An assessment of hospital maternal health services in northern Ghana: a cross-sectional survey

Background: Access to and delivery of comprehensive emergency obstetric and neonatal care (CEmONC) services are often weak in low and middle-income countries affecting maternal and infant health outcomes. There are no studies on resources for maternal healthcare in the Northern region of Ghana. This knowledge is vital for health service planning and mobilising funding to address identified gaps. We investigated the available resources for managing CEmONC and referral services in the region. Methods: This study involved a cross-sectional survey of maternity facilities in ten hospitals in the Northern region of Ghana, serving a population of 2,479,461, including 582,897 women aged 15–49. Public and faith-based hospitals were included in the study. We used the Service Provision Assessment tool to gather data for this study between October and December 2019. Given the small sample size, we used descriptive statistics to summarise the data using SPSS version 25 and Excel 2016. Results: A total of 22,271 ANC visits from women to these hospitals occurred in the past 3 months preceding the study; however, 6072 birth events (cases) occurred within the same period. All the hospitals had less than one general medical doctor per 10,000 population (range 0.02–0.30). The number of midwives per 10,000 population ranged from 0.00 (facility H and J) to 1.87 (facility E), and none of the hospitals had a university-trained nurse designated for maternity care. Only one hospital had complete equipment for emergency obstetric and newborn care, while four others had adequate emergency obstetric care equipment. The number of maternity and delivery beds per 10,000 population was low, ranging from 0.40 to 2.13. Conclusions: The management of emergency obstetric care and referrals are likely to be affected by the limited human resources and equipment in hospitals in Northern Ghana. Financial and non-financial incentives to entice midwives, obstetricians and medical officers to the Northern region should be implemented. Resources should be mobilised to improve the availability of essential equipment such as vacuum extractors and reliable ambulances to enhance referral services. Considerable health system strengthening efforts are required to achieve the required standards. This study used a cross-sectional survey in ten hospitals in the Northern region of Ghana to assess the available resources for maternity healthcare including CEmONC and emergency referral. The hospitals comprised all the nine district hospitals in the region; consisting of public (2) and faith-based private-public (7) district hospitals as well as the regional hospital (1). This is the definition of the Ghana Health Service and WHO [22, 31, 32]. Maternity beds refer to all beds available for all maternity or labour related conditions including beds for resting while in labour, beds for childbirth and those used during the immediate postpartum period prior to discharge. On the other hand, delivery beds refer to beds that are exclusively designated for actual childbirth alone. The Northern region of Ghana is divided into the district, municipal or metropolitan local government areas. The metropolitan area has approximately 250,000 people; a municipality has at least 95,000 people while a district has approximately 75,000 people [33]. For the purpose of this study, the “district” denotes all the three local government areas. There are nine district hospitals across the 16 districts of the region. This study involved all the nine district hospitals and the regional hospital. We used the Service Provision Assessment (SPA) tool to gather data for this study. The tool has been uploaded as Additional File 1. The SPA was developed by the Demographic and Health Survey (DHS) Program for assessing health facilities in developing countries [34]. The assessment is carried out with a questionnaire that includes questions on facility level infrastructure, resources and systems, maternal and child health services, family planning and sexually transmitted infections [34]. The tool is suitable for assessing maternal and newborn healthcare, child healthcare, infrastructure, resources, systems, and family planning service provision. In 2002, the SPA was used to assess health service delivery in Ghana. It has also been used extensively in other Sub-Saharan Africa countries such as the Democratic Republic of Congo, Egypt, Kenya and Malawi [34]. For this study, we revised the tool to fit the characteristics of the northern region’s health facilities. For instance, the tool aggregates responses when assessing the number of equipment, such as beds. However, most of the facilities assessed had few beds (e.g. 70% had less than five delivery beds), so we reported the actual beds per facility instead of categorising or grouping the number of beds. Prior to data collection, invitation letters explaining the study’s purpose were sent from the Northern regional health directorate to Medical Superintendents of the nine district hospitals and the regional hospital selected. All heads of the hospital maternal health services were pre-informed by the Medical Superintendents. Following introductions and explanation of the study, we followed up to seek their interest and availability, obtain written informed consent, and schedule appointments with them. We interviewed the heads of maternity health services of the hospitals; however, their assistants were interviewed when it was not possible to meet them in person. Reported figures (e.g. number of ANC visits and number of births) were verified from the records of each facility to ensure accuracy. All interviews took place at the hospitals during regular working hours. Midwives, all medical officers and midwives in Ghana obtain formal training on EmONC (e.g. caesarean section, vacuum aspiration, blood transfusion, assisted deliveries) as part of their formal training and are also guided by the Safe Motherhood Protocol of Ghana Health Service which dictates how to handle maternity conditions [35]. Data collection was undertaken by two authors (EKA and RMA). EKA is a social scientist public health researcher with much experience in survey administration. RMA is a clinical midwife with expertise in assessing maternity healthcare who clarified the terminology and fieldwork procedures. Interviews lasted between 25 and 40 min. Data collection took place between October 2019 and February 2020. Our unit of analysis was at the facility level. We used SPSS version 25 and Excel 2016 for analyses. Given the minimal sample size, we used descriptive statistics to summarise the various aspects of the data. The results are presented in frequencies and percentages. In order to make the findings comparable to global standards as recommended by the WHO [15, 36], in each district/region, we computed the total number of births and ANC attendance within the 3 months preceding the survey per 100,000 (Table 1), maternal healthcare staff per 10,000 (Table 2), and the number of available beds (maternity and delivery beds) per 10,000 (Table 3). Finally, we summarised the available signal functions and essential equipment for maternity healthcare (Table 4). Births and ANC attendance in hospitals providing comprehensive emergency obstetric and newborn care per 100,000 population *Population data provided from the 2010 Census report (Ghana Statistical Service [GSS], 2010) * Further demarcation of regional and district boundaries was done between 2008 and 2019, adding two new regions from the original Northern region and ten new districts within the original Northern region. However, these newly created districts/regions do not have district/regional hospitals, and the population continues to be served by the existing health facilities Maternal healthcare staff at the hospitals per 10,000 population *Population data provided from the 2010 Census report (Ghana Statistical Service [GSS], 2010) * Further demarcation of regional and district boundaries was done between 2008 and 2019, adding two new regions from the original Northern region and ten new districts within the original Northern region. However, these newly created districts/regions do not have district/regional hospitals, and the population continue to be served by the existing health facilities Available maternity and delivery beds per 10,000 population *Population data provided from the 2010 Census report (Ghana Statistical Service [GSS], 2010) * Further demarcation of regional and district boundaries was done between 2008 and 2019, adding two new regions from the original Northern region and ten new districts within the original Northern region. However, these newly created districts/regions do not have district/regional hospitals, and the population continues to be served by the existing health facilities Availability of equipment and products for maternal healthcare, including referral CEmONC (comprehensive emergency obstetric and newborn care) The study protocol was assessed, approved, and ratified by the Navrongo Health Research Centre of the Ghana Health Service [NHRCIRB347]; and the Human Research Ethics Review Committee of the University of Technology Sydney, Australia [ETH19–4201]. Each participant provided written consent to participate in the study and to publish the data in a form that does not identify them in any way. All hospitals were de-identified and assigned identification letters.