Ghana’s Ensure Mothers and Babies Regular Access to Care (EMBRACE) program: Study protocol for a cluster randomized controlled trial

Background: The United Nations’ Millennium Development Goals call for improving maternal and child health status. Their progress, however, has been minimal and uneven across countries. The continuum of care is a key to strengthening maternal, newborn, and child health. In this context, the Japanese government launched the Ghana Ensure Mothers and Babies Regular Access to Care (EMBRACE) Implementation Research Project in collaboration with the Ghanaian government. This study aims to evaluate the implementation process and effects of an intervention to increase the continuum of care for maternal, newborn, and child health status in Ghana. Methods/Design: We will conduct a cluster randomized controlled trial using an effectiveness-implementation hybrid design in Dodowa, Kintampo, and Navrongo, Ghana. We will provide an intervention package to women living in randomly allocated intervention clusters. The study population is women of reproductive age between the ages of 15 and 49 years. The package includes: 1) use of a new continuum of care card, 2) continuum of care orientation for health workers, 3) 24-hour health facility retention of mothers and newborns after delivery, and 4) postnatal care by home visits. We will measure maternal, newborn, and child health outcomes for both intervention and implementation impacts. The intervention outcomes are continuum of care completion rate, rate of postnatal care within 48 hours, complication rate requiring mothers’ and newborns’ hospitalizations, and perinatal and neonatal mortality. The implementation outcomes are intervention coverage of the target population, intervention adoption and fidelity, implementation cost, and sustainability. Discussion: In this trial, we will investigate how successful continuum of care can contribute to improving maternal, newborn, and child health outcomes. If successful, this model will then be implemented further in Ghana and other neighboring countries. Trial registration: Current Controlled Trials ISRCTN90618993. Registered on 3 September 2014.

We will conduct a cluster randomized controlled trial using an effectiveness-implementation hybrid design [36,61,62]. In this study design, we will assess the effects of both an intervention package and its implementation process. An effectiveness-implementation hybrid trial is an innovative design to shorten the time from intervention development through its implementation in a real-world setting [63,64]. This study design is categorized into three types. The type one design focuses on testing the intervention while observing or gathering information on the implementation. The type two design simultaneously tests the intervention and its implementation strategy. The type three design focuses more on testing implementation strategy while observing or gathering information on intervention outcomes [61]. In our study, we will adopt the type two hybrid design to determine both the impact of the intervention on CoC completion in MNCH and the acceptability of the implementation strategy by GHS. This design will assess the effectiveness of intervention and implementation by using cluster randomized allocation to divide participants into intervention and control arms. For trial sustainability and scalability, we will consider site-specific contexts in designing interventions [59]. We will conduct the study in three stages: baseline, implementation, and follow-up phases. Acceptability of the trials will also be evaluated through monitoring ongoing trials. We will conduct the study at three different sites in Ghana; Dodowa (Greater Accra region), Kintampo (Brong-Ahafo region), and Navrongo (Upper East region), where the GHS runs HDSS sites (Figure 1). These three sites were selected because highly reliable HDSS data are available. HDSS involves semi-annual recording of vital demographic events occurring among residents of all households in the HDSS area: pregnancies, births, deaths, and migration. Other data are also collected and updated regularly, such as economic status, morbidity, and vaccination. Map of health and demographic surveillance system sites. Ghana has health and demographic surveillance system (HDSS) sites in Dodowa, Kintampo, and Navrongo. The HDSS areas involve highly reliable semi-annual recording of vital demographic events occurring in residents of all households. At the Dodowa HDSS site, the population was approximately 115,000 in 2011. It is located about 40 kilometers away from Accra [65]. Consequently, pregnant women from this site often prefer to deliver at health facilities in Accra. Service deliveries are challenging since the site land size covers about 40.5% of the Greater Accra region. Most of the communities are scattered, containing small populations [65,66]. At the Kintampo HDSS site, the surveillance population was approximately 200,000 in 2011. This site is a multi-ethnic area and farming is the most important economic activity. Apart from the central area, most villages are not supplied with electricity and are reached by dirt roads. Access to health facilities is a challenge in Kintampo [65,67], and a home delivery is often the first choice [68]. At the Navrongo HDSS site, the surveillance population was approximately 153,000 in 2011 [65,69]. Navrongo is the first area where the Community-based Health Planning and Services (CHPS) program was launched in Ghana. In that context, Community Health Officers (CHOs) have contributed to improving health status in communities. Among the three HDSS sites, only Navrongo is on track to achieve MDG 5 [48]. As the MNCH services are national priority areas in Ghana, various MNCH interventions have been implemented in the three HDSS sites. Representative interventions include Mobile Technology for Community Health (MoTeCH, between 2009 and 2012) [70], Quality of prenatal and maternal care (Qualmat, between 2009 and 2014) [71], and Ghana Essential Health Interventions Programme (GEHIP, between 2009 and 2014) [72] in Navrongo; early neonatal vitamin A supplementation in improving child survival (Neovita, between 2010 and 2012) [73], and Newborn home intervention study project (Newhints, between 2008 and 2010) [74] in Kintampo; and the conditional cash transfer project (GLST, between 2009 and 2014) [75] and Neonatal Quality Improvement Programme (NQIP, between 2011 and 2012) in Dodowa. Each HDSS site contains two districts, and the districts comprise multiple sub-districts, which are the minimum health administration units in Ghana. We defined the sub-district as a cluster unit. Only Jema and Dumso (which used to be a part of Jema) sub-districts in the Kintampo site were combined to form a cluster. In the three HDSS sites, a total of 36 sub-districts were included. Of them, we excluded four sub-districts of Navrongo where another MNCH project is planned. In total, we chose 32 clusters for our study targets (Figure 2): eight clusters in Dodowa, 12 clusters in Kintampo, and 12 clusters in Navrongo. We allocated half of the sub-districts to the intervention arm and the other half to the control arm. At least one hospital was allocated to the intervention arm in each site, since the majority of facility deliveries are done in hospitals (Table 1). Due to the nature of the intervention, masking was not feasible. Survey flow diagram for participant recruitment and inclusion. Sub-districts were defined as a cluster unit. In total, 32 clusters were chosen as the study targets; eight clusters in Dodowa, 12 clusters in Kintampo, and 12 clusters in Navrongo. Half were allocated to the intervention arm. HDSS; Health and Demographic Surveillance System. Number of facilities (As of June 2014) CHPS, Community-based Health Planning and Services; HDSS, Health and Demographic Surveillance System. We matched the clusters within the site before randomization in the following aspects: population, number of deliveries in each cluster, and number of midwives available. Cluster randomization was preferred over individual-level randomization to minimize contamination, and also for the pragmatic purposes of a future scale-up of the intervention. The clusters were randomized by a data analyst, who was not a primary member of the study team, using computer-generated random sequences. Our study population is women of reproductive age between the ages of 15 and 49 years who live in the areas covered by the Dodowa, Kintampo, and Navrongo HDSS sites. Participants will be women who have given birth in the study area between 1 September 2012 and 30 June 2014 for the baseline period, and between 1 October 2014 and 30 September 2015 for the trial period. Exclusion criteria for the trials are women who refuse to participate in the intervention. We expect only a few women will meet it as the interventions will be implemented within the current service delivery system. In other words, they will receive the interventions unless they refuse national standard MNCH services. Exclusion criteria for the impact evaluations are those who decline to be interviewed or who have migrated out of the HDSS sites. We will also involve health service providers in the intervention to provide CoC services and to educate participant women in the CoC concept. We will deliver the interventions to all eligible women in the intervention arm. MNCH service providers will approach them when receiving MNCH care at health facilities or home. The expected recruitment period will be from August to September 2014 for the baseline survey, from October 2014 to September 2015 for the intervention, and from October to November 2015 for the follow-up survey. We will request informed consent form signatures from the survey participants. Additionally, we will obtain participants’ oral consent at their intervention entry point. Their participation will be voluntary and they are free to join or withdraw at any time. We developed a package of interventions based on a formative research survey conducted in three different HDSS sites (Table 2). In the survey, we identified the gaps, barriers, and promoters of service reception in continuous MNCH care. To address the actual MNCH conditions in Ghana and site-specific cultural differences in delivery location, geographic conditions, and health system capacity, we developed the following four interventions: the utilization of the CoC card (A-1), CoC orientation (A-2), 24-hour health facility retention of mothers and newborns after delivery (B-1), and PNC by home visits (B-2). We will implement A-1, A-2, and B-2 interventions in all three HDSS sites. However, B-1 will be implemented only in Dodowa and Navrongo because the number of midwives is not enough in Kintampo. Interventions of Ghana EMBRACE implementation research project ANC, antenatal care; CHN, Community Health Nurse; CHO, Community Health Officer; CHPS, Community-based Health Planning and Service; CoC, Continuum of Care; DHMT, District Health Management Team; SDHMT, Sub-District Health Management Team; GHS, Ghana Health Service; MNCH, maternal, newborn, and child health; PNC, postnatal care. In the control arm, Ghanaian government will provide conventional MNCH services. In most of the primary level health facilities, basic MNCH services will be provided. The delivery and laboratory services are provided mainly in the secondary and tertiary level health facilities. ANC services include general check-ups for mothers, urine tests, hemoglobin tests, prevention of mother-to-child transmission (PMTCT) of HIV, nutritional support, tetanus toxoid immunization, and health education for birth preparedness and maternal complications. Delivery services include skilled delivery, facility referral, and emergency obstetric care. PNC services include general check-ups for mothers and children, body weights, infant vaccinations, hemoglobin tests, nutritional support, health education for breastfeeding and child care, and family planning counselling. Home visit check-ups are also provided, however, home visit PNC within 48 hours is poorly adopted. In addition to the maternal health record book, health service providers will provide the MNCH CoC card to all women who receive ANC, delivery, and PNC assisted by SBAs (Table 3). When these women receive follow-up services, health service providers place a sticker on the card to show that they complied with the CoC; this includes four ANC visits, delivery assisted by SBAs, and three PNC visits (within 48 hours, at seven days, and at six weeks after delivery). Additionally, health service providers record on the CoC card: provision of essential services and health education; blood tests for assessing hemoglobin, blood group, and Rhesus factor; intermittent preventive treatment (IPT) for malaria; tetanus toxoid immunization; early initiating and exclusive breastfeeding; family planning; preparing items for delivery and baby; arrangement of transportation for delivery, caregiver, and calling health service providers after delivery; and the presence of danger signs during pregnancy and after delivery for mother and newborn. Maternal, newborn, and child health continuum of care card contents ANC, antenatal care; CoC, Continuum of Care; IPT, Intermittent Preventive Treatment; PNC, postnatal care. We will implement the MNCH CoC orientation in two stages. First, we will complete the training of trainers for supervisors of District Health Management Teams (DHMT) and Sub-District Health Management Teams (SDHMT). The trained supervisors will then conduct orientations for health service providers at hospitals, health centers, CHPS, and private clinics. We will focus mainly on introducing MNCH CoC concepts, their importance, the extent to which the MNCH CoC card could be used, and the protocols of other interventions. In this intervention, we will encourage mothers to stay with their newborns at health facilities after delivery for at least 24 hours for PNC. This intervention targets only the health facilities where SBAs provide delivery services. In this intervention, mothers and newborns will receive necessary care in a health facility during the 24 hours after delivery. During their stay, we will provide them nutritious drink supplements (for example, Milo (Nestlé S.A., Vevey, Switzerland)). After ensuring all of the necessary health check points, health service providers will discharge them if neither mothers nor babies show any danger signs. We will encourage CHOs to visit mothers and newborns for PNC within 48 hours after delivery. This intervention is composed of two steps. The first step is delivery notification; when home delivery occurs, community health volunteers, community key informants, or traditional birth attendants will inform the CHOs that the labor or delivery occurred in the community. The second step is PNC by home visits; CHOs visit mothers and newborns for PNC within 48 hours postpartum. Where necessary, we will provide beds for postpartum rest, rechargeable lamps or solar lanterns, torchlights for all B-1 intervention facilities, and motorbikes for the B-2 intervention facilities. We will also provide care materials for both eligible categories of health facility such as blood pressure apparatus for mothers and children, stethoscopes, thermometers, and pen lights. We will organize three types of supportive activities before the start of the intervention to implement the interventions smoothly into the communities. They include stakeholder meetings, community leader meetings, and community durbars. We will organize the stakeholder meetings in both the intervention and control arms. To avoid contamination of the intervention effect, the community leader meetings and community durbars will be held only in the intervention arm. In each HRC, we will form an intervention monitoring team. This team is responsible for monthly monitoring of all implementation activities of health service providers and for supervision of the DHMT/SDHMT supervisors. The EMBRACE researchers will meet monthly to discuss the issues raised during the monitoring and provide feedback to the monitoring team and the DHMT/SDHMT supervisors. We will assess the impact of the intervention by interview survey using a semi-structured questionnaire and HDSS data. We will also assess the acceptability of the implementation process by monthly monitoring and supervision, and by conducting a survey of eligible women at the HDSS sites. Data managers of the HRCs will review all of the collected data for accuracy and completion. Data will be entered into computers twice, using Visual FoxPro software (Microsoft, Washington, United States). Verification checks will be done to correct any discrepancies in records. We will assess the intervention efficacy using both survey and HDSS data, as described below: The questionnaire items are: sociodemographic and socioeconomic characteristics; MNCH services uptake; health complications during pregnancy, at delivery, and during postnatal period; pregnancy outcomes; and care-seeking behaviors. To control for the effects of health care provider and health facility characteristics on the study outcome, we will conduct health care provider assessment and health facility assessment at the baseline survey for both intervention and control arms. The data we will collect includes information on service provision, knowledge of CoC, supervision, community support, job satisfaction at health care provider level, infrastructure, human resources, MNCH service availability, annual statistics of key MNCH indicators, current practice of retaining mother and newborn for the first 24 hours, and provision of PNC by home visit within 48 hours postpartum. Prior to the survey, we will complete training for research assistants at each site. The EMBRACE researchers will develop the survey questionnaire through several workshops based on the formative research questionnaire. We will develop the questionnaire in English and the research assistants will interpret it orally when they conduct interviews with participants. As the local languages are not always written, we will not use translated questionnaires. The trained research assistants will pre-test the questionnaire with 20 eligible women and the researchers will revise it accordingly. We will assess the acceptability of the implementation process by monitoring, supervision, key informant interviews, and surveys, as described below: The primary outcome of the intervention is CoC completion rate of mothers and their children (Figure 3 and Table 4). The secondary outcomes include the PNC rate within 48 hours, the complication rate requiring mothers’ and newborns’ hospitalizations, and the PMR and NMR. The PMR is defined as fetal deaths during any period of pregnancy and newborn deaths within seven completed days after birth. The NMR includes early neonatal deaths occurring during the first seven days of life and late neonatal deaths occurring after seven days but before 28 completed days of life [76,77]. Conceptual framework. The intervention package includes the utilization of the CoC card, CoC orientation for health service providers, 24-hour health facility retention of mothers and newborns after delivery, and home visit PNC. The outcomes will be measured for both intervention and implementation aspects (ANC, antenatal care; CoC, Continuum of Care; NMR, neonatal mortality rate; PMR, perinatal mortality rate; PNC, postnatal care). Outcomes of Ghana EMBRACE implementation research project ANC, antenatal care; CoC, Continuum of Care; HDSS, Health and Demographic Surveillance System; MNCH, maternal, newborn, and child health; NMR, neonatal mortality rate; PMR, perinatal mortality rate; PNC, postnatal care; SBA, skilled birth attendant. The implementation outcomes will be measured by five outcomes: 1) intervention coverage of target population, 2) adoption and 3) fidelity in CoC card utilization or PNC within 48 hours by mothers’ retention at health facility or by home visit, 4) implementation cost, and 5) sustainability [64,78,79]. We made a calculation for two types of sample size to measure different outcomes. At first, to measure the CoC completion rate, we used the interview survey and calculated a total of 1,500 women for the sample size. Second, to measure the NMR, we used a HDSS data and calculated a total of 15,000 women for the sample size. To calculate the sample size of the CoC completion rate, we used the data of the formative research collected in the previous year in the same sites. According to the data, the CoC completion rate was 8.0%. The coverage of four ANC visits, delivery attended by SBAs, PNC within 48 hours postpartum, and PNC at two weeks were 86.6%, 75.8%, 13.0%, and 60.0%, respectively. The lowest coverage was identified at PNC within 48 hours postpartum, and if this coverage increases to more than that of PNC at two weeks, the CoC completion rate would also increase to 60.0%. For that, we estimated that coverage of four ANC visits would be improved from 86.6 to 95.0%, and calculated the sample size using an intraclass correlation coefficient (ICC) of 0.02675 determined in the formative research. The ICC was estimated by considering the differences in the sizes of clusters using multilevel regression with a random intercept at the cluster level. The confidence interval (CI) was 95%, and power was 80%. Adding 10% for potential attrition, the total sample size was calculated at 1,500 for each baseline and follow-up survey period. Therefore, we estimated approximately a 500-person sample size for each HDSS site (1,500 in total). In addition to the interview survey, we will also use the HDSS data to evaluate the effect of the interventions on perinatal and neonatal mortality. For the baseline and follow-up survey periods, HDSS data will capture 15,000 pregnancy cases, respectively. The sample size was estimated according to the following assumptions: a 25% reduction of PMR (from 31 to 23 per 1,000 pregnancies), with 95% CI, 80% power, 32 clusters, and an ICC of 0.0007256. The ICC was based on previous MNCH research conducted in the study area [74] and the sample size was calculated at 13,548. In addition, we added 10% for potential attrition. We will conduct baseline and follow-up interview surveys to assess changes in each outcome in both the intervention and control arms. To minimize overestimation of intervention impact, we will estimate all intervention impact outcomes with an intention-to-treat analysis. In this analysis, individuals’ outcome data are analyzed according to the allocated arm regardless of the place where they received care. Also, all eligible individuals are included in the analysis regardless of whether they provided outcome data [80]. We will conduct a descriptive analysis to assess the distribution of various factors related to the mothers and newborns under study. To evaluate the average effectiveness of interventions on CoC completion, we will apply the generalized estimating equations model with both continuous and binary outcomes [80]. This model uses data on all mothers, including those with incomplete data, over the period from 16 weeks of pregnancy to six weeks postpartum. Adjustment factors include basic demographic characteristics, socioeconomic characteristics, and facility characteristics. The data analysis will mostly be conducted using Stata version 13 (StataCorp LP, College Station, Texas, United States). PMR and NMR will be calculated based on the number of live births and perinatal deaths or neonatal deaths as a total of the three HDSS sites, as well as at each HDSS site. These outcomes will be compared before and after interventions. Implementation impact will be evaluated by means of descriptive statistical analysis. The qualitative data will be coded and categorized. Ethical approval was obtained from the Ethics Review Committee of the GHS (reference: GHS-ERC: 13/03/14), the Institutional Review Boards of Dodowa HRC (reference: FGS-DHRC: 280214), Kintampo HRC (reference: 2014–11), and Navrongo HRC (reference: NHRCIRB137) in Ghana, and from the Research Ethics Committee of the University of Tokyo in Japan (reference serial number: 10513). Informed consent will be obtained from all survey participants before their inclusion in the study. We will record the consent through a signature or thumbprint. Participants will be withdrawn from the study if they experience a serious or intolerable adverse event, develop or disclose symptoms or conditions listed in the exclusion criteria, or require early discontinuation for any other reason. This will not affect their normal service delivery at any of the health facilities. Permission for conducting the intervention study will be sought from the local health authorities and community leaders before initiating the study. This intervention package is not invasive. Thus, the participants will not be exposed to marked risks. By participating in this study, they benefit by improved CoC knowledge and care of mothers and newborns. The MNCH service providers will receive a training session about CoC and the procedures to be performed. Health facilities of the intervention arm will receive a set of PNC services, motorbikes, or rest beds if they are not available ones. We will introduce the intervention to the control arm participants immediately when a positive impact is identified. We will encourage referral to a health facility if we identify minor, acute, or chronic illness in mothers or newborns in either the intervention or control arms. All information obtained through this study will be confidential. Access to information will be limited to research assistants for conducting interviews and data entry management staff. Study records will be identified only by means of study identification numbers. The results of the study will be presented first to community members and their leaders in HDSS sites. In addition, policy briefs will be developed in collaboration with the Policy, Planning, Monitoring, and Evaluation Division of the GHS and submitted to the Office of the Director-General of GHS and the Family Health Division. Presentations will be made at the GHS Directors meeting, the Senior Managers Meeting, and international conferences. Trial findings will also be disseminated in scientific meetings and papers on the intervention impact on improving CoC, the impact of increased CoC on MNCH status, the acceptability of the interventions, and the strategies for Japan’s international health policy for MNCH.

N/A

• Dima Health