Health system strengthening to reduce maternal and perinatal deaths in Central Region of Ghana – project approach

Abstract

Background

Institutional maternal and perinatal outcomes are highly specific to local contextualized conditions and are suited to inform local policy. As institutional delivery increases, maternal health events are more likely to occur in facilities than at home.

Methods

This was a before-after impact evaluation study through the implementation of a 3-year quality improvement project undertaken in Central Region. The goal was to reduce institutional maternal mortality ratio and perinatal mortality rate over a 3-year period from 2021 to 2023 with 2020 as a baseline year. Baseline assessment was conducted by reviewing 2020 maternal death audit reports to identify key challenges associated with maternal deaths. Interventions were drawn to address identified challenges by key stakeholders. Sensitization meetings were held with stakeholders, health managers and health staff to discuss the interventions and solicited their support. Interventions implemented were life-saving skills and safe motherhood trainings for health staff across the region, introduction of maternal death tracker to improve surveillance on maternal death, clinical supervision of health staff at the peripheral level by specialist through telemedicine using WhatsApp platforms from 2021 to 2023.

Baseline indicators were collected from the District Health Information Management System-2 (DHIM2) in 2020. The outcome indicators were improved family planning coverages, increased skilled delivery rates and reduction in maternal mortality ratio and perinatal mortality rates. Process indicators were proportion of practicing midwives and clinicians trained in life saving skills, number of health centers conducting basic emergency obstetric care signal functions, percentage of audited maternal deaths entered DHIMS2.

Results

Over the course of the project, 4,800 health staff were trained, skilled birth rate improved from 61.6 to 63.9%, family planning acceptor rate from 28.5 to 45.1% and reported maternal deaths audited improved from 78.7 to 100%. The implementation of institutional audit recommendations also improved from 20 to 72%. Institutional maternal mortality ratio declined by 25.3% and neonatal deaths by 45.2%.

Conclusions

Quality improvement methods using the project approach to strengthen maternal health services can lead to improve mother and baby health outcomes. This approach when scaled up nationally may help increase the chances of meeting the sustainable development goal three, I recommend the implementation of the interventions to policymakers such as the Ministry of Health to support in reducing institutional maternal and perinatal mortalities.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12884-026-09593-1.

Introduction

In 2023, about 260 000 women died from pregnancy, delivery and related conditions globally. About 92% of all maternal deaths occurred in low- and lower-middle-income countries in 2023, and most of the deaths could have been prevented [1, 2]. Approximately 239,000 women die annually due to pregnancy and related complications in sub-Saharan Africa where women have over 200 times lifetime risk of dying compared to those in some developed countries [1, 2].

Additionally, the global coverage of skilled attendance at birth is estimated at 81%, compared with only 60% in sub- Saharan Africa [3]. As institutional delivery rates continue to rise, more individuals with pregnancy related complications will access treatment and mortality events are more likely to occur in facilities than at home. Facility-based events are highly specific to local contextualized conditions, and are therefore well-suited to inform local policy makers, clinicians, and programs to target specific health system strengthening efforts. Programs targeting health system strengthening can improve service quality by addressing local conditions. Ghana’s maternal mortality ratio declined from 760 per 100,000 live births in 1990 to 310 per 100,000 live births in 2017 [4]. In 2008, the Government of Ghana introduced the free maternal health policy to address financial barriers to maternal health services and reduce preventable maternal deaths [5, 6]. World bank estimates put maternal mortality ratio (MMR) for 2023 at 234/100,000 live births [7] showing a steady reduction in MMR, however considering the rate of decline, much effort will be needed for Ghana to achieve the Sustainable Development Goal of 70/100,000 live births in 2030.

In the Central Region of Ghana, annual institutional MMR has averaged 114 per 100,000 live births between 2015 and 2020 [8]. In 2020, a total of 75 institutional maternal deaths mainly from hemorrhages, hypertensive disorders, and septic abortions occurred. Six administrative districts in central region of Ghana namely Mfantseman, Cape Coast, Abura-Asebu-Kwamankese, Gomoa West and Komenda-Edna-Eguafo Abirem accounted for over 60% of the cases, poor referral practices and institutional factors relating to quality of care contributed to this [8]. To improve quality of care, quality improvement (QI) approach can be used. QI is a systematic approach, guided by data, to improve health care delivery and outcomes. The QI initiatives can be led by a team at community, sub-district, district and regional levels of care. As part of their responsibilities, QI teams can review data, undertake small tests of change, and monitor progress, such focus which requires time and a coordinated mindset and effort, can create significant transformations [9]. A team made up of the authors and specialists conducted a baseline study in 2020 through review of maternal death audit reports. The review identified challenges such as poor institutional surveillance on maternal deaths leading to poor accountability for maternal deaths through audits. For example, some deaths were not comprehensively reported and captured, often misclassified, lost in transition between facilities, sometimes audits were not conducted. Institutional sub-optimal technical and clinical capacity in handling obstetric emergencies was identified as a challenge in district hospitals, health centers and lower-level facilities [10].

Contributing further to the burden was the weak and inadequate clinical supervision by senior clinicians and specialists over clinical teams at all levels of service delivery. Health centers lacked the capacity to provide Basic Emergency Obstetric and Newborn Care (BEmONC), due to inadequate equipment and supplies, weak obstetric competences such as poor skills in conducting manual removal of retained placenta, and other products of conception [11]. In addition, anemia in pregnancy was common, a study conducted in Central Region identified that 47.1% of pregnant women had anemia [12], institutional access to blood and/or blood products for emergency obstetric care services was challenged by storage capacity, and non-availability of blood processing equipment and supplies. Utilization of obstetric services was hampered by socio-cultural beliefs and poverty levels in communities [13], this resulted in late attendance to the facilities with complications which sometimes resulted in institutional maternal deaths. The above identified and enumerated gaps needed to be addressed to improve maternal and perinatal health outcomes.

Quality improvement tools were employed in 14 Asian and African countries with support of German Federal Ministry for Economic Cooperation and Development. Five main methodologies were used to improve the quality of care in these countries: capacity-building and supervision, governance and regulation, systemic QI at facility level, support to infrastructures, and community support. Additionally, morbidity and mortality reviews with improvement in maternal and perinatal death surveillance were implemented [14].

A nutritional innovations project study conducted by Bigool et al. between 2014 and 2017 in two regions in Ghana used quality improvement approach that focused on improving antenatal care attendance and institutional births found positive association of the project impact on probability of antenatal care attendance by 12% and institutional births by 8% [15].

Our quality improvement project aligned with what has been described in the above studies. The goal was to reduce institutional maternal mortality ratio and perinatal mortality rate from 2020 levels by at least 15% over a 3-year period. The specific objectives were to improve facility surveillance and accountability on maternal deaths, to institute clinical supervision of maternal health services by obstetrician/gynecologists, to improve the knowledge and skill sets of doctors, midwives and other staff on life saving skills and safe motherhood protocols and to improve on access to Basic Emergency Obstetric and Newborn Care (BEmONC) services by creating Model Health Centers in the Region. The plan was to integrate these interventions into our routine health system beyond the project period. This paper aims to demonstrate the impact of implemented interventions on maternal and perinatal health outcomes in Central Region.

Methods

Study settings: The project was undertaken in Central Region, which is one of the sixteen administrative regions of Ghana (Fig. 1) with 2020 population of 2,982,658. The region is divided into 22 Metropolitan, Municipal and District Assemblies.

Fig. 1.

Map of Ghana showing the location of Central Region

The health system of the region mirrors that of the country with a five-tier healthcare delivery system: community -based health planning and services (CHPS) compounds offering close to client services in community settings, health centers that serve as first point of referrals for CHPS compound at sub-district levels, and district, regional and teaching hospitals offering primary, secondary, and tertiary level care respectively [5]. Different ranges of maternal and child health services were provided at each level; antenatal care, post-natal care, family planning services were provided at the CHPS level with occasional skilled delivery services. Health centers provide antenatal, supervised delivery, postnatal, family planning services, treatment of common infections in pregnancy, childhood immunization services and treatment of common childhood infections. Though health centers are expected to provide basic emergency maternal and newborn care (BEmONC), they however could not provide all the signal functions due to capacity challenges. The district hospitals provide comprehensive emergency obstetric and newborn care (CEmONC) while the regional and teaching hospitals provide specialist and intensive care services in addition to CEmONC. The health sector at each level enters aggregate service delivery data at the end of every month into our nationally accepted health information system known as DHIMS2. Data on various indicators can be extracted from DHIMS2, analyzed and used for various purposes such as design interventions to address service gaps, research purposes such as secondary data and others.

Study design

The study design was a before-after impact evaluation.

Baseline assessment

Baseline assessment was conducted and data on measurable indicators were collected from DHIMS2 by the authors in year 2020. The baseline assessment utilized quality improvement approach by reviewing 2020 maternal death audit reports which identified the following challenges; weak maternal mortality surveillance, sub-optimal clinical competencies in addressing maternal and perinatal care, weak or lack of clinical supervision of health staff in maternal health space by specialist and senior clinicians, lack of BEmONC services at the health centers and inadequate access to blood and blood products during emergencies. Baseline indicators were skilled-delivery rate, family planning acceptor rate, maternal mortality ratio, still birth rate and neonatal mortality rate assessed from DHIMS2 in year 2020.

Intervention design

A meeting was organized by the authors with a team of specialists to discuss the findings from the maternal death audits that generated change ideas to address identified challenges. These ideas were translated into implementable interventions referencing other QI interventions implemented in 14 Asian and African countries. These interventions included capacity-building and supervision, governance and regulation, systemic QI at facility level, support to infrastructures, community support, morbidity and mortality reviews and maternal and perinatal death surveillance [14].

Sensitization meetings were held where the quality improvement ideas and interventions were discussed with the hospitals, health directorates and other stakeholders. The authors then adopted national training guidelines such as safe motherhood and life-saving skills training manuals to plan training for staff at all levels of service delivery to address the sub-optimal clinical competencies on maternal health services. To improve surveillance on maternal death and ensure accountability for maternal deaths, a tracking system was designed to improve maternal death surveillance and reporting in all facilities known as maternal death tracker. The Region was divided into five zones and the health facilities and districts in a zone had their zonal leaders which were made up of one district director of health service, one medical superintendent of a hospital within the zone and an obstetrician gynecologist. The leaders were to coordinate maternal health services, create a telemedicine WhatsApp platform with membership of midwives, doctors, and health managers from all the facilities within the zone, ensure sharing of information and resources such as blood products, support in the management of difficult cases and facilitate referrals through information shared on the WhatsApp platforms. The teams from facilities and districts within the zones were to form a unit and organize regular meetings to discuss maternal and perinatal health issues. A regional maternal and newborn taskforce with oversight responsibility for this project was formed, the taskforce was responsible for regional enquiry into maternal and newborn deaths, coordination of competency trainings, zoning, technical support to zonal facilities, follow up on implementation of recommendations and action plans from death audit reviews as well as the evaluation of the project. A regional maternal and newborn health coordinator was identified to coordinate the work of the regional taskforce. Process indicators were selected to measure each of the interventions such as the number of midwives and clinicians trained in life saving skills, number of health centers conducting basic emergency obstetric care signal functions, percentage of audited maternal deaths entered into the district health information management system-2 (DHIMS2), impact indicators were selected based on the expected goals of the project which were skilled delivery rate, family planning acceptor rate, maternal mortality ratio, still birth rate and neonatal mortality rate assessed from DHIMS2.

Implementation phase

The implemented activities are presented below.

Training of frontline health staff

A practical and competency-based lifesaving skills training module was developed for medical officers, physician assistants, midwives, staff nurses, and community health nurses, by the regional obstetricians group based on identified service gaps. The training content was standardized, facilitated by at least two obstetricians per training session and focused mainly on practical learning sessions and case studies. All the hospitals and district health directorates within the zones planned and set dates for the training sessions, the team of the obstetricians/gynecologists moved to the hospitals and the districts health directorates to train staff. The district health directorate gathered staff such as midwives, community health nurses and others from the health centers and CHPS compounds to a venue at the district level for the training. In the hospitals, the training took place there and staff involved were doctors, nurses, midwives and others. The training duration was for two days. The training was funded by district health directorates, hospitals, and polyclinics managements from internally generated funds. Training agenda is presented in the appendix.

Supportive supervisory visits

The region was zoned into five administrative locations based on their geography and referral patterns by the authors. Due to inadequate number of obstetricians, each zone had at least two obstetricians drawn from the teaching and other hospitals, assigned for supportive supervisory visits, maternal and perinatal mortality audits, and capacity building workshops. The specialists also assisted with written presentations, discussions, and management of obstetric emergencies on zonal social media platforms. Zonal social media mainly through WhatsApp platforms discussed complicated cases and lower-level facilities received management support from the specialists on the platform, arranged for referrals and shared resources such as blood products.

Figure 2 gives a summary of the study starting from identified challenges from assessment, implemented interventions, output and outcomes with the linkages.

Fig. 2.

Diagrammatic description of the project showing the various stages of implementation

Surveillance on maternal deaths

the authors adopted a maternal death tracker (Fig. 3) for daily maternal deaths reporting. Facilities reported maternal deaths verbally to the regional maternal newborn health coordinator within twenty (24) hours of occurrence. The information was then recorded on the maternal death tracker (i.e., day of the month, number of maternal deaths and facility recording the death) for dissemination on managerial platforms in the region for visibility on the burden and initiation of maternal death audit processes. The days of each month were color coded to distinguish dates with deaths and no deaths. A red-coded date indicated maternal deaths, whilst a green-coded date showed no deaths (Fig. 3).

Fig. 3.

Description of the maternal death tracker for the project

Maternal and perinatal death audit reviews

A three-tier mortality audit system was adopted for auditing maternal, stillbirths and newborn deaths: A facility level audit was conducted within seven days after the occurrence of deaths, with the zonal obstetricians in attendance, to identify issues with clinical care processes, available competencies, and supplies. The second tier is a quarterly zonal death audit meeting which seeks to find common issues within zones that accounted for deaths, discuss and identify corrective action. This meeting also discussed implementation status of audit recommendations for the facilities that reported maternal deaths during the quarter. A regional death enquiry meeting that found issues and gaps for regional level actions was the third tier. The facility audit team was as outlined in the Ghana’s maternal and perinatal mortality audit guidelines. The zonal audit team constituted obstetricians, medical superintendents of hospitals and polyclinics, district directors, public health nurses, deputy directors of nursing services, clinical coordinators, medical officers, and maternity ward in-charges. The regional director of health services chaired the regional enquiry meetings every month, with members of the regional maternal and newborn task force team comprising (deputy director public health, deputy director clinical care, chief nursing and midwifery officer, chief pharmacist, maternal and newborn coordinator, and an obstetrician from teaching hospital in attendance.

Regional BEmONC facilities

The authors with support of district managers identified 48 health centers in all the districts as model health centers that operated 24-hour services. These health centers serve as relatively resourced first point referral facilities for lower-level peripheral facilities. District directors coordinated the functionality of these facilities through staff re-deployments, clinical skills set training through attachments to hospitals for identified clinical competency gaps, and stakeholder engagements for supply of needed equipment and supplies. These model health centers ran 24-hour service and had relatively more staff such as midwives than the other health centers. The staff in the model health centers were trained by the obstetricians to provide at least five of the seven BEmONC signal functions, namely intravenous administration of uterotonics, parenteral administration of anticonvulsants and antibiotics, manual removal of placenta and neonatal resuscitation. These facilities were stocked with logistics such as magnesium sulphate, oxytocin and intravenous infusions by health managers. District medical superintendents or heads of hospitals provided clinical supervision over these facilities.

Stakeholder engagements

The regional health directorate and the authors organized annual stakeholders meeting on the burden of maternal and newborn deaths. The stakeholders included traditional leaders, queen-mothers, religious leaders, media, and local government leaders. The meetings provided community perspectives on mortality determinants, and support to address critical community factors.

Data capture and management

No individual data was collected, and the data collected was secondary data from the facilities. Trained health staff captured aggregated facility level data on maternal health services and outcomes from patient folders, antenatal care, delivery, post-natal care, and admission registers and entered them into the Ghana Health Service DHIMS2 monthly. Districts provided verifiable and routine written reports on the implementation of zonal, district and sub-district level activities including zonal quarterly meetings. Additionally, training coordinators and/or facilitators provided written reports using standard reporting templates that showed the number of staff trained and their cadres which were analyzed and synthesized together. Regional enquiry meetings were recorded by the regional maternal health coordinator. Supervision and monitoring of model health centers by authors assessed the functionality of 24 h service provision, availability of selected essential commodities and number of staff including midwives as well as capacity to perform at least five of the signal functions. Their assessment report was used to determine the functionality of the model health centers. Data collection tool was prepared by authors to extract data from DHIMS2 and analyzed. Data analyzers known as Pivot table in DHIMS2 were used to generate the indicator results. In estimating maternal mortality ratio (MMR), the indicator was typed on the Pivot table, then the organizational unit, which is Central region selected, command was given and it autogenerated MMR by dividing the total maternal deaths by the total live births and multiplying by 100,000. Hence, the Pivot table in DHIMS2 was used to compute the impact indicators such as the family planning acceptor rate, skilled birth rate, neonatal mortality rate and still birth rate yearly for the three years. The family planning acceptor rate was computed by dividing total family planning acceptors by women in fertility age, neonatal mortality rate was computed by dividing total neonatal deaths by total live births multiplied by 1000 and still birth rate was the total still births divided by total live births multiplied by 1000 as computed by DHIMS2. This was how the baseline indicators as well as the three yearly indicators were computed. The total number of deaths and total deaths audited were also extracted from DHIMS2. Continuous variables had summary tables of frequencies, totals, and percentages presented. For categorical data, summary tables, proportions and percentages were presented. Bar charts and line graphs were also used to depict the data. A two-sample t-test with equal variances was used to compare and determine associations, confidence intervals and statistical significance. P-value of less than 0.05 (p-< 0.05) denotes statistical significance.

Ethical approvals

The study was not based on individual data collection, so no consent was needed. Data used for the study was secondary data assessed from our health information system known as DHIMS2. Hospitals and health facilities routinely enter health data into DHIMS2. This report involves real routine public health and retrospective work and did not require ethical approval. Again, appropriate administrative permissions were obtained from the Regional Health Directorate.

Results

Capacity Building of staff to improve clinical competence for maternal health care.

Over the course of the project, a total of 4,800 health staff were trained. Midwives were the majority (74.7%) of all health staff trained (Table 1).

Table 1.

Cadre of Staff trained in Life Saving Skills, Central Region, 2020–2023

Cadre	2020	2021 (%)	2022 (%)	2023 (%)	Total (%)
Midwives	0	1091 (79.5)	1172(75.1)	1320 (70.7)	3,583 (74.7)
Medical Officers	0	44 (3.2)	64 (4.1)	62 (3.3)	170 (3.5)
Physician Assistants	0	32 (2.3)	84 (5.4)	97 (5.2)	213 (4.4)
Nurses	0	173 (12.6)	222 (14.2)	273 (14.6)	668 (13.9)
Others	0	32 (2.3)	19 (1.2)	115 (6.2)	166 (3.5)
Total	0	1372 (100.0)	1561(100.0)	1867 (100.0)	4,800(100.0)

Surveillance, audits and model health centers

Facility maternal and perinatal death audits meetings and implementation of recommendations improved over the course of the project; 78% facility audits were done in 2020 and 100% in 2021, 2022 and 2023. Though no quarterly zonal audits were done in 2020, 100% was done in 2021, 2022 and 2023. Again, no regional enquiry meetings were held in 2020 but 18, 15 and 10 enquiries were made in 2021, 2022 and 2023, respectively. Number of Model health centers improved from 5 in 2020 to 48 in 2023 (Table 2) through resourcing and equipping 48 health centers to improve management of maternal health services at the lower level and ensure good and prompt referrals to the next level.

Table 2.

Audits and Model Health Centers, Central Region, 2020–2023

Indicator	2020 (baseline)	2021	2022	2023
% Maternal deaths audited	78	100	100	100
% Quarterly zonal Audit meetings held	0	100	100	100
No. of Regional death enquiry	0	18	15	10
No. of Model Health Centers (BEmONC)	5	26	48	48

Intermediate outcomes

Birth attendance improved from 61.6% in 2020 to 63.9% in 2023 (an.

improvement of 2.3%). Family planning acceptor rate improved from 28.5% in 2020 to 45.1%.

by 2023 (i.e., an increase of 58%) Fig. 4.

Fig. 4.

Trend in Family planning acceptor rate by year

Outcome indicators

Over the period of the project, the total live births increased by 12.3% from 68,705 in 2020 to 77,111 by 2023 (Table 3). Institutional maternal mortality ratio declined from 75 (109.2/100,000LB) in 2020 to 56 (72.6/100,000LB) constituting 25.3% decline (Fig. 5). Neonatal deaths declined from 453 (6.6%) in 2020 to 248 (3.2%) in 2023, being 45.0% decline.

Table 3.

Livebirths, stillbirths, maternal deaths, and family planning outcome

Indicator	2019	2020 baseline	2021	2022	2023	Average
Total Live Births (N)	66,513	68,705	75,674	76,139	77,111	72,828.4
Total Institutional Maternal Deaths (N)	70	75	79	69	56	69.8
Institutional Maternal Mortality Ratio	105.2	109.2	104.4	90.6	72.6	96.4
Total Institutional Still Births (N)	888	925	975	858	774	884
Institutional Still Birth Rate (%)	13.17	13.5	12.9	11.3	9.9	12.1
Total Neonatal Deaths (N)	498	453	520	364	248	416.6
Neonatal Mortality (%)	7.5	6.6	6.9	4.8	3.2	5.8
Family Planning New Acceptors (N)	79,928	85,239	106,749	129,340	141,157	108,482
Family Planning Acceptor Rate (%)	32.1	28.5	35.1	40.8	45.1	43.8
Skilled Birth Rate (%)	58.9	61.6	65.4	64.8	63.9	62.9

Fig. 5.

Trend in institutional maternal mortality ratio by years

Table 4 shows the statistically significance level of the outcome indicators. Overall, all the four major indicators reached statistically significant difference between the years 2020 to 2023.

Table 4.

Significance of the Outcome Indicators

Indicator	2020	2022	2023	Difference	95% C. I	p-values
Institutional Maternal Mortality Ratio	109.2	90.6	72.6	36.6	5.3–6.1	0.02
Institutional Still Birth Rate (%)	13.3	11.1	9.9	3.4	2.3–4.5	0.0006
Neonatal Mortality (%)	6.6	4.8	3.2	3.4	2.6–4.1	0.0007
Family Planning Acceptor Rate (%)	28.5	40.8	45.1	-16.6	16.3–16.8	0.0000

Discussion

To further improve on maternal health outcomes and specifically reduce maternal deaths in the Central Region of Ghana, the study used quality improvement tools and specific interventions to help reduce institutional maternal mortality ratio and perinatal mortality. The quality improvement tools systematically improved care by standardizing processes and structure that helped reduced variation and achieve predictable results as part of measures to improve maternal outcomes in the healthcare systems (18). Again, the interventions implemented are known and proven to improve maternal health services and contribute to reducing maternal deaths [16].

The project emphasized the role of health leadership in reducing maternal and perinatal deaths and improving maternal and child health services. Key components of the project included surveillance on maternal deaths and capacity building for health staff on safe motherhood and life-saving skills. Others were supervision of maternal and perinatal health services by specialists using telemedicine as well as ensuring accountability for maternal and perinatal deaths through audits at all levels. The implemented interventions were funded from internally generated funds and not donor funded, this has an implication for sustainability of these interventions and can be replicated by other regions in the country. The interventions were implemented by the institutions in the region every year for the three years, this was to ensure ownership by health facilities and sustainability beyond the project period. An earlier study conducted in Eastern Region of Ghana on maternal death audit review showed that supervision of maternal health services by senior clinicians and specialist improved on the quality of care received by women [16] and hence contributed to reducing maternal deaths consistent with the findings of this project.

Over the course of the project, maternal deaths and neonatal deaths declined by 25.3% and 45.0%, respectively. Institutional MMR for 2023 was 72.6/100,000LB compared to the national figure of 108.6/100,000LB, still birth rate was 9.9 compared to the national figure of 10.2 and neonatal mortality rate was 3.4/1000LB compared to national figure of 5.13/1000LB [17]. These improvements might be attributable to the implementation of the various strategies in the project, key among them is the annual capacity building on safe motherhood and lifesaving skills of health staff at all levels. This might have led to early identification and referral of women and neonates with problems for prompt management as shown in a study conducted in Malawi [18] that recommended need-based knowledge and skills update of midwives in reducing maternal deaths.

Skilled attendance at birth improved through staff postings and re-deployments to address equity gaps. The maternal death tracker platform aided with local level facility management actions to address deficiencies with obstetric emergency care processes and services. Facility mortality audits aided in the identification of obstetric competency gaps, from peripheral levels to district hospitals. These identified emergency obstetric care management deficiencies, together with findings from facility assessments and specialists’ facility supervisory visits, informed the content of our life saving skills trainings as found in other studies [19].

Zonal mortality audits and reviews improved networking among zonal facilities, obstetric pre-referral communications and the sharing of resources to improve maternal health services. Network of care for maternal and child health services is known to improve quality of care, continuity of care, and maternal and newborn outcomes. This created intentional connections between people and services, strengthened the functional aspects of health systems and incorporated relational aspects of health delivery. The network supports quality health service delivery and strengthens referral systems within the local context, thus promoting continuity of care [20]. Sharing resources like blood within the zonal networks improved obstetric hemorrhage related case outcomes. Accountability for maternal and neonatal deaths have been shown to improve health outcomes [16, 18] and this was employed at all levels of the health system. Regional level enquiries improved facility level accountability, stakeholder engagements, human resource and equipment allocations to various districts and facilities improved care for women and neonates.

Family planning acceptor rates improved significantly by 58% over the duration of the project, through improved provider training, commodity availability, and local level service innovations to promote demand generation. However, the improvement in family planning coverage did not translate into declines in our annual total deliveries reported. Other studies conducted confirmed that fertility is shaped by multiple interacting determinants such as contraception, marriage, postpartum insusceptibility and sexual exposure among others, hence an increase in family planning coverage can be offset by reductions in postpartum abstinence or breastfeeding leading to higher fertility. High family planning coverage may sometime reflect birth spacing rather than a desire to stop childbearing [21, 22].

Conclusion

Several lessons were learnt from this project and key among them is the role leadership can play at all levels to drive change and implement innovative strategies to improve maternal and perinatal health outcomes. Leveraging technologies such as telemedicine through WhatsApp platforms and networking of facilities to share knowledge, resources and strengthen referral systems can improve quality of care for maternal and child health services, resulting in improved outcomes. This has policy implication of building capacity of leadership at all levels on maternal and child health to lead in implementing innovative strategies to reduce maternal and perinatal deaths. The policy of forming networks among health facilities and improving relationships among service providers need to be explored to improve health outcomes.

Study limitations

The limitation of using DHIMS2 data includes facility coverage issues especially for some private facilities, however district health directorates collaborated with the private sectors in their areas of operations to capture their data for entry into DHIMS2. As in the case of many secondary data sources, there may be data quality issues such as data errors, data incompleteness and late reporting. Regular data verification and validation exercises were conducted to address these data quality issues.

Authors’ contributions

A.O.S was involved in conceptualization of research, implementation, data analysis manuscript preparation and review, G.S was involved in research implementation, manuscript preparation and review, A.A was involved in research implementation and manuscript review, V.M was involved in research implementation, data analysis, manuscript preparation and manuscript review, E.K.A was involved in conceptualization, implementation and manuscript preparation and A.R.O was involved in data analysis, manuscript preparation and manuscript review.

Funding

The study was funded by the Central Regional Health Administration, Ghana.

Data availability

The data that support the findings of this study are available from the authors but restrictions apply to the availability of these data, which were used under license from the Ghana health Service dhims2 which is the national health data repository system. Data are, however, available from the authors upon reasonable request and with permission from the Ghana Health Service. Example from: https:// www.dhims.chimgh.org. You can contact the author at akos_owususarpong@yahoo.com.

Declarations

Ethics approval and consent to participate

Not applicable

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.