialized care in certain regions⁷. Virtual care can address some of these barriers. There is robust evidence for remote delivery of interventions such as peer support, internet-based guided self-help, and psychological therapy^6,8. Yet, minimal research has focused on how to leverage technology to optimize access to treatment within a stepped model of care delivery. Online platforms have been created to provide patient-facing options for planning treatment and accessing the right type of care; however, few of these platforms have been rigorously evaluated⁹.

Collaborative care is a patient-centered healthcare model based on the principles of chronic disease management where two or more healthcare providers, including those from primary and specialist care, share care plans that are coordinated via a trained behavioural healthcare coordinator¹⁰. Collaborative mental health care—involving a multi-professional approach, structured evidence-based assignment of interventions, and systematic follow-up—has consistently been shown to be superior to usual care for depressive and anxiety symptoms in non-perinatal populations¹¹. Some perinatal collaborative care models have been trialed, usually involving a behavioural care coordinator embedded in an obstetrical care setting, sometimes with psychological treatment available¹². However, these have focused mainly on pregnancy, since obstetrical care tends to stop at ~6 weeks postpartum, and not all trials have had positive results. A recent systematic review suggested this may be in part because the key elements of collaborative care have been inconsistently applied across interventions¹².

In Ontario, Canada’s largest Province, there is currently no patient-facing system-level mechanism that coordinates perinatal mental health care. Based on the promise of the collaborative mental health care model and the need for more research in this area, we developed the Reproductive Mental health of Ontario Virtual Intervention Network (MOVIN) as a Province-wide intervention to improve the perinatal mental health of our population. MOVIN leverages technology to reach pregnant and postpartum people outside of specific clinic visits or obstetrical or primary care practices. Patients can self-refer on MOVIN’s web-based platform and view psychoeducational resources there. They can then meet virtually with a care coordinator to create a personalized treatment plan that includes existing virtual and in-person resources in their communities. A psychiatrist oversees treatment planning, communicates with and provides advice to the patient’s primary care clinician, and can provide direct psychiatric consultation as needed. Automated symptom checks help to identify when to iterate treatment plans over time. In this way, MOVIN applies the key components of successful collaborative mental health care.

This pilot randomized controlled trial (RCT) of MOVIN aimed to evaluate the feasibility of proceeding to a large-scale RCT. The results show that the MOVIN protocol is feasible to implement, including with respect to recruitment and research questionnaire follow-up, and high intervention acceptability on the part of the participants. Although pilot trial efficacy results are not meant to be definitive, it is also highly encouraging that those randomized to the MOVIN group have lower levels of depression symptoms at 24-weeks post-randomization compared to participants allocated to a usual care control condition.

Methods

Study design

This pilot randomized, controlled, parallel-group trial randomized women and gender-diverse persons with perinatal depression from across Ontario, Canada’s largest Province (pop. ~15 million) to MOVIN or a control condition (1:1 ratio). Outcome data were collected at 12- and 24-weeks post-randomization. The Research Ethics Board at Women’s College Hospital (WCH) in Toronto, Ontario (REB#2020-0054-B) approved the trial. Funding for the RCT was from the WCH Foundation, which had no role in the study design, collection, analysis or interpretation of data, or the decision to submit the results for publication. The trial was registered at clinicaltrials.gov (NCT04680585) first posted 2020/12/23. Recruitment started 2021/04/12.

Participants

Inclusion criteria were: (1) adult 18 years or older, (2) pregnant or up to 1 year postpartum (including through adoption or surrogacy, and inclusive of cis-women, non-binary and transgender people), (3) Edinburgh Postnatal Depression Scale (EPDS) score >12, which has high specificity for depression (0.95)¹³ (clinical diagnosis was not required, so as to best approximate real-world implementation); and (4) registered with a primary care provider (family physician/nurse practitioner) in Ontario (in Canada health services are organized, licensed and funded provincially). Requiring a primary care provider was necessary, to be able to implement the collaborative mental health care model. While obstetrical care in Canada often includes the screening and identification of depression and anxiety in pregnancy and at the postpartum visit (~6 weeks postpartum), it does not routinely include ongoing mental health management or prescribing of psychotropic medications. These activities are primarily in the scope of the primary care provider.

Exclusion criteria were active suicidal ideation, mania, psychosis, substance or alcohol use disorder, inability to complete relevant study procedures or measures online (expected to be a small minority: 98.2%–93.4% for low-income—of Canadian households have access to home internet in ages 25–44 years¹⁴, although there was budget for devices and data plans if necessary), or inability to participate in English. Neither the MOVIN intervention components nor the research questionnaires were translated due to budgetary constraints for the pilot.

Participants were recruited through investigator and professional networks (e.g., public health nurse listserv, obstetrical and midwifery clinics, delivery hospitals across Ontario), digital advertisements (e.g., Facebook), and social media (e.g., Twitter, baby/mommy blogs). Recruitment materials directed participants to a study landing page on the institutional website where participants could submit their interest in the study to the research team. This recruitment plan reflected anticipated real-world implementation where anyone in the health jurisdiction could self-identify online without going through a healthcare provider (a known barrier to case identification), but healthcare providers could also direct their patients there. A highly trained research team member conducted consent and eligibility procedures by telephone, with principal investigator (or delegate) confirmation. Informed consent was obtained prior to the initiation of any study procedures.

The secure encrypted web-based clinical trial data management system REDCap was used to conduct randomization and manage data collection. REDCap is fully configurable and incorporates sophisticated data validation rules to ensure high-quality data is captured for remote web-based data entered by participants and research staff. Data were hosted, managed and stored in a REDCap database at Women’s College Hospital in Toronto, Ontario.

Interventions

The experimental condition was the MOVIN intervention for 24 weeks post-randomization. Based on collaborative mental health care models¹⁰, the intervention involves a care coordinator and psychiatrist who maintain a registry of patients for whom they provide evidence-based treatment recommendations in a stepped-care approach. They support the patient’s primary care clinician and provide direct specialist care when needed. The premise is that clear workflow procedures and personnel to facilitate straightforward access to treatment based on a patient’s level of clinical severity and treatment preferences may result in more timely and effective treatment, leading to improved clinical outcomes¹⁵.

Each participant randomized to MOVIN was invited to create a secure account on the MOVIN platform, accessible via web- and mobile-enabled devices and cloud-hosted with built-in privacy and security protocols for healthcare data backed by real-time auditable privacy policies. The platform contained links to publicly available educational materials and treatment services in Ontario, curated by the study team with input from persons with lived experience. This included links to resources for specific communities, including persons of colour, gender and sexual minorities, and those with a disability, and links to multi-lingual resources focusing on the top 5 languages other than English most commonly spoken in Ontario (French, Spanish, Punjabi, Arabic, Chinese).

When participants logged into the platform for the first time, they were automatically prompted to fill out a scale (using the ten questions from the EPDS) to evaluate their level of symptoms. They received automated feedback about the meaning of their score and educational resources were suggested for them based on reported symptoms. For example, if a participant reported difficulty with sleep, then sleep resources would be flagged for them. Participants could also optionally create a profile to enter information about their psychiatric history and treatment preferences.

After completing the symptom scale +/ their profile, participants were prompted to schedule a telephone call or secure video visit with a MOVIN care coordinator to collaboratively create a treatment plan. The role profile for the care coordinator was based on that of the behavioural healthcare manager in collaborative mental healthcare, including: (1) treatment planning and coordinating treatment, (2) proactively following up to gauge treatment response, (3) organizing communications between primary care provider and psychiatrist consultants as needed, and (4) offering brief counseling using evidence-based techniques¹⁶. This role profile can be applied to clinical social workers, licensed counsellors, nurses, psychologists and others. In the pilot study, there was one MOVIN care coordinator, a registered nurse (RN) with expertise in perinatal health and training in evidence-based counselling techniques.

Treatment planning followed a stepped-care framework based on illness severity (Table S1). Cut-off scores on the EPDS for varying levels of symptom severity were used as a guide (e.g., EPDS < 13 representing milder symptom levels, 13–18 representing mild to moderately severe symptoms and 19 or greater representing moderate to severe symptoms)¹⁷. Recommended interventions comprised existing Ontario-based services accessible at no cost to the participant, pre-specified in the study’s standard operating protocols according to level of severity. However, recommendations were individualized and took into account patient preference and prior experience. The care coordinator also had the flexibility to find and recommend additional services to meet participant needs, outside of the services outlined in the standard operating protocol, consulting with the psychiatrist as needed.

Psychosocial interventions (e.g., peer support, public health nursing support) were recommended regardless of symptom severity, and on their own for people with milder symptomatology. An example of peer support included as an option in the study’s protocols was Postpartum Support International, which provides a telephone helpline, online support group, chat rooms and discussion boards (peer support), including across Canada at no cost to the user (www.psi.org). Ontario also has public health resources in each region to support perinatal people, which includes nurse home visiting programs, online and in-person nurse-led support groups (Ontario’s Healthy Babies Healthy Children Program). This was the standard recommendation for public health nursing support.

Formal psychological interventions (e.g., cognitive behaviour therapy) were generally recommended for those with moderate or severe symptoms (or for those who had tried psychosocial interventions already to minimal effect)^18,19. At the time of the pilot RCT, there were several publicly-funded psychological services available at no cost to the end-user, including internet-based cognitive behaviour therapy (iCBT) programs covered by the Province, as well as online group therapy offered through one of Ontario’s specialty perinatal mental health care programs.

For participants with more severe symptoms (e.g., EPDS 19 or above) or for those who had had inadequate response to a psychological intervention, the care coordinator would usually encourage the participant to book a visit with their primary care clinician for further assessment and consideration of medication treatment. In these cases, a letter was sent to the primary care clinician detailing the current symptoms and reasons why a visit was recommended (e.g., to check for reversible conditions such as hypothyroidism, to consider adding an antidepressant medication to the participant’s treatment plan). The program also facilitated MOVIN psychiatrist-to-primary care clinician e-consultation and direct patient video-consultations with the MOVIN psychiatrist (no waiting list) if additional advice was needed. As MOVIN was not set up as a crisis service, if significant safety issues arose (e.g., active suicidal ideation), the participant was directed to their nearest emergency department.

Once the participant and care coordinator agreed on a plan, details about each recommended service and how to access it were placed in a “Care Plan” in the MOVIN platform for the participant to view as needed. A letter was also sent to the primary care clinician reporting on what was recommended. Participants could also request that a copy be sent to their obstetrical care provider (e.g., midwife, obstetrician) if still being followed in that setting.

After initial treatment recommendations, all participants were monitored every 2–4 weeks, at the discretion of the care coordinator (i.e., every 2 weeks when a person is most symptomatic; every 4 weeks would be more common if symptoms were mild initially and/or once the participant started to improve). For monitoring, participants received an email prompting them to login to their account in the MOVIN platform, where they were asked if they thought that their symptoms were improving, worsening, or staying the same, and then offered to complete the EPDS again. Participants whose symptoms were worsening/not improving were prompted to reconnect with the care coordinator to adjust treatment (e.g., help address barriers to access of a recommended service, move to a higher care step). Communications were sent to the primary care clinician for updates, or if additional actions were recommended (e.g., as per above in terms of recommendation to consider medication, or direct psychiatrist consultation).

Symptoms were monitored until 2 consecutive EPDS scores were <10, whereupon the care coordinator would arrange for discharge from MOVIN and offer the option to reconnect if symptoms recurred prior to the end of the study period. For those with EPDS > 10 at the end of the study period, updates to primary care clinicians were provided with recommendations for future care.

Those in the control condition were given a PDF hand-out describing what treatment options might be helpful based on their initial EPDS score that had been collected through the trial’s baseline research assessment. The handout comprised a list of resources about depression and anxiety in pregnancy and postpartum and an up-to-date listing of treatment services available in Ontario (Table S2). Participants in the control condition did not interact further with the research team, except for completion of their research measures by self-report. In their informed consent, all participants were informed that responses to electronic research survey questionnaires would not be monitored in real-time. Participants were provided with an up-to-date list of national and province-specific mental health resources with websites and phone numbers for crisis lines in the case of need for immediate assistance.

Research measures

Prior to randomization, participants were asked to complete a baseline questionnaire via REDCap, querying demographic, obstetrical and mental health history variables, baseline measures of clinical depressive and anxiety symptoms, quality of life, relationship function and parenting stress. Depressive symptoms, expected to the primary outcome of a future RCT, were measured using the EPDS. Anxiety symptoms were measured using the Generalized Anxiety Disorders-7 (GAD-7) scale, a self-report scale with good discriminate validity in perinatal populations²⁰. Quality of life was measured using the EQ-5D-5L, a multi-attribute utility instrument for measuring quality-adjusted life year^18,21. The dyadic consensus subscale of the Dyadic Adjustment Scale (DAS), a self-report measure of relationship adjustment with an intimate partner, was used to measure relationship distress (among those with a partner only)¹⁹. The Parenting Stress Index short form (PSI-SF), a 36-item measure consisting of three sub-scales: parental distress, dysfunction in the parent-child relations and difficult child to measure parenting stress was administered at baseline to postpartum participants only²². A MINI International Neuropsychiatric Interview was telephone-administered by highly trained research staff to evaluate for past and current psychiatric diagnoses according to the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-TR) criteria.

The primary outcome was feasibility, to determine how well the protocol could be implemented and inform a future large-scale evaluation. Feasibility metrics were recruitment rate, adherence to follow-up research measures, and, for the MOVIN group, acceptability/adoption of the intervention. Participants completed an acceptability questionnaire based on the Technology Acceptance Model (TAM) that evaluates perceived usefulness, perceived ease of use, social norms, attitudes toward use and behavioral intention for use²³. TAM-based acceptance questionnaires have been used in the evaluation of mental health interventions^24,25. A second questionnaire developed for the study queried ease of access, perceptions of layout, presentation and content, and interactions with the care coordinator. We also measured the proportion of participants who logged in, connected with the care coordinator, completed their patient profile, and completed follow-up monitoring in the platform. Success criteria for proceeding to a large RCT were: (1) n ≥ 60 recruited, (2) ≥75% retention, and (3) ≥80% acceptability (MOVIN group).

Secondary outcomes were clinical symptoms (EPDS, GAD-7), quality of life (EQ-5D-5L), relationship distress (DAS), parenting stress (PSI-SF, postpartum only), and health service use (within and outside of MOVIN) at 12- and 24-weeks post-randomization. Adverse events were assessed systematically in questionnaires querying obstetrical, neonatal and birth outcomes at both follow-up time points, and could also be spontaneously reported by participants.

Sample size

A range of 20–40 participants in a pilot study allows for sufficient variability in assessing the process of an intervention²⁶. We aimed to randomize 100 participants (50/arm), conservatively estimating a 75% retention rate, which would result in 37 participants/arm at 24 weeks post-randomization. Initially, the trial was registered to recruit 80 participants; however, this was amended to 100 participants at about 6 months into recruitment (September 20, 2021) as recruitment was proceeding rapidly, and recruiting closer to the upper range of participants (i.e., outcomes for as close to 40 participants as possible) would be optimal to understand feasibility.

Randomization

Randomization was 1:1 in randomly varying block sizes using a computer-generated random allocation sequence, stratified by pregnancy status (pregnant or postpartum) at enrolment. A research team member informed participants of their allocation status and provided access to the MOVIN or control condition. It was not possible for participants to be blinded, but they were not informed of the study’s explicit hypotheses.

Statistical analyses

Proportions and means were calculated to determine recruitment and research questionnaire follow-up rates, metrics of adoption, and Likert-scale responses from the acceptability questionnaires. We conducted intention-to-treat analyses for clinical outcomes, which were analyzed over time by intervention group using univariate linear mixed models (LMM) adjusted for fixed effects (time of response and interaction between time and intervention condition) and random effects (subject). This model handles continuous longitudinal data in an easy, valid, and flexible manner and can be used for data with an unequal number of measurements per subject²⁷. The proportion of individuals with EPDS ≤ 12 at each endpoint was also compared between groups using a chi-square test of proportions.

Results

Recruitment was from April to November 2021, with follow-up to June 2022. From 293 self-referrals to the study website, 186 underwent full eligibility screening, of whom 147 were eligible and 101 were randomized (n = 48 MOVIN, n = 53 control) (Fig. 1). No potential participants required study-provided devices or data plans. Follow-up questionnaires were completed by 94% at 12 weeks and 80% at 24 weeks post-randomization. There were two withdrawals in the control condition; all others who did not complete questionnaires were lost to follow-up.

Fig. 1. Consort diagram for the pilot randomized controlled trial of the Reproductive Mental health of Ontario Virtual Intervention Network (MOVIN).

EPDS Edinburgh Postnatal Depression Scale.

Most participants (mean age 34.1, SD 4.15; 47.5% pregnant) were married or common law (96.0%), heterosexual (86.9%, 11.1% bisexual), had post-secondary education (86.1%) and had an annual household income of >$80,000 CAD/year (79.2%). About 12.9% of participants were born outside of Canada, and for 28.7%, English was not the primary language spoken in the home. Participants were from all across Ontario, with 52.4% living in Toronto, Ontario’s largest city (where ~44.4% of Ontarians live)²⁸. Sociodemographic and psychiatric history characteristics were similar between groups (Table 1).

Table 1.

Baseline characteristics, presented as N (%) unless otherwise specified

	MOVIN (n = 48)	Control (n = 53)
Socio-demographics
Mean age (SD) in years	34.2 ± 4.5	34.0 ± 3.7
Living with a partner	46 (95.8)	46 (86.8)
Heterosexual	39 (84.8)	47 (88.7)
Pregnancy status at baseline
Pregnant	22 (45.8)	22 (41.5)
Postpartum	26 (54.2)	31 (58.5)
Index pregnancy is first pregnancy	10 (20.8)	20 (37.7)
Completed postsecondary education	41 (85.4)	46 (86.8)
Household Income <$80,000 CAD/annum	7 (14.6)	11 (20.8)
Born in Canada	42 (89.4)	46 (86.8)
English as first language	35 (72.9)	37 (69.8)
Self-identified race or ethnicity (self-identified)
Asian (Chinese, Filipino)	3 (4.2)	3 (5.7)
Black (Caribbean, North American)	1 (2.1)	2 (3.8)
Indigenous	0	1 (1.9)
Latin-American	2 (4.2)	2 (3.8)
South/Southeast-Asian	1 (2.1)	5 (9.4)
West Asian	1 (2.1)	1 (1.9)
White only	39 (81.3)	36 (67.9)
Prefer not to answer	1 (2.1)	3 (5.7)
Psychiatric history
Lifetime psychotropic medication	28 (58.3)	33 (62.3)
Lifetime psychiatric hospitalization	1 (2.1)	6 (11.3)
Lifetime suicide attempt	4 (8.3)	5 (9.4)
Current episode MINI diagnosis
Generalized anxiety disorder	20 (41.7)	20 (39.2)
Major depressive disorder	11(22.9)	10 (19.2)
Obsessive-compulsive disorder	0 (0.0)	4 (7.7)
Panic disorder	2 (4.2)	2 (3.8)
Post-traumatic stress disorder	2 (4.2)	3 (5.8)
Social anxiety disorder	4 (8.3)	3 (5.9)

Participant views of MOVIN were positive, with over 80% of respondents agreeing that MOVIN enhanced the management of their mental health concerns and that they intend to use MOVIN in future (Table 2). Views of the layout and content were also positive (Table 3). The care coordinator role was rated very highly, with ~90% of respondents reporting that the care coordinator was easy to contact, able to understand their situation and comfortable to communicate with, and provided useful help. About 75% felt they had easier access to healthcare services because of the care coordinator, that they would not have received this type of treatment without MOVIN, and that the treatments recommended were easy to access. However, participants were not as sure that their overall access to healthcare services was better because of MOVIN (40.5% neutral, 31.0% agreed). Open-ended comments showed occasional errors with login and authentication, a desire for improved optimization on mobile, and constructive suggestions for navigating the resource section (Table S3).

Table 2.

Patient acceptance questionnaire adapted from Yan and Or²³, based on the technology acceptance model (n = 41)

	Very strongly disagree	Strongly disagree	Disagree	Neutral	Agree	Strongly agree	Very strongly agree
MOVIN improves the ability to manage mental health concerns	1 (2.4)	0 (0.0)	0 (0.0)	8 (19.5)	18 (43.9)	6 (14.6)	8 (19.5)
Saves time managing mental health concerns.	1 (2.4)	0 (0.0)	0 (0.0)	11 (26.8)	13 (31.7)	11 (26.8)	5 (12.2)
Enhances managing mental health concerns	1 (2.4)	0 (0.0)	0 (0.0)	7 (17.1)	18 (43.9)	10 (24.4)	5 (12.2)
MOVIN is useful in managing mental health concerns	1 (2.4)	0 (0.0)	0 (0.0)	11 (26.8)	15 (36.6)	9 (22.0)	5 (12.2)
I find it easy to get MOVIN to do what I want it to do	0 (0.0)	0 (0.0)	3 (7.3)	12 (29.3)	15 (36.6)	9 (22.0)	2 (4.9)
It is easy for me to become skillful at using MOVIN	0 (0.0)	0 (0.0)	2 (4.9)	11 (26.8)	16 (39.0)	10 (24.4)	2 (4.9)
I find MOVIN easy to use	0 (0.0)	0 (0.0)	1 (2.4)	10 (24.4)	16 (39.0)	9 (22.0)	5 (12.2)
Using MOVIN is a good idea	0 (0.0)	0 (0.0)	0 (0.0)	6 (14.6)	15 (36.6)	11 (26.8)	9 (22.0)
Using MOVIN is a wise idea	0 (0.0)	0 (0.0)	0 (0.0)	6 (14.6)	14 (34.1)	12 (29.3)	9 (22.0)
I like the idea of using MOVIN	0 (0.0)	0 (0.0)	0 (0.0)	3 (7.3)	17 (41.5)	11 (26.8)	10 (24.4)
Using MOVIN is pleasant	0 (0.0)	0 (0.0)	1 (2.4)	11 (26.8)	13 (31.7)	9 (22.0)	7 (17.1)
People who influence my behaviour would think I should use MOVIN	0 (0.0)	0 (0.0)	0 (0.0)	16 (39.0)	13 (31.7)	6 (14.6)	6 (14.6)
People who are important to me would think I should use MOVIN	0 (0.0)	0 (0.0)	0 (0.0)	11 (26.8)	16 (39.0)	7 (17.1)	7 (17.1)
I intend to use MOVIN.	0 (0.0)	0 (0.0)	3 (7.3)	5 (12.2)	18 (43.9)	8 (19.5)	7 (17.1)
I predict I will use MOVIN	0 (0.0)	0 (0.0)	3 (7.3)	5 (12.2)	19 (46.3)	7 (17.1)	7 (17.1)
Plan to use MOVIN	0 (0.0)	0 (0.0)	3 (7.3)	6 (14.6)	15 (36.6)	9 (22.0)	8 (19.5)

Results are presented for the 24-week follow-up time point for n = 34 and the 12-week time point for n = 7 who completed the 12-week but not the 24-week post-randomization follow-up. Results are presented as N(%) unless otherwise specified.

Table 3.

Participant perceptions of the MOVIN platform and care model questionnaire designed for the study (n = 44)

	Strongly disagree	Disagree	Neutral	Agree	Strongly agree
MOVIN layout and presentation
MOVIN was easy to use and navigate	0 (0.0)	5 (11.4)	7 (15.9)	20 (45.5)	12 (27.3)
The information was laid out in a logical and organized manner	0 (0.0)	3 (6.8)	8 (18.2)	20 (45.5)	13 (29.5)
The colours and fonts made MOVIN easy to read	0 (0.0)	1 (2.3)	9 (20.5)	21 (47.7)	13 (29.5)
I liked the look and visual presentation of MOVIN	0 (0.0)	2 (4.5)	14 (31.8)	15 (34.1)	13 (29.5)
MOVIN sections and content
The screening was useful	0 (0.0)	0 (0.0)	7 (15.9)	29 (65.9)	8 (18.2)
The repeat screening option was useful	0 (0.0)	2 (4.5)	8 (18.2)	27 (61.4)	7 (15.9)
The patient profile section was useful	0 (0.0)	0 (0.0)	14 (31.8)	24 (54.5)	6 (13.6)
The resources and educational materials were helpful	0 (0.0)	2 (4.5)	6 (13.6)	24 (54.5)	12 (27.3)
The resources and educational materials were easy to understand	0 (0.0)	0 (0.0)	4 (9.1)	25 (56.8)	15 (34.1)
The resources and educational materials were free from judgement or bias	0 (0.0)	0 (0.0)	3 (6.8)	27 (61.4)	14 (31.8)
MOVIN care coordinator
Easy to contact the care coordinator	1 (2.3)	0 (0.0)	2 (4.7)	16 (37.2)	24 (55.8)
Care coordinator was able to understand my situation	1 (2.3)	0 (0.0)	2 (4.7)	17 (16.8)	23 (53.5)
Comfortable communicating with the care coordinator	1 (2.3)	0 (0.0)	2 (4.7)	15 (34.9)	25 (58.1)
The Care Coordinator provided useful help	1 (2.3)	1 (2.3)	5 (11.6)	13 (30.2)	23 (53.5)
Care coordinator helped me feel better	1 (2.3)	0 (0.0)	8 (18.6)	14 (32.6)	20 (46.5)
MOVIN access to health services
Easier to access health care services because of the care coordinator	2 (4.7)	1 (2.3)	7 (16.3)	16 (37.2)	17 (39.5)
I would not have received this type of treatment without MOVIN	2 (4.7)	1 (2.3)	11 (25.6)	14 (32.6)	15 (34.9)
Treatment (s) suggested was/were easy to access	1 (2.3)	3 (7.0)	8 (18.6)	17 (39.5)	14 (32.6)
My access to health care services was better because of MOVIN	3 (7.1)	9 (21.4)	17 (40.5)	13 (31.0)	0 (0.0)

Results are presented for the 24-week follow-up time point for n = 36 individuals and the 12-week time point for n = 8 individuals who did not complete the 24-week timepoint. Results are presented as N(%) unless otherwise specified.

About 85.4% (n = 41) of MOVIN group participants logged in at least once, mostly by computer (n = 16), smartphone (n = 14) or both (n = 11, one of whom also used a tablet). All 41 completed the patient profile, six of whom required a reminder or assistance from the care coordinator to complete it. All participants completed follow-up symptom monitoring, which included repeated administration of the EPDS scale, at least once (66% completed it 7 or more times; maximum 12 times). Initial EPDS scores input into the platform by participants were in the mild range for 7, moderate range for 24 participants and severe range for 10 participants.

Almost all participants who logged in to the platform (n = 39/41, 95.1%) and most participants overall (n = 39/48, 81.3%) had a call with the care coordinator. This included 5/7 (71.4%) in the mild range on initial EPDS and all in the moderate (24/24, 100%) and severe (8/8, 100%) ranges on their initial platform EPDS. The median number of care coordinator meetings was 3 (IQR 2–6), so most participants had the opportunity to discuss stepped-care adjustments to their treatment plan.

The most frequently recommended intervention was the online group therapy program (n = 23), followed by the internet-based cognitive behaviour therapy option (n = 18). Most participants who met with the care coordinator (n = 37/39) received recommendations in addition to the pre-specified ones, mostly for unguided self-help (n = 33) and other psychological services (n = 34) (Table S1). Reasons for this included waitlists and participant preference for synchronous individual therapy or local in-person services. The care coordinator also recommended services for related issues (intimate partner violence, couples counselling, partner mental health, parenting, eating disorders, and cannabis use). The MOVIN psychiatrist provided primary care provider e-consultation once (2.1%) and direct consultation for three participants (6.3%). Direct consultations with the psychiatrist occurred when there was a need for physician input, including when participants’ primary care clinicians were unavailable and when there was high symptom severity over multiple follow-ups with minimal improvement. Participants with MOVIN psychiatrist involvement (directly or via e-consultation with the primary care provider) improved substantially, with mean (SD) EPDS scores decreasing from 20.3 (3.1) initially to 9.3 (2.1) at final platform follow-up.

Based on the research follow-up questionnaires, mean EPDS score in the MOVIN group decreased from 16.5 (Standard Error, SE 0.64) to 11.5 (0.67) at 12 weeks and 9.78 (12.1) at 24 weeks post-randomization. The mean difference (95% confidence interval) comparing the MOVIN group to the control group was −1.44 points (−3.21 to 0.34) at 12 weeks, and −2.32 points (−4.23 to −0.42) at 24 weeks post-randomization (Table 4). MOVIN participants were more likely to have EPDS ≤ 12 at 12 weeks (n = 26/43, 60.5% vs. n = 21/52, 40.4%, chi-square 3.80, p = 0.05) and 24 weeks post-randomization (n = 27/36, 75.0% vs. n = 23/45, 51.1%, chi-square 4.83, p = 0.03). Mean GAD-7, EQ-5D-5L visual analog scale, and PSI-SF scores were also lower in the MOVIN group, although not always statistically significant. There were no major changes in DAS scores in either group over time.

Table 4.

Maternal clinical outcomes

	Group	Mean (Standard error)	Adjusted mean difference (95% CI)
EPDS
Baseline	Intervention (n = 48)	16.5 (0.64)	−0.55 (−2.28 to 1.17)
	Control (n = 53)	17.1 (0.60)
12 Weeks	Intervention (n = 43)	11.5 (0.67)	−1.44 (−3.21 to 0.34)
	Control (n = 52)	12.9 (0.61)
24 Weeks	Intervention (n = 36)	9.78 (0.72)	−2.32 (−4.23 to −0.42)
	Control (n = 45)	12.1 (0.65)
GAD-7
Baseline	Intervention (n = 48)	10.9 (0.64)	−0.82 (−2.58 to 0.94)
	Control (n = 52)	11.7 (0.62)
12 Weeks	Intervention (n = 41)	8.27 (0.69)	−1.36 (−3.21 to 0.48)
	Control (n = 50)	9.63 (0.63)
24 Weeks	Intervention (n = 37)	6.21 (0.72)	−2.52 (−4.46 to −0.58)
	Control (n = 44)	8.74 (0.67)
EQ-5D-5L VAS
Baseline	Intervention (n = 47)	60.9 (2.54)	4.53 (−2.35 to 11.42)
	Control (n = 53)	56.4 (2.40)
12 Weeks	Intervention (n = 41)	66.3 (2.70)	10.04 (2.78 to 17.29)
	Control (n = 48)	56.3 (2.51)
24 Weeks	Intervention (n = 36)	70.4 (2.85)	6.33 (−1.30 to 13.97)
	Control (n = 43)	64.1 (2.62)
Dyadic Consensus Scale of the DAS (participants with a partner at baseline only)
Baseline	Intervention (n = 46)	49.1 (1.18)	−1.52 (−4.81 to 1.77)
	Control (n = 46)	50.6 (1.18)
12 Weeks	Intervention (n = 41)	49.4 (1.21)	−1.12 (−4.47 to 2.23)
	Control (n = 44)	50.5 (1.19)
24 Weeks	Intervention (n = 34)	50.6 (1.23)	0.01 (−3.43 to 3.45)
	Control (n = 40)	50.6 (1.21)
PSI-SF (participants postpartum at baseline only)
Baseline	Intervention (n = 26)	97.7 (3.03)	−0.08 (−8.26 to 8.09)
	Control (n = 31)	97.8 (2.78)
12 Weeks	Intervention (n = 24)	89.4 (3.10)	−5.25 (−13.6 to 3.06)
	Control (n = 30)	94.7 (2.81)
24 Weeks	Intervention (n = 19)	83.8 (3.31)	−8.52 (−17.3 to 0.27)
	Control (n = 25)	92.3 (2.95)

Intervention and control groups are compared over time using mixed linear models, with comparisons presented as mean differences (and 95% confidence intervals, CI).

EPDS Edinburgh Postnatal Depression Scale, GAD-7 Generalized Anxiety Disorder-7, VAS Visual Analog Scale, DAS Dyadic Adjustment Scale, PSI-SF Parenting Stress Index-Short Form.

For participant-reported health service use, a higher proportion in the MOVIN group reported accessing peer support and internet-based CBT (i.e., guided self-help), whereas they were less likely to have accessed group therapy and psychiatrist services (Table 5). There were no serious adverse events, and no adverse events assessed as being causally related to the study.

Table 5.

Participant self-reported health service use during the current episode of treatment, reported as N (%) based on what participants indicated they had accessed at 12- and then at 24-weeks post-randomization

	12 weeks post-randomization		24 weeks post-randomization
	MOVIN (n = 37)	Control (n = 43)	MOVIN (n = 36)	Control (n = 43)
Peer support	12 (34.3)	9 (23.7)	9 (32.1)	7 (20.0)
Group therapy	3 (8.8)	10 (25.6)	5 (18.5)	14 (38.9)
Internet-based CBT	18 (50.0)	5 (11.4)	18 (52.9)	9 (21.4)
Therapist	38 (79.2)	40 (75.5)	30 (62.5)	37 (69.8)
Antidepressant medication	13 (30.2)	16 (32.7)	12 (40.0)	12 (32.4)
Psychiatrist	9 (25.0)	15 (36.6)	11 (30.6)	17 (38.6)

CBT cognitive behaviour therapy.

Discussion

This pilot RCT aimed to examine the feasibility of a protocol for evaluating MOVIN for perinatal depression across a large health jurisdiction. Target recruitment was met within 6 months, participants found the intervention acceptable, and there was adequate research questionnaire adherence, supporting a future large definitive trial to confirm the pilot RCT’s promising efficacy results. Improvements to the model needed for the large trial mainly include a need for attention to technical aspects of the platform to improve ease of consistent access and navigation of resources.

The acceptability and preliminary efficacy results align with those of collaborative care models for depression, which show that it is superior to usual care for improving symptoms and other patient-focused outcomes^29,30. The results also align with literature that increasingly supports virtual care as a care-delivery model with high uptake and patient satisfaction in perinatal populations³¹. The fact that only about 5% of MOVIN group participants had a consultation with the MOVIN psychiatrist is consistent with other research on escalation rates to MDs in non-MD-led models of care³². This represents a promising model from the perspective of efficiency—i.e., potential to have lower waitlists for the limited specialized care available—and best utilization of provider expertise.

Different from most collaborative care models, where psychosocial and psychological interventions are delivered within the primary care setting, MOVIN-recommended interventions were drawn from existing community-based services that patients could access independently. This may possibly explain why only ~30% of participants agreed that MOVIN improved their overall access to healthcare services, as being in MOVIN did not mean that they would have preferential access to such services (except for consultation with the MOVIN psychiatrist, the three times that it was required). However, this approach was logical in our jurisdiction, given that most primary care teams do not have psychosocial and psychological services available for perinatal mental health³³, but there are multiple services available in the community through other public and private agencies. Of potentially even greater interest, while the control group participants were also informed in their handout about the same core services that the care coordinators were recommending via MOVIN and they did use services (supporting the idea that MOVIN itself did not necessarily improve a person’s “access” to services), the pattern of service use was different. Control group participants were less likely to reporting accessing peer support and iCBT, and more likely to access group therapy and psychiatrist services compared to the MOVIN group. They did not improve as much as the MOVIN group, though, suggesting that assistance with navigating potential treatment options, with follow-up when someone is not improving, may be the key ingredient in improving outcomes, which is consistent with the expected mechanism of a collaborative care intervention.

In planning for a future definitive trial, the applicability of the pilot trial methods and findings must be considered. Recruitment was rapid, and participants were diverse (~25% of participants identified as a visible minority, consistent with the population makeup of the Province)³⁴. However, it will be important to enact a comprehensive strategy around diversity and inclusivity in a larger trial, including ensuring gender-inclusive language in all materials, and population-specific resources (e.g., by gender, sexuality, cultural background), with input from persons with lived experience. Of note, immigrant women were only about 12% of the sample, whereas at least 30% of deliveries in Ontario are to immigrant women³⁵. Ideas for improving the inclusivity for immigrant and non-English speaking communities specifically include: (1) lowering the EPDS cut-off score for entry into the program (e.g. an EPDS of >9 instead of >12), as this may be more sensitive in certain Asian, south Asian and Southeast Asian populations³⁶, (2) use of the translated version of the EPDS (it has been translated into 58 languages) so that prospective patients can self-screen in their language of choice, and (3) providing translation services for interactions with the care coordinator (these are available publicly for hospital-based clinical visits in Ontario, currently).

Another consideration for the future trial is that the care coordinator was a nurse in the pilot RCT. This may be helpful in the Ontario setting for implementation since Ontario has perinatal public health nursing programs that could potentially assign staff to take on this type of role. However, in collaborative care models, other licensed providers (e.g., clinical social workers) have also been trained in the care coordinator profile. As such, a future large trial with multiple care coordinators could consider individuals from varied backgrounds to optimize options for implementation. As health services are planned and funded provincially in Canada for the most part, this could help other Canadian Provinces and Territories think about how the model could be optimally implemented in their setting if the results are positive.

Engagement was high—over 85% logged into the MOVIN platform and met with the care coordinator, so received the intervention as intended. However, there were technical aspects of the platform that could be improved. Better features for navigating the platform resources (i.e., including a feature to search for resources more readily) would be helpful and there was a clear preference for a mobile app (or at least better optimization of the mobile version), aligning with the types of technologies many patients are now using increasing often in their day-to-day. The finding that the some of the pre-specified psychological interventions were not always accessible or acceptable to participants is also important to consider. The care coordinator sourced additional interventions for almost all participants, including for related issues such as intimate partner violence. In the pilot, the MOVIN psychiatrist approved any new options considered to ensure that they were evidence-based interventions as per clinical practice guidelines. As resources in any health jurisdiction change constantly, it will be necessary in the future trial (or in any real-world implementation of this model) to have robust processes for identifying appropriate, accessible, treatment services and keeping them current.

There were some additional considerations that arose around the outcome measurement and timing. There was high follow-up at 12 weeks post-randomization, but the follow-up at 24 weeks post-randomization was also reasonable, suggesting that the longer follow-up time frame would be feasible for a primary outcome of a larger trial. This is important because a longer time frame is likely more reasonable for patients to receive a full course of treatment. Further, the use of the EPDS as the only primary outcome for a future trial also may be important to reconsider, given that a symptom scale based on the EPDS is used in the intervention group for ongoing treatment planning and iteration (i.e., if one group is completing the outcome measure repeatedly, but the other is not, this could bias the results). It may be important to consider adding a rater-reported outcome for an upcoming future definitive trial. In addition, as per the discussion above about lowering the threshold of EPDS score for entry into the trial (to >9, from >12), it may also be useful to consider lowering the EPDS score for remission as well (i.e., to ≤9 from ≤12).

Finally, this was a pilot study, so the efficacy estimates were preliminary and not meant to be definitive. There were also some imbalances between groups that were apparent despite randomization. For example, the proportion with a history of psychiatric hospitalization was higher in the MOVIN group. Since past psychiatric hospitalization likely increases risk for poor outcomes, this could have biased the estimate against MOVIN’s effectiveness (i.e., the effect size could be greater than what we observed herein). In a definitive trial, a larger sample size will likely lead to improved balance between the groups, and the analytic plan can be developed to address any remaining observed imbalances.

In summary, building on these promising results, the next step involves expanding to a larger trial to definitively evaluate the MOVIN model’s efficacy and capacity for real-world implementation. For this next step, the five large perinatal psychiatry treatment centres in Ontario are working together to engage a roster of specialized MOVIN psychiatrists to oversee the program over the longer term. The larger definitive trial will include a cost-effectiveness evaluation to inform decision-makers about whether and how to implement the program sustainably. It is also possible that in the future, the model of care could be extended in multiple ways, including for pre-pregnancy counselling, prevention, perinatal psychiatric disorders other than depression, and for different populations, including fathers requiring mental health care and/or caregivers requiring information and support.

Author contributions

S.N.V. and A.D. designed the intervention and led the conceptualization of the study, with input from C.L.D., S.G., T.J., and N.I. S.N.V. led the conduct of the study, with substantial support from A.D., M.A., K.B., S.L.G., M.M., and V.S. S.N.V. and S.L.G. led the analysis. All authors contributed to the interpretation of the study results. S.N.V. drafted the manuscript, and all authors revised the manuscript for important intellectual contributions. All authors approved the final draft of the manuscript.

Peer review

Peer review information

Communications Medicine thanks Zoe Darwin and the other, anonymous, reviewers for their contribution to the peer review of this work.

Data availability

We are unable to make the data for this study publicly available, as participants did not specifically consent to the sharing of their data at the time of initial data collection. However, data sharing agreements may be approved for specific purposes with research ethics board approval and appropriate contractual arrangements in place. Please contact simone.vigod@wchospital.ca for inquiries. Data are stored on secure servers at Women’s College Hospital in Toronto, Ontario.

Competing interests

S.V. reports royalties from UpToDate Inc. for authorship of materials related to depression and pregnancy. S.G. reports royalties from UpToDate Inc. for authorship of materials related to depression and pregnancy, the Canadian Pharmacists Association, and Norton for chapters on depression, and honoraria for presentations or writing from LifeSpeak. No other authors have competing interests to declare.

Supplementary information

The online version contains supplementary material available at 10.1038/s43856-026-01381-4.