Facilitators and barriers to the implementation of maternal and perinatal death surveillance and response in Ethiopia: a systematic review

Abstract

Introduction

Ethiopia was one of the pioneer countries to implement the World Health Organization’s Maternal and Perinatal Death Surveillance and Response (MPDSR) system to accelerate the reduction in maternal and perinatal mortality. However, there has been no systematic evaluation of its implementation in Ethiopia and that’s why we conducted this systematic review.

Methods

A systematic review was conducted to synthesize the evidence on coverage, facilitators or barriers to MPDSR implementation in Ethiopia. We searched PubMed, Embase, Web of Science, PubMed Central, and Google Scholar using relevant key terms. In addition, Ethiopian Public Health Institute websites searched for additional data. Articles published before 2013 excluded from this review. The methodological quality of the studies assessed using the Joanna Briggs Institute’s quality appraisal tool. For quantitative studies, descriptive analysis conducted; thematic synthesis used for qualitative studies.

Results

From twenty studies included, 12 only reported maternal death reviews while eight included maternal and perinatal death reviews. During the reporting period, the coverage of maternal and perinatal deaths remained less than 22.1% and 12.1% of the expected deaths respectively. Reported facilitators were community involvement, sufficient capacity building, and supportive supervision. Reported barriers were lack of conducive learning environment, fear of blame and litigation, lack of financial resources, high staff turnover, and defensive attitudes and practices.

Conclusions

Despite all efforts, the uptake of MPDSR has been low. Addressing identified barriers and utilizing identified facilitators essential for optimising MPDSR implementation in Ethiopia.

Registration

PROSPERO Registration Number: CRD42022315199.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12978-025-02115-w.

Plain language summary

Ethiopia was one of the pioneer countries to adopt the World Health Organization’s Maternal and Perinatal Death Surveillance and Response (MPDSR) system to accelerate the reduction in maternal and perinatal mortality. Although several studies on MPDSR implementation and/or barriers and facilitators to MPDSR in Ethiopia have been conducted, there is no comprehensive assessment of data to inform decision-making on the future of MPDSR. This review was conducted according to the Preferred Reporting Items for Systematic reviews and Meta-Analysis (PRISMA) guidelines. The review protocol was registered in PROSPERO (CRD42022315199). The PubMed (Medline), Embase (OVID), Web of Science (Core Collection), CINAHL, Cochrane Library (incl. CENTRAL), Google Scholar, PubMed Central, Academic Search Premier, Preprint Citation Index and Ethiopian Public Health Institute websites were used for searching. All identified articles exported to Covidence, and duplicates removed. Two reviewers (MY and CS) independently screened and reviewed the studies and reports against the inclusion criteria, and independently extracted information using predetermined inclusion criteria. The descriptive analysis conducted for quantitative studies while thematic synthesis done for qualitative studies and presented the main (sub-) themes in text and quotes. Over the reporting period, the coverage of maternal and perinatal deaths remained less than a quarter approximately of the expected deaths. Reported facilitators were community involvement, sufficient capacity building, and supportive supervision. Reported barriers were lack of conducive learning environment, fear of blame and litigation, lack of financial resources, high staff turnover, and defensive attitudes and practices. Therefore, addressing the identified barriers and utilizing identified facilitators paramount to optimise the MPDSR implementation in Ethiopia.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12978-025-02115-w.

Introduction

Pregnancy-related mortality remains a public health problem across the world, one of the major reasons being a need for higher quality maternity care [1]. Globally, there were 287,000 maternal deaths (800 per day), mainly from preventable causes, in 2020 [2]. In addition, there were 2.3 million neonatal deaths (6,300 per day) in 2022 and 1.9 million stillbirths in 2021 [3]. Sub-Saharan Africa alone accounts for over half of all maternal deaths, with a combined maternal mortality ratio of 545 per 100,000 live births [4]. Similarly, neonatal mortality and stillbirths remain high in sub-Saharan Africa, with 27 neonatal deaths per 1000 live births and 17 stillbirths per 1000 live births [3].

The situation in Ethiopia is not different. Despite a reduction in the maternal mortality ratio from 635 to 267 per 100,000 live births between 2010 and 2020 [2], in 2020, Ethiopia had over 10,000 women who died of pregnancy related causes [2, 5]. In addition, the perinatal mortality rate was 33 deaths per 1000 live births [6].

Beyond counting the numbers, strategies to reduce maternal and perinatal deaths require a systematic, qualitative approach to identify biomedical causes, determinants, and contributing factors. This may help design local context-tailored interventions and improve quality of care, whereby future similar deaths could be prevented [7]. In 2013, to accelerate the mortality reduction, the World Health Organization (WHO) introduced a maternal death surveillance and response (MDSR) system in countries with a high death burden —the 2017 update included perinatal deaths to become the maternal and perinatal death surveillance and response (MPDSR) system [8, 9].

Ethiopia was one of the pioneer countries to implement MDSR through piloting this approach in four larger regions (Oromia, Amhara, Southern Nations, Nationalities and People’s Region (SNNPR), and Tigray) in 2014, with a national rollout in late 2015. MDSR was integrated into national public health emergency management (PHEM) under the Integrated Disease Surveillance and Response system [10]. In Ethiopia, MPDSR is an integral part of quality of care improvement efforts to reduce maternal deaths, as well as preventable stillbirths and early neonatal deaths. Although several studies on MPDSR implementation and/or barriers and facilitators to MPDSR in Ethiopia have been conducted, there is no complete assessment of data to inform decision-making on the future of MPDSR [10–16]. Therefore, the objective of this systematic review was to provide comprehensive evidence on the coverage of MPDSR and its facilitators and barriers in Ethiopia.

Research questions

What is the coverage of MPDSR in Ethiopia?

What are the barriers and facilitators influencing MPDSR implementation in Ethiopia?

Methods

This review was conducted according to the Preferred Reporting Items for Systematic reviews and Meta-Analysis (PRISMA) guidelines [17]. The review protocol was registered in PROSPERO (CRD42022315199).

Inclusion criteria

All studies, including published and unpublished reports on MPDSR and/or facilitators or barriers were included. To provide a complete overview of the evidence on the topic, we included quantitative and qualitative studies and annual reports with relevant data on the condition surrounding maternal and perinatal deaths in Ethiopia. Those without relevant data on the coverage of MPDSR, facilitators or barriers were excluded. To align with the global release of MDSR, publications since 2013 were included. The MPDSR system focused on analysing causes of death and implementing respective targeted interventions. Coverage was defined as the number of deaths captured and notified or reviewed against the expected deaths according to the national demographic health survey. In Ethiopia, the MPDSR system is designed to capture and analyse all maternal and perinatal deaths occurring at a health facility and the community level. The details about the review and reporting procedures found in national technical guidance for MPDSR [18].

Information sources and search strategy

A systematic search of PubMed (Medline), Embase (OVID), Web of Science (Core Collection), CINAHL, Cochrane Library (incl. CENTRAL), Google Scholar, PubMed Central, Academic Search Premier and Preprint Citation Index was conducted on 29 September 2023 and was re-run 17 February 2025 using a combination of key terms developed in consultation with a medical information specialist librarian of Leiden University Medical Center, the Netherlands (JS). We used a combination of ‘maternal and perinatal death surveillance and response’ and ‘Ethiopia’ for developing the search terms appropriate to the respective databases. The search was conducted using the following terms: maternal mortality, perinatal deaths, death reviews, death audits, surveillance, pregnancy, obstetrics, Africa, and Ethiopia; the full search strategy is provided in Appendix S1. In addition, reference lists of the included studies as well as the official website of the Ministry of Health and the Ethiopian Public Health Institute were searched for additional relevant studies or reports.

Study selection

All identified articles exported to Covidence (www.covidence.org), and duplicates removed. Two reviewers (MY and CS) independently screened titles and abstracts. All potentially relevant articles or articles that could not be excluded based on the abstract only retained for full-text review. MY and CS independently reviewed the full text against the inclusion criteria. Differences between both reviewers during the full text review resolved by discussion with a senior reviewer (AKT).

Assessment of methodological quality and data extraction

Two reviewers (MY and CS) assessed the methodological quality of the studies using the Joanna Briggs Institute (JBI) critical appraisal tool for assessing qualitative studies which consists of ten questions; all seven qualitative papers assessed accordingly. For quantitative papers and MPDSR reports, we used the JBI critical appraisal tool for cross-sectional studies using the checklist of eight questions and all studies and reports assessed accordingly. Details are in Table S1 and Table S2.

Both reviewers independently extracted information on study design, study setting, data collection period, study participants, maternal deaths, perinatal deaths, relevant elements related to MPDSR implementation, facilitators and barriers on a previously developed template in the systematic review data repository (srdr.ahrq.gov) platform [19]. Conflicts during data extraction resolved by repeated discussion among reviewers until unanimity reached.

Data analysis

The coverage of maternal and perinatal deaths analysed using descriptive statistics and are presented in tables and graphs. For the barriers and facilitators identified in qualitative studies, we used a conceptual thematic analysis approach [20] to categorise facilitators and barriers into themes and sub-themes after reading the major themes highlighted in the papers and discussed among peers (MY and CS). We then organized and analysed these themes using thematic synthesis, and presented the main (sub-) themes in text and quotes.

Results

Search outcome

The initial search generated 2829 records. After deduplication, 1387 records were imported into Covidence for screening using title and abstract review. Of these, 1337 were excluded and 50 records were retrieved for full-text review. After full text review, 30 studies were excluded. The main (86.7%) reasons for exclusion were lack of any data on MPDSR coverage or facilitators or barriers. Twenty papers met the inclusion criteria and were included in the review (Fig. 1).

Fig. 1.

PRISMA flow chart of included studies

Study characteristics

From the 20 studies, eleven were conducted at the national level. Twelve focused solely on maternal death reviews, while eight included reviews of perinatal deaths in addition to maternal deaths. Seven were national reports, and seven were qualitative studies involving key informant interviews or focus group discussions (Table 1).

Table 1.

Characteristics of MPDSR studies in Ethiopia [n = 19]

Authors and year	Region	Scope	Data collection methods	Sample unit
Abebe et al. 2017 [10]	Oromia, Amhara, SNNP, Tigray	Maternal	In depth interviews	Health workers
Ayele et al., 2019 [14]	Tigray	Maternal and perinatal	Focus group discussions, [in-depth] interviews, record review	Health facility and health workers
Bogale et al., 2020 [21]	Addis Ababa	Maternal	Checklist based interviews and record review	Health facility and health workers
Cetin et al., 2022 [16]	National	Maternal and perinatal	In depth interviews	Health workers
Endris et al., 2023 [22]	National	Maternal	Record review, checklist based interviews	Health facility and health workers
EPHI 2016 [23]	National	Maternal	Maternal death reporting	Maternal deaths
EPHI 2017a [24]	National	Maternal	Maternal death reporting	Maternal deaths
EPHI 2017b [25]	National	Maternal	Maternal death reporting	Maternal deaths
EPHI 2018 [26]	National	Maternal	Maternal death reporting	Maternal deaths
EPHI 2020 [27]	National	Maternal and perinatal	Maternal [perinatal] death reporting	Maternal and perinatal deaths
EPHI 2022 [28]	National	Maternal and perinatal	Maternal [perinatal] death reporting	Maternal and perinatal deaths
EPHI 2024 [29]	National	Maternal and perinatal	Maternal [perinatal] death reporting	Maternal and perinatal deaths
Hadush et al., 2020 [30]	National	Maternal	Maternal [perinatal] death reporting, Record review	Maternal deaths and health workers
Melberg et al., 2019 [11]	Addis Ababa and Oromia	Maternal	In depth interviews	Health workers
Melberg et al., 2020 [15]	Addis Ababa and Oromia	Maternal and perinatal	In depth interviews	Health workers
Tariku et al., 2021 [31]	Amhara	Maternal	Self-administered	Health workers
Tesfaye T. et al., 2015 [13]	Dire Dawa	Maternal	In depth interviews, record review	Health workers
Tesfay N. et al., 2024 [32]	National	Maternal	Survey	Health facility
Tura AK et al., 2020 [12]	Harari and Oromia	Maternal	Commentary	Health workers
Zemen E. et al., 2024 [33]	Oromia	Maternal and perinatal	Record review, In-depth interviews	Health facility and health workers

SNNP Southern Nations Nationalities and Peoples, HFs Health Facilities, HWs Health Workers, HCs Health Centres, MDs Maternal Deaths, PD Perinatal Deaths, MPDs Maternal and Perinatal Deaths

Assessment of methodological quality of included studies

Based on the methodological quality assessment of the studies using the Joanna Briggs Institute (JBI) critical appraisal tool for assessing qualitative studies, all seven qualitative papers scored above 60% and passed for inclusion (Table S1). The quantitative papers and MPDSR reports were also scored above 75% and passed for inclusion (Table S2).

Coverage of deaths in the MPDSR system

While the first four national reports focused on maternal deaths only [23–26], the latest three national reports included both maternal and perinatal death reviews [27–29]. The coverage of notifications and reviews of maternal death started from 0% to 1% in 2014, and increased until these proportions reached 15% and 13% in 2018 [26], after which these proportions declined again to 13% and 4% in 2020. However, after 2021, the coverage of notifications and reviews increased progressively resulted these proportions reached 22% and 16% (Fig. 2).

Fig. 2.

Trends of maternal deaths notification and review coverage in Ethiopia from 2014–2024

The coverage of notifications of perinatal deaths increased from 3.8% in 2021 to 12.2% in 2024. Perinatal death reviews increased from 0.1% in 2018 to 5% in 2024 [29]. The most recent MPDSR national report of 2024 revealed that from an estimated 115,957 perinatal deaths, only 14,155 (12.2%) were notified weekly and 5,671 (5.3%) underwent case-based reviews, implying that only 40.1% of the notified deaths underwent such a detailed review [29]. Reports highlighted substantial regional variations in the coverage of death notifications and reviews [23–29]. The highest review coverage (74% in 2016, 33% in 2017, and 63% in 2018) was seen in Dire Dawa City Administration in the first three reports and in Harari region (63% in 2020, 52% in 2021 and 68.7% in 2024) in the last three national MPDSR reports (Fig. 3).

Fig. 3.

Trends of maternal death review coverage by regions in Ethiopia from 2016 to 2024

Similarly, the highest notification and review coverage was reported in Harari region, with 100% of the estimated maternal deaths being notified and 68.7% reviewed in 2024. On the other hand, 20 out of 127 zones and two town administrations did not report any deaths at all in 2021 [28].

Facilitators of MPDSR implementation

All seven qualitative studies [10–16] reported facilitators of MPDSR implementation (Table 2). These were grouped into three categories: political/leadership level, health system level, and health worker level. Abebe et al. [10] and Tura et al. [12] reported that strong political support and active communication and coordination by chairpersons and stakeholder collaboration in implementing the responses were facilitators [10, 12–14]. Reported health system level facilitators were: integration of MPDSR into the PHEM system, community involvement in MPDSR, sufficient capacity building among committee members and the presence of supportive supervision to health facilities [10, 13, 14]. Health worker-related facilitators were: active participation and commitment by staff as committee members during death review meetings and health workers coping strategy to the burden of audits by many people [10, 12, 15, 16].

Table 2.

Facilitators for MPDSR implementation in Ethiopia 2024

Main Theme	Sub-theme	Studies
Political support and facility leadership level factors	Strong political commitment	Abebe et al. [10] and Tura AK et al. [12]
	Active communication and coordination of MPDSR	Abebe et al. [10], Ayele et al. [14], Tura AK et al. [12], and Tesfaye T. et al. [13]
	Stakeholders collaboration	Abebe et al. [10]
Health system level factors	Integration of MPDSR into the PHEM system*	Abebe et al. [10] and Tura AK et al. [12]
	Community involvement in MPDSR	Ayele et al. [14] and Tesfaye T. et al. [13]
	Sufficient capacity building	Abebe et al. [10], Ayele et al. [14], Melberg et al., 2019 [11], and Tesfaye T. et al. [13]
	Regular supportive supervision	Abebe et al. [10], Melberg et al., 2019 [11], and Tesfaye T. et al. [13]
Health worker level factors	Active participation	Cetin et al. [16] and Melberg et al. [15]
	Coping burden of audits	Cetin et al. [16]
	Motivation and commitment	Abebe et al. [10], Cetin et al. [16] and Tura AK et al. [12]

*PHEM system feed in to MPDSR system in reporting and handling data

Barriers to MPDSR implementation

Barriers to MPDSR implementation were reported in all seven qualitative studies [10–16] (Table 3). Barriers were summarized into three main themes: leadership and governance level, health system level, and health worker level. Politicization of maternal and perinatal deaths, role confusion between PHEM, which is under the Ethiopian Public Health Institute and maternal and child health (MCH) coordinators, situated under the Ministry of Health in monitoring the response implementation, lack of adequate legal frameworks to prevent punitive action, and poor handover system of the tasks in health facilities were leadership and governance barriers [10, 11, 14]. Health system level barriers were: a lack of conducive learning environment, lack of financial resources for implementing recommendations, high staff turnover, competing priorities (given outbreaks of COVID-19 and measles during the study period), ignoring death on arrival for not counting deaths, lack of appropriate feedback to lower health facilities, and failure to implement recommendations [10, 12–14, 16]. At the health worker level, reported barriers of fear of blame and litigation, gaps in knowledge and skills to perform death reviews, lack of motivation and commitment, defensive attitudes and practices, workload of focal persons, and poor attendance of review meetings [11, 12, 14–16].

Table 3.

Barriers to MPDSR implementation in Ethiopia 2024

Main Theme	Sub-theme	studies
Leadership and Governance level factors	Politicization of maternal and perinatal deaths	Tura AK et al. [12] and Melberg et al., 2019 [11]
	Role confusion	Abebe et al. [10]
	Lack of adequate legal framework	Melberg et al. [15] and Tura AK et al. [12]
Health system level factors	Lack of conducive learning environment	Tura AK et al. [12] and Cetin et al. [16]
	Lack of financial resource	Tesfaye T. et al. [13], Abebe et al. [10], and Ayele et al. [14]
	High staff turnover	Tesfaye T. et al. [13] and Abebe et al. [10]
	Poor handover system	Abebe et al. [10]
	Competing priorities	Abebe et al. [10]
	Ignoring death on arrival	Tura AK et al. [12]
	Lack of appropriate feedback to lower health facilities	Tura AK et al. [12]
	Failure to implement recommendations or changes	Tariku MK et al. [31]
Health worker level factors	Fear of blame and litigation	Abebe et al. [10], Ayele et al. [14], Cetin et al. [16], Melberg et al., 2019 [11], Melberg et al. [15], and Tura AK. et al. [12],
	Knowledge and skills gap	Tariku MK et al. [31]
	Defensive attitude and practice	Ayele et al. [14], Melberg et al., 2019 [11], and Tura AK et al. [12]
	Lack of the safety of health workers	Cetin et al. [16], Ayele et al. [14], Tura AK et al. [12], and Melberg et al., 2020 [15]
	Workload	Cetin et al. [16] and Melberg et al., 2019 [11]
	Poor attendance	Tura AK et al. [12]

Discussion

Main findings

This systematic review synthesized the evidence from 13 studies and 7 national reports on the implementation of MPDSR and related facilitators and barriers in Ethiopia, an early adopter of MPDSR. While MPDSR was in place since 2014, national reports covered less than 22% of the expected deaths, indicating that the MPDSR system has yet to achieve its primary objective of notifying and reviewing all maternal and perinatal deaths [18, 28]. Identified facilitators and barriers centred around three themes of political leadership and governance, health system, and health worker level factors.

The strong political commitment in prioritizing and supporting the MPDSR system, aligning it with the national PHEM system to enhance surveillance of reportable deaths played a crucial role in introducing the system across Ethiopia. However, the challenges in the notification and reporting of cases that arose from role confusion between PHEM and MCH as well as the politicizing of maternal deaths with commitment slogans such as “no mother should die while giving birth” influenced health workers behaviours [10, 11]. Thus, political commitments should focus on fostering an inclusive environment for effective death reviews and responses [10, 12, 16].

MPDSR implementation requires a conducive learning environment where all feel safe in discussing cases without fear, in order to facilitate discourse concerning the circumstances surrounding deaths [18, 34, 35]. However, Ethiopian health workers reported fear of blame, litigation and punishment in the event of disclosing the identified gaps during death reviews, particularly in cases where care was delayed [10–13, 15, 16, 36]. As the MPDSR system is also used by cadre of health system and local leadership to identify health workers who committed errors for discipline or censure, health workers are focused on protecting their own interests by avoiding reporting of maternal deaths to absolve themselves from political and personal accountability [11]. There is contradiction between the experiences of Ethiopian health workers and the purpose of MPDSR to facilitate learning. Understanding how the current blame culture can be minimized is needed to improve the uptake and effectiveness of the MPDSR system.

The regular capacity-building training sessions within MPDSR are important to support MPDSR committee members and leaders’ professional development [11, 13, 14, 18]. They can help fill knowledge or skills gaps and minimize the blame culture that might occur due to high staff turnover and poor handover system in health facilities [13]. The MPDSR related capacity building programmes in place should ensure that new personnel are adequately trained in the procedures and make sure that death review environments are free from blame and shame culture.

Absence of a designated budget to support the implementation of the entire MPDSR process including capacity building of MPDSR committee impedes the effective implementation of MPDSR. Minimal resources were key health system barriers, specifically in implementing recommendations identified during review [10, 13, 14]. Moreover, most of the responses require multi-sectoral collaborations including the community representatives to implement local context interventions [13]. Failure to review deaths was reported as well and therefore appropriate feedback and lessons learned were not provided for these cases [12]. There is no formal mechanism to consolidate this information and assess the efficacy of the lessons learned from the death reviews. This is primarily due to the absence of a designated reporting format as many health facilities only provided feedback to health offices and lower-level health facilities [18].

Strengths and limitations

To ensure that the results are representative, this review encompassed all studies and reports pertaining to the implementation of MPDSR in Ethiopia. Fear of blame, litigation, and a defensive attitude can all contribute to underreporting of maternal and perinatal deaths and reviews, which could have impeded the data included in this review. As the national reports have not reported the status of response implementation, identifying reports related to response implementation at national level was difficult in this review. This review also included national reports and an unpublished thesis, none of which were peer reviewed and therefore, data quality may affect the results.

Interpretation

Despite numerous efforts to implement MPDSR in Ethiopia, the programme has captured, notified, or reviewed less than 22% of expected maternal and perinatal deaths. This indicated that the poor performance of the system due to the potential barriers. The MPDSR must go beyond having national guidelines or strategies to reduce maternal and perinatal mortality. To achieve SDGs targets in maternal and perinatal deaths, there should be preparations including regular training programmes and effective monitoring of MPDSR implementation through supportive supervision involving community representatives and relevant stakeholders to direct the efforts towards implementing the local context responses recommended after death review. In addition, maternal death reviews at regional level using the confidential enquiry into maternal deaths (CEMD) approach were recently introduced in eastern Ethiopia and may be utilized as an additional means to identify the clear gaps in providing quality care and address local-level barriers [37]. Given the chart anonymization and the review committees being established at the regional level, the problem of blaming, defensive attitude, and litigation that are reported at the health facility level could perhaps be minimized through CEMD [12, 38]. Together it is hoped that these efforts should help to reducing deaths through improving quality of care by refining MPDSR system.

Conclusions

In Ethiopia, the MPDSR system has captured, notified or reviewed less than 22% of expected maternal and perinatal deaths. The MPDSR implementation in Ethiopia is influenced by barriers and facilitators that sit under the domains of political/leadership and governance, health systems, and individual level factors. To optimize the MPDSR system in Ethiopia, efforts need to expand beyond collecting maternal and perinatal death reports by focusing on addressing the present blame culture and creating a more conducive learning environment for death reviews and implementing responses. Focusing on good leadership, health worker capacity building, community involvement, multi-sectoral collaborations and sufficient resources will also help to refine the MPDSR system.

Authors’ contributions

MY, AKT, CS, MK, and TvdA conceived the review. MY and CS screened articles and extracted data. MY drafted the manuscript, which was revised by TvdA, AKT, JS, BJ and MK. All authors approved the final version for submission.

Funding

No specific funding was received for this review. MY’s PhD work, of which this study is part, was funded by the LUMC Global programme.

Data availability

The data will be available from corresponding author based on reasonable request.

Declarations

Ethics approval and consent to participate

Not applicable.

Competing interests

The authors declare no competing interests.