Responding to the challenge of Adolescent Perinatal Depression (RAPiD): protocol for a cluster randomized hybrid trial of psychosocial intervention in primary maternal care

1. Among adolescents presenting with depression during pregnancy, the intervention package will, at 6 months postpartum, produce significant improvement in depression symptoms as assessed by the EPDS. For the purpose of this study, a mean total EPDS score difference of 2.0 between the arms will be regarded as a clinically meaningful difference in depression symptoms (see “Sample size determination for the RCT”).

2. The second primary hypothesis is that mothers receiving the study intervention will have significantly better parenting skills at 6-month postnatal follow-up compared to those receiving usual care as assessed with the HOME-IT [18, 19].

A series of formative activities took place at the beginning of the project. We conducted engagement meetings with key stakeholders consisting of decision-makers at the State Primary Health Care Board and various cadres of maternal frontline providers, including supervisory physicians, facility managers, community health workers (CHO), and community health extension workers (CHEW). Following the meetings, key informant interviews were held with frontline maternal care providers and with adolescent mothers. The key informants were selected from among providers and adolescent trial participants in our previously concluded RCT [15]. In the interviews with providers, we explored facility manpower and workload issues; experience with the delivery of treatment during the RCT, in particular engagement with the adolescents; and facility organizational issues, including those relating to the scheduling of clinic appointments for trial participants. With the mothers, who were adolescents during the trial, we explored perception of the appropriateness and usefulness of the services received by them, especially with regard to coping with parenting roles following childbirth; their level of satisfaction with the care they received; and what was lacking in the care received. With both groups we explored how evidence-based care for perinatal depression, including the provision of parenting skills, could be integrated into routine maternal care for adolescent mothers. Finally, we conducted a theory-of-change workshop with selected stakeholders to map the route from intervention design to delivery to expected outcome, identifying facilitators and potential barriers to successful project implementation.

This is a single-blind, cluster randomized trial. The study is taking place in primary maternal care clinics (MCCs) in Oyo State, Nigeria. These clinics were selected from across the 11 local government areas (LGAs) in and around the city of Ibadan. In Oyo State, maternal and child health care services are provided in community MCCs by primary health care workers (PHCW), with the different cadres consisting of nurse/midwives, CHO, and CHEW. However, the CHO and CHEW are the main direct clinical service providers. Along with training to provide care for persons with a range of common health problems presenting in primary care, all have also received basic midwifery training. Supervision is routinely provided to the frontline providers by general physicians, one each supervising a group of clinics located in the local government area. When required, these physicians make referrals to specialists, including psychiatrists, at either one of the two institutions with mental health specialists in the city.

Adolescents in the study intervention arm receive a manualized package of care that consists of: behavioral activation and problem-solving treatment; parenting skills training; and social and parenting skills support provided by a “neighborhood mother”. The behavioral activation and problem-solving treatment is delivered in six sessions during the antenatal period, consisting of three weekly sessions followed by three fortnightly sessions for those with EPDS scores of 12–17, or all six sessions delivered weekly for those with EPDS scores > 17. Supplemental sessions may be delivered if, during the 6-week postnatal visit, the providers find that the mother still has significant levels of depression (EPDS score of 6 or more). The number and frequency of these sessions are determined by the providers based on the level of depression (see Fig. 2). Treatment adherence is promoted by providers calling or sending text messages to their patients reminding them of appointments and agreed homework tasks from the PST sessions. This approach and the format of the problem-solving treatment are similar to what we used in our earlier RCT for perinatal depression. The parenting skills training is delivered in two ways: as a component of the problem-solving treatment which is provided in face-to-face sessions; and through mobile phone calls and texts (as appropriate) delivered in the postnatal period. There is a core set of themes to guide the provider during the calls, but attention to the particular needs or deficits of individual adolescent mothers is also addressed during the calls. The core themes consist of: personal and health care needs during pregnancy, including nutrition, rest, exercise, avoiding alcohol, and self-medication; preparing for childbirth; early signs of common ill health during pregnancy and in the infant; care of the newborn; infant feeding (including good and common bad practices to avoid); immunization schedule; stimulating and responding to infant’s needs; and dispelling cultural myths and taboos that are either harmful or unhelpful to good parenting. Many of the materials to inform the contents of the parenting skills training have been designed to address deficits in parenting skills among adolescent mothers as observed in our previous RCT. The materials also reflect information obtained during the formative qualitative activities. Providers make fortnightly calls starting shortly after the 6-week postnatal clinic visit of the mother and child, and continue until at least the 6-month postnatal outcome assessment. (We will encourage providers to continue with this, as needed, even after the discontinuation of the mother from the trial at the end of the 6-month postnatal period.)

A third component of the intervention package is the engagement of a “neighborhood mother” in the provision of social support and complementary parenting skills training to the adolescent. The idea of a “neighborhood mother” came from the extensive preliminary qualitative interviews with the adolescents when two observations were made: it is not uncommon to find that pregnant adolescents have been ostracized by their biological parents due to conflict related to the pregnancy, and may also not have support from the spouse’s parents; and as a result of this ostracization, the adolescent is often lacking in support from an experienced female who may themselves have nurtured children and can guide the adolescent in the basics of child care. For the purpose of this trial, the adolescent in the intervention arm is encouraged to identify a female in the neighborhood, who may or may not be biologically related, but who the adolescent can repose confidence in for needed social and instrumental support. The woman so identified is invited to the clinic with the adolescent for briefing by the maternal care provider and is enlisted to provide hands-on parenting skills training and support to the adolescent. The neighborhood mother agrees to work with the maternal care provider to address any identified skills deficits the adolescent may have. At each clinic visit and during postnatal telephone calls, the provider checks with the adolescent how the relationship with the neighborhood mother is going. The provider also makes regular telephone contact with the neighborhood mother to exchange experience on progress with addressing the adolescent’s needs.

Participants in the control arm receive usual care. As in the intervention clinics, providers in the control clinics have had prior training in the use of the mhGAP-IG. Usual care for perinatal depression in these facilities thus consists of the basic specifications of the mhGAP-IG for treating depression, and these include psychoeducation, reactivation of the social network, and addressing current psychosocial stressors. Providers decide on the number of sessions even though, if implemented according to the guide, patients with depression are expected to be seen a number of times. In this arm, there are no structured sessions of behavior activation and problem-solving treatment, no structured parenting skills training, and no engagement of a “neighborhood mother” in the provision of care to the adolescent.

We are conducting a comprehensive process evaluation to assess the barriers and facilitators of scaling up the intervention. The component of the project is informed by the CFIR [28], a state-of-the-art approach for exploring implementation issues. Guided by this framework, we will explore factors such as: intervention characteristics and the process of delivering the intervention (e.g. how engaging the intervention components are to providers and users); outer setting (e.g. current standard of care, workload); inner setting (e.g. organizational structure of the facilities and acceptability of the new intervention approach to both providers and patients); characteristics of key individuals (e.g. provider attitude, patients’ compliance and adherence to treatment, etc.); and process of service delivery (e.g. fidelity to guideline specifications, supervision frequency and content, etc.). We will explore these factors with providers (N = 15), adolescents who recover from depression (N = 10), and those who do not recover (N = 10) 6 months after childbirth.

Monitoring of adherence to trial protocol specifications is being conducted by the Trial Manager. This follows a structured format consisting of qualitative interviews and quantitative assessments. Detailed assessment of fidelity in the delivery of the intervention specifications is being conducted. This involves the live observation and rating of 20 randomly selected therapy sessions by different providers. For this purpose, we are using the 18-item Enhancing Assessment of Common Therapeutic factors (ENACT) tool [29] to evaluate the extent to which providers are applying the skills acquired to conduct focused psychological assessment of and intervention for the adolescents. ENACT is a tool that has been developed to provide reliable and valid assessment of therapist competence in a variety of cultural and service settings. This is being done by research supervisors, appropriately trained for the purpose, sitting in during clinical encounters between maternal care providers and the depressed pregnant adolescents.

To explore factors relating to the barriers and facilitators at various stages of project execution, qualitative interviews will be transcribed. Interviews will be conducted in Yoruba and will be translated into English, following which back-translation checks will be applied. The data generated will be analyzed using thematic analysis with the assistance of a qualitative software package, MAXQDA. The exploration of the findings will utilize the realist approach to address the question of “what works for whom and in what circumstance” [30, 31].

The analysis and presentation of the trial will be in accordance with CONSORT guidelines [32], with the primary comparative analyses being conducted on an intention-to-treat basis and due emphasis placed on confidence intervals for the between-arm comparisons. A full analysis plan will be developed prior to completion of data collection and discussed and agreed with the TSC.

We will use descriptive statistics to assess the balance between the trial arms at baseline for both MCC and individual participant characteristics. In order to take appropriate account of the hierarchical nature of the data, we will use multivariate mixed-effects regression models (logistic or linear dependent on outcome type) to estimate recovery from depression at 6 months for the intervention group versus the control arm, adjusting for baseline depression and randomization variables. In a secondary analysis, we will further adjust for any variables that were imbalanced between trial arms at baseline. We will analyze continuous outcomes (EPDS and GAD-7) over the 6-month postnatal follow-up period using repeated-measures analysis by including follow-up occasion as a random effect in the regression model. The primary outcome will first be analyzed by intention to treat without imputation. However, we will conduct sensitivity analyses to assess the potential effect of missing data using multiple imputation methods. No interim analysis is planned except when required to explore any undesired effect that may arise during the course of the trial and such analysis is deemed necessary and approved by the TSC.

We will use descriptive statistics to explore provider competence, as assessed by ENACT. We will evaluate the provider determinants of competence by determining the demographic predictors of scores on ENACT using regression analysis.