Antenatal couples’ counselling in Uganda (ACCU): study protocol for a randomised controlled feasibility trial

Background: Common avoidable factors leading to maternal, perinatal and neonatal deaths include lack of birth planning (and delivery in an inappropriate place) and unmet need for contraception. Progress has been slow because routine antenatal care has focused only on women. Yet, in Uganda, many women first want the approval of their husbands. The World Health Organization recommends postpartum family planning (PPFP) as a critical component of health care. The aim of this trial is to test the feasibility of recruiting and retaining participants in a trial of a complex community-based intervention to provide counselling to antenatal couples in Uganda. Methods: This is a two-group, non-blinded cluster-randomised controlled feasibility trial of a complex intervention. Primary health centres in Uganda will be randomised to receive the intervention or usual care provided by the Ministry of Health. The intervention consists of training village health teams to provide basic counselling to couples at home, encouraging men to accompany their wives to an antenatal clinic, and secondly of training health workers to provide information and counselling to couples at antenatal clinics, to facilitate shared decision-making on the most appropriate place of delivery, and postpartum contraception. We aim to recruit 2 health centres in each arm, each with 10 village health teams, each of whom will aim to recruit 35 pregnant women (a total of 700 women per arm). The village health teams will follow up and collect data on pregnant women in the community up to 12 months after delivery and will directly enter the data using the COSMOS software on a smartphone. Discussion: This intervention addresses two key avoidable factors in maternal, perinatal and neonatal deaths (lack of family planning and inappropriate place of delivery). Determining the acceptability and feasibility of antenatal couples’ counselling in this study will inform the design of a fully randomised controlled clinical trial. If this trial demonstrates the feasibility of recruitment and delivery, we will seek funding to conduct a fully powered trial of the complex intervention for improving uptake of birth planning and postpartum family planning in Uganda. Trial registration: Pan African Clinical Trials Registry PACTR202102794681952. Approved on 10 February 2021. ISRCTN Registry ISRCTN97229911. Registered on 23 September 2021 A completed SPIRIT Checklist is available in Additional file 1. The CONSORT flow diagram for the study is shown in Fig. ​Fig.22. ACCU CONSORT 2010 flow diagram This is a feasibility non-blinded cluster-randomised controlled clinical trial of a complex intervention being conducted in primary health centres in Uganda. It is not possible to blind participants because the intervention is delivered through the local health centre. The use of cluster RCT design in this study is justified by the fact that the different components of the intervention package may not be delivered directly to individual participants but only applied at a cluster level [34, 35]. The unit for the administration of the intervention is a health centre rather than individuals. The cluster design minimises contamination by individuals to family, friends and community members since our intervention requires behavioural change as an outcome (uptake of birth planning and PPFP) [34, 35]. The unit “cluster” in this trial is a health centre III including the staff and patients in its catchment area. On average, the catchment area of a health centre III covers a population of 7846 women aged 15–49 years [36]. A health centre III cares for 4 to 6 parishes which are composed of 2–4 villages. The study area is in Mbarara District (rural) and Mbarara city (urban). According to the National Population and Housing Census (2014) Sub-County Report, women of reproductive age make up 25% (119,220/472,629) of the total population in the region [36]. This study targets pregnant women (and their partners) living in Mbarara in south-western Uganda. The primary participant will be the pregnant woman. If she meets the inclusion criteria, she will be included regardless of whether or not her partner consents to participate. Health centres to be included should meet the following criteria: Individual participants to be included should meet the following criteria: An individual will be excluded if she: In the catchment area of each of the four selected health facilities, all parishes will be selected for recruitment and follow-up of participants. In each parish, two VHTs will be invited to take part in the study (a total of 10 VHTs per area). The VHTs will be selected in a meeting organised at the health facility. The participating VHTs will identify all pregnant women and their partners from their community registers of antenatal couples. VHTs will visit the pregnant women and their partners in their villages to screen them against the eligibility criteria, explain the study and request their consent to participate. Each screening will be recorded in a questionnaire on their smartphone using the COSMOS app, which will record the reason for exclusion of any woman who is not included. Eligible women will be asked for their written consent to participate (Additional file 2). If they consent, their partner will also be invited to consent, in particular, to answer the questions posed by the VHTs. If the partner does not consent to take part himself, the woman will still be included in the study (as an intention to treat analysis). We will minimise selection bias by having a standardised data collection protocol, similar sources and methods of data collection in both intervention and control areas. The VHTs from both the intervention and control areas will be invited to attend a 2-day face-to-face training on the study, taking informed consent and data collection using the COSMOS app on a smartphone. Table ​Table11 shows the training schedule. Training schedule for VHTs on taking informed consent and data collection The complex intervention consists of the following components. In addition to the training provided to all VHTs (described above), in each intervention area, the VHTs will also be trained to provide basic information and counselling to antenatal couples in the community, encouraging them to attend a formal antenatal clinic, deliver in an appropriate place according to their level of risk, attend postnatal care clinic and use postpartum family planning. The information delivered by VHTs will also include birth preparation and complication awareness. The VHTs will take part in an additional face-to-face training course lasting 3 days, covering general counselling skills, couples’ counselling, identifying high-risk pregnancies, postpartum family planning, health education and gender-based violence (see Table ​Table2).2). VHTs in the intervention group will have several opportunities to deliver counselling to the woman (with or without her partner) and to show her the health education films on family planning on their smartphone: at the baseline antenatal visit, at the second antenatal visit, and if still relevant at the first and second postpartum visits. They will encourage couples to attend antenatal clinics together at the intervention health facility, where a trained health worker will be able to provide further couples’ counselling. VHTs in the control group will not be able to show the films and will not have been trained to provide counselling but may still provide general health promotion advice according to the standard training that all VHTs receive. Training schedule for VHTs in the intervention High-risk pregnancies Birth planning GBV gender-based violence, PPFP postpartum family planning, REDI rapport building, exploration, decision-making, implementation, RH reproductive health Health workers at the intervention health centre IIIs will be invited to attend a 6-day refresher training on general communication skills, couples’ counselling, information and practical skills for antenatal risk assessment, birth preparedness and postpartum contraception provision (see Table ​Table3).3). Health workers will be asked to provide couples’ counselling to couples who come together to the antenatal clinic. Training schedule for health workers Post-course evaluation ACCU study administration Closure To compensate for providing couples’ counselling at weekends, the intervention health centres will receive 600,000 Ugandan shillings (about 120 UK pounds) monthly. Intervention health facilities will be provided with a TV screen and health education films about the contraceptive implant. They will be asked to screen the films daily. The films will also be shown in the waiting room of the antenatal clinic and of the postnatal clinic, so the couples will see them if and when they attend antenatal and postnatal clinics. The films have undergone an intensive development in Uganda using the person-based approach to intervention development [31] and have been approved by the Ministry of Health. Supportive supervision visits will be undertaken at least every 2 months to VHTs and health facilities in the intervention arm, to ensure that the intervention is being delivered properly. Clusters are at the level of the health centre III (sub-county). In all the villages surrounding a health facility, all the couples will be in the same group. Two urban and two rural health centre IIIs will be randomly selected from a list of all health centre IIIs in Mbarara District (rural) and Mbarara city (urban), in south-western Uganda. We will use a stratified randomisation according to urban/rural location. Participating health facilities will be randomised during a meeting in which the district health officer is invited to draw pieces of paper out of a hat, to determine whether each health centre is assigned to the intervention or control arm—so that the randomisation process is open and understood by all. It will not be possible to blind participants or care providers. If two of the selected sub-counties share a boundary, we will only include women intending to attend ANC and PNC in their area, in order to minimise contamination. All trial procedures are shown in the SPIRIT figure (Table ​(Table4).4). The questionnaires will be completed using the COSMOS digital data entry platform on smartphones (https://cosmos.tetratech.com/). This allows data entry even in the absence of a mobile phone network and upload of data at a later time into the COSMOS database. All participating VHTs will be provided with a smartphone which will remain the property of the research team until the end of the study, when the VHTs will be allowed to keep them. VHTs will pilot the use of smartphones for data entry using a training version of the COSMOS app for 2–3 weeks before starting recruitment. VHTs will be given paper case record forms to use as a backup if there are technical problems which prevent data entry on the phones. The research assistant would then enter data from paper case report forms into the electronic database. VHTs will be trained to store paper forms securely until they can be passed on to the research team. ACCU SPIRIT figure showing the schedule of enrolment, interventions and assessments After a woman has consented to take part, the VHT may continue immediately with the baseline visit or may arrange to return at a later time. However, the baseline visit should be conducted no later than 7 months of pregnancy, to allow time for the couple to engage in counselling and decision-making before the delivery. The VHT will take the woman (and her partner, if he is present) through the baseline questionnaire (Additional file 3) about her demographic characteristics, any previous pregnancies, the current pregnancy and her future intentions regarding the place of birth and postpartum family planning. Village health teams in all areas will attempt to follow up on pregnant women and their partners a total of four times. VHTs will perform one antenatal follow-up visit at least 1 month after the baseline (data collection about use of antenatal clinic and intentions to use birth planning and PPFP). The postnatal visits will take place at about 1 week, 6 months and 1 year after giving birth (to collect data on place of delivery, uptake of postpartum contraception, perinatal and neonatal mortality). On each occasion, they will enter data into the relevant electronic questionnaire on their smartphone. VHTs in the intervention group will deliver additional counselling and will encourage couples to attend the health facility for couples’ counselling if needed, for couples who have not yet taken up postpartum contraception. In each health facility, health workers will be asked to report aggregate data every month on the number of women attending ANC alone or with their partners, the number of couples counselled about PPFP, the number of deliveries and the number of women provided with PPFP. All participating health centres receive a payment of 100,000 Ugandan shillings per month (about 20 UK pounds) for their help with data collection. In Uganda, domestic violence is common. According to the 2016 Uganda Demographic Health Survey [37], 56% of ever-married women and 44% of ever-married men have experienced physical, sexual or emotional violence by their current or most recent spouse/partner, and one in five have experienced physical violence in the preceding year. Therefore, it is likely that some of our participants will experience this during the course of the study, although we hope that our intervention may actually reduce the risk by fostering good communication and mutual understanding between partners in a couple. Nevertheless, at follow-up visits, the VHT will always ask briefly to speak separately with each member of the couple and will then ask screening questions about domestic violence, only if it is possible to speak in private. If VHTs become aware of domestic abuse occurring within one of the included couples, they will counsel and facilitate the victim to report the case within existing structures—initially, this would be the focal person for gender-based violence at the health facility, or the Local Council I leader who is responsible for addressing these issues and offering support as appropriate, and/or the community development officer or police. If the victim does not feel ready, the VHT will follow up and counsel her/him, with the goal of supporting her/him to disclose through the appropriate channels. The research team will monitor and report the progress of that referral as part of adverse event reporting, including a report of whether on review it is associated with being in the study or not. Any serious adverse event (resulting in hospitalisation or death) will be reported to the Mbarara University Research Ethics Committee (MUREC) using their standard serious adverse event reporting form. Data entry directly into the COSMOS system minimises the risk of data loss. We will control the quality of quantitative data collection using internal controls on the COSMOS system—for example, by checking the global positioning system (GPS) location of the interview and the time taken for the interview. We will also conduct spot checks—a researcher will repeat data collection for a random sample of about 5% of participants from each village health team member to ensure that the data has been captured accurately—if significant discrepancies are found, a larger sample will be repeated for that VHT. In the electronic database, we will use standardised data capture with drop-down boxes and data entry validation. Regular debriefing sessions will be held with VHTs for review of experience and discussion of difficulties. In complex intervention trials, it cannot be assumed that the delivery of an intervention in “real life” will be exactly as planned in the design stage of a trial. Process evaluation can lead to a greater understanding of what works or does not and provide meaningful interpretation of the effects of an intervention to inform future implementation [33, 38]. During the intervention, we will conduct a qualitative process evaluation aiming to identify operational reasons for failure or success in the implementation of the intervention and of the trial as a whole. We will assess the intervention fidelity and experiences of the intervention (in the intervention group) through ongoing monitoring and supportive supervision visits. Qualitative data will be collected using an interview guide to capture changes in implementation or contextual factors. The interview schedules will be altered in response to emerging themes developing from analysis which will take place concurrently with data collection. This will ensure that emerging themes from earlier data can be investigated in later data collections. We will hold the interviews and focus groups at least 3 months and up to 12 months after the delivery. We will start interviews and focus group discussions at least 6 months after the start of the trial and intervention. Interviews will be conducted by a member of the research team with training and experience in qualitative research, who is fluent in the local language, Runyankole. The interviews will take place in a location such as the participant’s home or private room at a health centre, guaranteeing privacy. They will be audio-recorded and will be transcribed verbatim and translated into English where English was not the language used during the interview, omitting any names so that transcripts are anonymous. Field notes will be taken to record non-verbal communication. The following will be the primary feasibility outcomes: The secondary outcomes (estimates of effect size for the main trial) will be the following: There are four clusters—two intervention and two control areas—in the feasibility trial. In each of these areas, we will engage 10 VHTs to recruit participants. Each VHT will aim to recruit 35 women over a 6-month period, so in total, we will aim to recruit about 1400 women. A formal sample size calculation is not required for a feasibility study, but a sample of this size will allow us to estimate the key parameters for a future full-scale trial, including the variability, clustering and retention rate. For the process evaluation, we will use a “purposive” sampling approach to select respondents (according to socio-demographic and organisational factors expected to influence the delivery or effectiveness of the intervention). We plan six separate FGDs, each of six to ten participants, two each for the following categories of individuals: (a) postpartum women—the age will determine their allocation to subgroups: adolescent girls (≤ 20 years), women 21 years and above; (b) men who are partners or husbands of women who deliver during the study; and (c) VHTs involved in the study. We will conduct about 20–30 in-depth interviews with clinic managers, health service providers, VHTs and couples. We will interview postpartum women and their partners, who took up, and who did not take up, the offer of couples’ counselling at home and at ANC and the offer of a contraceptive method, in order to understand the factors influencing their decisions, their experience of the intervention and their suggestions for improving it. We will also ask about the degree of contamination between study areas and will assess this as a feasibility outcome. The primary analysis will determine whether the study is feasible. Data will be explored descriptively and graphically for all feasibility outcomes including the intervention uptake, adherence, attrition, retention and the number of participants recruited per site. We will estimate the variance of proposed outcome measures; alongside previous literature, this will be helpful in a future sample size calculation. We will also estimate the intra-cluster correlation coefficient (ICC), and its 95% confidence interval, at both the randomisation and VHT level. Whilst these estimates are not by themselves sufficient for a future calculation, they can be compared to the existing literature and give some reasonable bounds for future estimates as well as allowing us to explore at which level clustering may be present so that this may be factored into the future design. All participant data will be included in the analyses, including those who have withdrawn, unless the participant specifically requests that their data be removed. The pattern and frequency of missing data will be explored descriptively to determine whether there are outcomes or items on instruments that participants opted not to complete. This may inform the decision about measures for a full trial. In-depth interviews and focus group discussions will be transcribed verbatim and translated into English, where English was not the language used during the interview, into Microsoft Word documents and analysed in Atlas.ti®. Thematic content analysis will be used [39]. Individual interviews or focus group discussions will constitute the units of analysis. At least two investigators will code the first three interviews/focus groups and will independently propose a coding framework. The coding framework will be discussed with the wider team before being finalised and applied to the interviews. Quantitative data on the role of costs both formal and informal in relation to the intervention will help in the design of a cost-effectiveness study. The cost-effectiveness of the intervention versus control depends on the additional cost of the intervention plus any knock-on effects on other costs. In this feasibility stage, both of these will be explored. We will evaluate the different ways of delivering the intervention in relation to its cost, given that the method of delivery may have to vary by place. The standard questionnaires will include questions about the costs of different components of care in both groups. The aim will be to assess in the process evaluation the likely cost of the intervention if delivered in different ways and to guide how it might be delivered most efficiently in the main trial. The knock-on effects of such an intervention, which may be diffuse (such as travel to the venue, time off work and cost of measures to encourage spouse attendance) will also be explored qualitatively in the process evaluation. We will also explore in an open-ended way what other non-healthcare costs might be incurred by trial participants. One perspective often adopted for costing is that of the health care provider, and this will be a perspective included in the main trial. Cost per birth averted is likely to be the measure of cost-effectiveness in the main trial but the appropriateness of this will also be explored particularly with the local health service leaders and taking account of other relevant work in the field and in terms of policy alternatives. Cost per quality-adjusted life year (QALY) is not usually seen as appropriate for interventions to do with fertility control. Thus, consideration will be given to the inclusion of other secondary outcomes by use of an impact inventory which will list all the effects, both to do with outcomes and resource use, of the intervention. The eventual cost-effectiveness analysis will compare the results from the trial with those in the relevant literature. Each participant will be given a unique identification number. Data entered via the COSMOS database is automatically backed up to 3 different data centre regions. All collected data is automatically deleted from mobile devices after successful synchronisation to the COSMOS database. Downloaded data and manually entered data will be stored on password-protected computers. All access to data is very tightly and strictly controlled by the University of Southampton. Transcripts from interviews and focus groups will also be anonymised and shared between the research groups. At the time of publication, we will make the anonymised data freely available on a data repository of the University of Southampton. COSMOS uses built-in advanced encryption technology (AES256) to safeguard data and meet obligatory commitments towards data protection. Data entry directly into the COSMOS system minimises the risk of data loss or breach of confidentiality. The VHTs will use a lockable cupboard to keep documents. The research assistant will monitor this and will collect study documents on a regular basis and transfer the documents to the secure locked cabinet in the study office. The study database will be password-protected and will be stored on a password-protected computer. Data entered into the study database will be anonymised. Qualitative data will be recorded on a digital voice recorder, which will also be stored securely in a locked cabinet in the research office. The recordings will be transferred to a secure password-protected computer. They will be transcribed and translated into a Word document on the same computer where the recordings are downloaded, omitting any names so that transcripts are also anonymous. Once the transcripts have been checked and corrected, if necessary, the voice recordings will be deleted. Randomised controlled trials have special ethical concerns and the ethical reason for randomly assigning the intervention arm is because we expect both arms to be at a state of equipoise. For the purpose of this feasibility trial, the research participants in the control arm have access to standard care that is deemed adequate healthcare by the Ugandan Ministry of Health. The study was approved by MUREC (the Mbarara University of Science and Technology Research Ethics Committee, Ref #. 16/06-20) and the University of Southampton, Faculty of Medicine Ethics Committee (Ref #. ERGO 54459.R3). All the study methods and procedures will be conducted in accordance with MUREC’s guidelines and regulations. Written informed consent will be obtained from participants of the study. In the case of illiterate participants, they will be asked to use a thumbprint, and an independent witness will sign to confirm that they gave their consent freely in line with the MUREC guidelines. We will train all research assistants and village health teams in relevant principles of good clinical practice, particularly taking informed consent, and confidentiality. All protocol modifications will be communicated to MUREC for approval before they are implemented, and protocol modifications relevant to health care provision will be communicated to study sites. The currently approved version of the protocol is version 2, dated July 29, 2020. We are required to report any deviations from the protocol to the Mbarara University of Science and Technology Research ethics committee. The study will comply with the data protection guidelines of the University of Southampton. This trial has an independent Trial Steering Committee (TSC) that acts as the oversight body on behalf of the sponsor and funder. At least half of the members, including the chair, are independent. The committee will meet at least once a year. No Data Monitoring and Ethics Committee will be convened for ACCU, as this role will be assumed by the TSC. The role of the Trial Steering Committee is to ensure the trial is on track as set out in the protocol, to ensure the safety of the patients, to help deal with any problems or issues that might arise and to advise the principal investigators and co-investigators. Among the terms of reference of the steering committee, as stipulated in their charter, is to recommend whether to continue or terminate the study or further adapt it based on safety and efficacy considerations.