Maternal death review in Africa

1. Introduction 

Maternal death review (MDR) is a key element of a strategy to improve the quality of maternal healthcare services by focusing on the causes of deaths and what could have done to avert each death. MDR does more than count maternal deaths; it looks beyond the numbers to study the causes and avoidable factors behind each death, leading to actions to improve quality of care based on the findings. Universal reporting of maternal death is an important element of health management information systems as long as it leads to the allocation of resources and taking responsibility for actions to prevent these avoidable deaths [1]. Striving for the highest quality reporting, in terms of universal coverage and reliability, is essential to inform actions to prevent maternal deaths, and to support the human right to be acknowledged in life and in death [2].

South Africa is currently the only Sub-Saharan African country to institutionalize MDR into the health system. Since the introduction of Confidential Enquiry of Maternal Death (CEMD) in 1997, a majority of maternal deaths are reported, the causes of death are analyzed and reviewed, and actions are taken to improve the quality of maternal health care. MDR needs political buy-in and the force of legislation to work well. South Africa sets an example and inspires other low and middle resource countries in Africa to follow suit.

In many African countries maternal deaths have been reviewed at some health facilities. There are also documented successes in reducing maternal death, improving quality of care, and increasing client satisfaction and use of available services [3], [4]. However, in nearly all African countries, MDRs are carried out ad hoc, reporting is incomplete, and there is limited political, institutional or legal support. Institutionalizing MDR into the health system has been difficult for technical, human, and financial resource reasons, as well as inattention to quality and accountability of maternal health care.

2. Methods 

2.1. Introducing maternal death review 

The process of introducing MDR systematically into all African countries started in 2003 when 3 UN agencies (WHO, UNFPA, and UNICEF) collaborated with development partners and professional bodies to introduce MDR, aiming for institutionalization of MDR at the health system level. Four orientations for national MDR committees and program managers from 34 countries were held in 2003 and 2007 (Table 1). All countries drafted provisional national plans. During the orientations, 5 MDR methods were introduced: verbal autopsy; facility-based MDR; near miss review; confidential enquiry of maternal death; and criterion-based clinical audit. “Beyond the Numbers” a WHO publication published in October 2004 was the main reference for orientations [5].

Table 1. Countries invited to participate, attending orientation, and/or participating in evaluation of MDR in Sub-Saharan Africa, 2003–2007.

	Country	Invited	Oriented	Evaluated
	Angola	Y	Y	Y
	Benin	Y	Y	N
	Botswana	Y	Y	Y
	Burkina Faso	Y	Y	Y
	Burundi	Y	Y	Y
	Cameroon	Y	Y	N
	Cape Verde	Y	Y	N
	Central African Rep	Y	Y	Y
	Chad	Y	N	Y
	Congo	Y	Y	Y
	Comoros	Y	N	Y
	Côte D'Ivoire	Y	Y	Y
	Democratic Rep. Congo	Y	Y	Y
	Djibouti	Y	N	N
	Equatorial Guinea	Y	N	N
	Eritrea	Y	N	Y
	Ethiopia	Y	Y	Y
	Gabon	Y	Y	N
	Gambia	Y	N	N
	Guinea Bissau	Y	Y	Y
	Guinea (Conakry)	Y	Y	N
	Ghana	Y	Y	Y
	Kenya	Y	Y	Y
	Lesotho	Y	Y	N
	Liberia	Y	N	N
	Madagascar	Y	Y	N
	Malawi	Y	Y	Y
	Mali	Y	Y	Y
	Mauritania	Y	Y	N
	Mozambique	Y	Y	Y
	Namibia	Y	Y	Y
	Niger	Y	Y	Y
	Nigeria	Y	Y	N
	Rwanda	Y	Y	Y
	São Tomé & Principe	Y	N	Y
	Senegal	Y	Y	Y
	Sierra Leone	Y	N	Y
	Somalia	Y	N	Y
	South Africa	Y	N	Y
	Swaziland	Y	Y	N
	Sudan	Y	N	N
	Tanzania	Y	Y	Y
	Togo	Y	Y	Y
	Uganda	Y	Y	N
	Zambia	Y	Y	Y
	Zimbabwe	Y	Y	Y
	Total	46	34	30

3. Results 

3.3. Implementation of maternal death review 

3.3.1. National committees 

National committees have been set up in 7 countries (23%) to plan, coordinate, and implement MDR activities. Among them, 6 meet regularly. Seven countries reported that provincial or district MDR committees have been set up. However, in some countries where MDR started at the provincial or district level (such as Angola, Guinea, Malawi, Mozambique, Tanzania), and where there are provincial and/or district committees, national MDR committees may not yet be in place.

3.3.2. National guidelines 

Thirteen countries (43%) have developed national MDR guidelines and 11 (37%) have begun to implement the guidelines. MDR guidelines must be disseminated widely to local health authorities.

3.3.3. Training 

While having the necessary national capacity is important, skill transfer from national to local level is essential for scaling-up. In Zimbabwe, a 2-day national workshop introduced the concept of MDR and the different MDR approaches. Participants included provincial medical directors, provincial nursing officers, district medical officers, chief nursing officers for city health departments and hospital matrons (directors of nursing). The goal of the workshop was to advocate and get the commitment of managers to support the MDR in their provinces and health facilities. In addition, representatives of 7 countries took a study tour to South Africa after the orientation. They observed the processes and practices of conducting CEMD at local level, and conferred with program managers and healthcare providers about their experience and difficulties. They were able to compare the successful process of CEMD in South Africa with their own situation (Table 2).

Table 2. Comparison of MDR process in eight countries with South Africa, August 2004.

	Key steps of implementing MDR	Ethiopia	Kenya	Malawi	Mozambique	Swaziland	Tanzania	Uganda	Zambia	South Africa
	1. Active concerned group created	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
	2. Development of Tools	Not started	Started	Not Started	Not Started	Done	Not started	Ongoing	Not started	Done
	3. Coverage
	i. Pilot phase	No	Yes	No	No	Started at	No	Dist. Hosp	No	Done
	ii. Implementation	No	No	No	No	all levels	No	No	No	Done
	iii. National level	No	No	No	No		No	No	No	Done
	iv. District level.	No	No	No	No		No	No	No	Done
	4. Government endorsement of MDR	Yes	Yes	?	Yes	Yes	?	Yes	Yes	Yes
	5. Consultation
	i. Professional bodies	Yes	Yes	No	No	Yes	No	No	Yes	Yes
	ii. Civil society	No	No	No	No	No	Yes	No	No	Yes
	iii. Donor agencies	Yes	Yes	No	Yes	Yes	No	Yes	Yes	Yes
	6. Legal Reforms for MDR
	i. Recognizing RH Rights	No	No	No	No	No	No	No	No	Yes
	ii. MDR as notifiable	No	No	No	No	Yes	No	No	No	Yes
	7. Governmental Structure for SMP
	i. Ministry of women affairs	No	No	No	No	No	No	No	No	No
	ii. Directorate	No	No	No	No	No	No	No	No	No
	iii. Unit/Department	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
	8. Training
	i. Orientation/introduction	No	Yes	?	No	Yes	No	No	No	Yes
	ii. In-service/C.M.E.	No	Yes	No	No	Yes	No	No	No	Yes
	iii. Pre-service
	a. Nurses	Yes	Yes	Yes	Yes?	?	Yes	Yes	Yes	Yes
	b. Doctors	Yes?	?	?	?	?	?	?	?	Yes
	c. Paramedicals	Yes	Yes	?	Yes	Yes	Yes	Yes	Yes	Yes
	9. Community involvement
	Preparing the program	No	No	No	No	No	No	No	No	Yes
	Creating awareness	No	No	No	No	No	No	No	No	Yes
	MDR Implementation	No	No	No	No	No	No	No	No	Yes
	10. Implementation stages	No	No	No	No	Yes	No	No	No	Yes

3.3.4. Implementation 

Program managers acknowledged that MDR was not being conducted regularly in all countries, even though all were aware of its role in improving maternal health services and reducing maternal morbidity and mortality. National policy should insist as a minimum that all central, provincial, and district hospitals should conduct MDR (Table 3, Table 4).

Table 3. Status of implementing maternal death review in thirty Sub-Saharan African countries, Nov 2008: Part 1.

	Country	MDR policy	MDR in MNCH package	MD a notifiable condition	MD reviewed and analyzed	MDR report from health facilities to district	Any MDR report produced	Recommendations implemented	Government budget for MDR	Financial support from partners
	Angola	N	N	Y	Y	Y	N	N	N	Y
	Burkina Faso	Y	N	Y	Y	Y	Y	Y	Y	Y
	Botswana	Y	Y	N	Y	N	N	N	Y	Y
	Burundi	N	N	Y	Y	Y	N	Y	N	Y
	Chad	N	N	N	N	N	N	N	N	N
	Central African Rep.	N	N	Y	N	Y	N	N	N	N
	Comoros	N	N	N	N	Y	N	N	N	N
	Congo	Y	Y	Y	Y	Y	Y	Y	Y	Y
	Côte d'Ivoire	N	N	Y	N	Y	Y	N	N	Y
	Dem. Rep. Congo	N	N	Y	Y	N	N	N	N	Y
	Eritrea	Y	Y	Y	N	Y	Y	Y	N	Y
	Ethiopia	N	N	N	N	N	N	N/A	N	Y
	Ghana	Y	Y	Y	Y	Y	Y	Y	Y	Y
	Guinea	N	N	Y	N	Y	N	N	N	Y
	Kenya	Y	N	N	Y	Y	Y	Partially	N	Y
	Malawi	N	Y	Y	Y	Y	N	Y	Y	Y
	Mali	Y	Y	Y	Y	N	Y	Y	N	Y
	Mozambique	N	Y	Y	Y	Y	N	Y	N	Y
	Namibia	N	N	N	N	Y	N	N	N	Y
	Niger	N	N	Y	N	Y	Y	N	N	N
	Rwanda	N	N	N	N	N	N	N	N	N
	Senegal	Y	N	Y	N	Y	Y	N	N	Y
	Sierra Leone	N	N	N	N	N	N	N	N	N
	Somalia	N	N	N	N	N	N	NA	N	N
	South Africa	Y	Y	Y	Y	Y	Y	Y	Y	Y
	São Tomé & Principe	N	N	Y	N	Y	Y	Y	N	N
	Tanzania	Y	Y	Y	Y	Y	Y	Y	Y	Y
	Togo	N	N	Y	N	Y	N	N	N	N
	Zambia	Y	N	Y	Y	Y	Y	Y	N	Y
	Zimbabwe	Y	Y	Y	Y	N	Y	N	N	Y

Table 4. Status of implementing maternal death review in thirty Sub-Saharan African countries, Nov 2008: Part 2.

	Country	MDR method	National MDR committee	National committee meet regularly	District and provincial committee	Professional body involvement	National guideline developed	National guidelines in use	Number received MDR training	MDR work plan updated
	Angola	FB MDR, VA	N	N/A	Y	N	Y	Y	5	N
	Burkina Faso	FB MDR	Y	Y	Y	N	Y	Y	?	Y
	Botswana	FB MDR	Y	Y	N	Not available	Y	Y	N/A	Y
	Burundi	FB MDR	N	N	N	Y	N	N	5	Y
	Chad	Not specified	N	N	N	N	N	N	NA	N
	Central African Rep.	Not specified	N	NA	N	N	N	NA	0	N
	Comoros	Not specified	N	N	N	N	N	N	N	N
	Congo	Not specified	N	N/A	N/A	Y	Y	Y	?	N/A
	Cote d'Ivoire	Not specified	N	N	N	N	N	N	?	N
	Dem. Rep. Congo	FB MDR	N	N	N	N	N	N	6	N
	Eritrea	FB MDR	N	N	Y	N	Y	Y	300	N
	Ethiopia	FB MDR	N	N	N	N	N	N/A	12	N
	Ghana	FB MDR	N	N	Some	Y	Y	Y	Y	Y
	Guinea	Not specified	N	N	Y	Y	N	N	15	N
	Kenya	FB MDR	Y	Y	1 province	Not active	Y	Y	100	Y
	Malawi	FB MDR	N	N	Y	N	N	N	N	N
	Mali	Not specified	Y	Y	N	Y	Y	Y	5	Y
	Mozambique	FB MDR	N	N	Y	Y	Y	N	NA	N
	Namibia	Not specified	N	N	N	N	Y	N	NA	N
	Niger	Not specified	Y	N	N	N	N	N	N/A	N/A
	Rwanda	FB MDR	N	N	N	N	N	N	3	N
	Senegal	FB MDR	Y	Y	N	Y	Y	Y	15	N
	Sierra Leone	Not specified	N	N	N	N	N	N	NA	N
	Somalia	Not specified	N	N	N	N	N	N	N	N
	South Africa	CEMD	Y	Y	Y	Y	Y	Y	Many	Y
	São Tomé & Principe	Not specified	N	N	N	N	N	N	NA	N
	Tanzania	FB MDR	N	N	Y	Y	Y	Y	?	Y
	Togo	Not specified	N	N	N	N	N	N	6	N
	Zambia	FB MDR	N	N	Y	Y	Y	Y	270	N
	Zimbabwe	FB MDR	N	NA	N	Y	N	NA	NA	N

Although 15 countries (50%) reported that maternal deaths are reviewed and analyzed at national, local or health institution level, only two countries report maternal deaths.

Twenty (67%) countries reported that while maternal death cases are reported from health facilities to district level, only 10 (33%) of them reported that the data had been analyzed. However, 14 (47%) reported that MDR reports have been produced, and 12 (40%) reported that, based on the results of MDR, recommendations have been made and implemented, at least in the institutions involved in the program.

Only 7 countries reported that the government has allocated budget for conducting MDR. In 18 countries financial support for MDR was provided by development partners.

3.3.5. Type of review selected 

Facility-based MDR is the most common method, chosen by 15 (50%) countries, followed by CEMD in South Africa and Swaziland, and facility-based MDR with verbal autopsy in Angola.

3.3.6. Collaboration with professional bodies 

Ten countries (33%) reported that the professional bodies and medical and OBGYN associations have supported MDR activities.

4. Lessons learned 

5. Discussion 

A national interest group and dedicated national committee supported by the government and professional bodies are a critical driving force behind MDR in those countries where progress is being made. Development of national guidelines may require a long process of consensus building among partners. The Kenyan group that participated in the first MDR orientation organized by the UN in 2003, met weekly, for about one year, to develop the national guidelines. A strong national interest group on MDR keeps pressure to act on the Ministry of Health, facilitates the development of necessary instruments and guidelines, and trains program managers and healthcare providers. In Kenya there was no government budget to support this interest group, but the professional commitment of the members working on their own time ensured the drafting of the national guidelines.

The national interest group, or some version of it, may become official. In South Africa, the Minister of Health appointed a National Committee on Confidential Enquiries into Maternal Deaths (NCCEMD). The NCCEMD is responsible for the confidential enquiry into maternal mortality in South Africa. Their task is: "To make recommendations, based on the confidential study of maternal deaths to the Department of Health such that the implementation of the recommendations will result in a decrease in the maternal mortality." The recommendations are specific so that their implementation is measurable.

While a national policy and guidelines are critical they do not guarantee MDR implementation at scale. Dissemination of the guidelines and program support to the program managers at the provincial and district level are also essential. A specific budget line for MDR in the health sector budget helps to ensure that it is not lost among competing priorities.

Government ownership and commitment are critical. Among the 7 countries with government allocations for MDR, all have a system in place to report maternal death from health facilities to district level and all have received financial support from development partners. Six have a national MDR committee, 6 have made maternal death a notifiable condition, 6 have developed national guidelines, and 5 update their work plan regularly.

On the other hand, among the 19 countries without government financial support, only 4 have developed national guidelines, 4 have a national MDR committee, and 4 have implemented recommendations after analysis.

MDR is more sustainable when integrated into the maternal and reproductive health program (skilled care, emergency obstetric care, and newborn care) rather than a vertical program. Causes and solutions are then more likely to be related to the availability, use, and quality of maternal and reproductive health care, especially the life saving services [8], [9].

Most countries indicate that external technical support is required to develop national MDR policy and strategies. Capacity building through training, experience exchange, and technical assistance to set up the MDR guidelines, database, and legal framework is needed. In addition, adequate financial support from the government and development partners can help to prioritize MDR. Active external support and follow-up to national MDR groups help ensure that technical support is provided at the right time. Facilitators of the MDR orientation workshops and program managers from South Africa visited selected countries during 2004 and 2006 and worked with key policy makers and program managers. These visits helped consolidate the local MDR processes.

Generally speaking, more affluent countries with better health systems are more likely to undertake MDR, yet MDR may escape their attention. For example, in Namibia, where nearly 80% of women deliver with skilled attendants, MDR is not yet a national policy. While Eritrea, where less than a third of women deliver with skilled attendants, has made significant efforts in that direction.

Developing a nationwide system of MDR is not without constraints. Systemic challenges commonly mentioned include lack of national policy and guidelines and a responsible oversight body, inadequate resources allocated by governments and development partners, shortage of staff, and limited awareness of the benefits and methodology of MDR due to lack of exposure and training. Coordination, so that the different parts of MDR proceed in parallel is also important and sometimes difficult. Lack of a legal framework may result in misconceptions and fears regarding possible punitive measures, and the perception that audits are judgments on the actions of professional medical staff.

In addition, frequent turnover of program managers and staff, competing priorities, incomplete and inconsistent reporting in some districts, and a nonparticipating private sector threaten the implementation of MDR (and other reproductive health programs).

Another critical gap in nearly all countries is that although maternal deaths are reported from health facilities to district health offices, the regular review of deaths does not take place. Hence causes of deaths and recommendations for actions—usually done by senior obstetricians and program managers—are lacking. Few obstetricians work in government and are usually stretched too thin to have time for reviews and providing critical feedback to the reporting institutions. Most government doctors also have private practices, and unless their services are adequately compensated it is unrealistic to expect them to allocate their own time regularly for MDR.

Failure to act on recommendations, perhaps due to lack of will or resources, is commonly reported and can be discouraging.

Only 30 (65%) of the 46 countries responded, despite frequent reminders. Among the nonrespondents, some (e.g. Nigeria and Swaziland) are known to have some MDR activities. The survey lacks the in-depth answers that would have resulted from a face-to-face interview, and the quality of completed questionnaires varies significantly. Nevertheless, we believe the survey to be sufficiently representative to guide future action.

The most common method chosen by countries in Africa is facility-based MDR. This method is easier to conduct than other methods and can be initiated at the district and provincial level or in a single institution. Not including near misses in the review involves fewer cases and is thus less burdensome. However, it does not need to be carried out in isolation. While moving from data collection to action, facility-based MDR may generate establishment of clinical standards and guidelines, which is the basis for a criterion-based clinical audit. And nationwide CEMD becomes feasible when a majority of districts are conducting facility-based MDR and most maternal deaths are captured by the system. Facility-based MDR can also be combined with community verbal autopsy and criterion-based audit of management issues.