SEARCH

SEARCH BY CITATION

Keywords:

  • maternal mortality;
  • surveillance system;
  • causes;
  • Morocco

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Objective

To describe the development of the maternal death surveillance system (MDSS) in Morocco and discuss the initial results.

Method

The nationwide MDSS was implemented in 2009 with the involvement of health professionals and local authorities. It comprises (i) notification of all deaths of women of reproductive age (from 15 to 49 years); (ii) a preliminary survey to identify pregnancy-related deaths; (iii) a confidential enquiry into all pregnancy-related deaths. The information thus obtained describes socio-demographic characteristics of the women, their obstetric and medical history, the mode of delivery, its follow-up and the medical cause of death.

Results

From 1st of January 2009 to 31st of December 2009, 3814 deaths of women of reproductive age were recorded, and a total of 436 pregnancy-related deaths were identified, with 73.4% of those occurring in health facilities. Among the 313 reviewed records, 80.8% were direct obstetric deaths, and 13.5% were classified as indirect. Haemorrhage was the first direct obstetric cause of death (33%). Heart disease was the main indirect obstetric cause of death (39% of indirect causes).

Conclusion

The Moroccan MDSS is a powerful tool for understanding the causes and circumstances of maternal deaths. However, challenges remain regarding the full coverage of the system, the decentralisation of the data entry and analysis and the completeness of medical records.

Objectif

Décrire le développement du système de surveillance de la mortalité maternelle (SSMM) au Maroc et discuter des premiers résultats.

Méthode

Le SSMM à l’échelle nationale a été mis en place en 2009 avec la participation des professionnels de la santé et des autorités locales. Il comprend (i) la notification de tous les décès de femmes en âge de procréer (de 15 à 49 ans), (ii) une enquête préliminaire pour recenser les décès liés à la grossesse, (iii) une enquête confidentielle sur tous les décès liés à la grossesse. Les informations ainsi obtenues décrivent les caractéristiques sociodémographiques des femmes, leurs antécédents obstétricaux et médicaux, le mode d'accouchement, le suivi et la cause médicale du décès.

Résultats

A partir du 1er janvier 2009 au 31 décembre 2009, 3.814 décès de femmes en âge de procréer ont été enregistrés et au total 436 décès liés à la grossesse ont été identifiés dont 73,4% survenant dans des établissements de santé. Parmi les 313 dossiers examinés, 80,8% étaient des décès obstétricaux directs et 13,5% ont été classés comme indirects. L'hémorragie était la première cause obstétricale directe de décès (33%). La maladie cardiaque était la principale cause obstétricale indirecte de décès (39% des causes indirectes).

Conclusion

Le SSMM marocain est un outil puissant pour comprendre les causes et les circonstances des décès maternels. Toutefois, des défis subsistent en ce qui concerne la couverture complète du système, la décentralisation de la saisie et l'analyse des données et l'exhaustivité des dossiers médicaux.

Objetivo

Describir el desarrollo de un sistema de vigilancia de mortalidad maternal (SVMM) en Marruecos y discutir los resultados iniciales.

Método

El SVMM a nivel nacional se implementó en el 2009 con la participación de profesionales sanitarios y autoridades locales. Incluye (i) la notificación de todas las muertes de mujeres en edad reproductiva (de los 15 a los 49 años); (ii) un ensayo preliminar de identificar las muertes relacionadas con el embarazo; (iii) una encuesta confidencial sobre todas las muertes relacionadas con el embarazo. La información obtenida describe las características sociodemográficas de las mujeres, su historia obstétrica y médica, el tipo de parto, su seguimiento y la causa médica de la muerte.

Resultados

Entre el 1 de Enero y el 31 Diciembre del 2009 se registraron 3814 muertes de mujeres en edad reproductiva y se identificaron un total de 436 muertes relacionadas con el embarazo. De las 313 historias revisadas, un 80.8% eran muertes por causas obstétricas directas y un 13.5% indirectas. La hemorragia era la primera causa obstétrica directa de muerte (33%). Las enfermedades cardiacas eran la principal causa indirecta de muerte (39% de las causas indirectas).

Conclusión

El SVMM de Marruecos es una herramienta poderosa para entender las causas y circunstancias de las muertes maternas. Sin embargo continúan existiendo retos en lo que respecta a la cobertura del sistema, la descentralización de la entrada de datos y el análisis y la integridad de las historias médicas.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Knowledge of maternal mortality ratios (MMR) has not always been considered essential for the development of strategies to reduce them or to monitor progress in their reduction (Graham et al. 1996; Campbell 1999). The need to monitor progress towards the achievement of the Millennium Development Goals (MDG) and the impact of Safe Motherhood programmes was strong leverage to request MMR estimates from all countries (Graham & Hussein 2006; Ronsmans & Graham 2006). The system for registering births and deaths in most countries (109 of 172) is not reliable enough to produce robust levels and ratios (WHO, UNICEF, UNFPA, & The World Bank 2010). Data from national and local demographic studies are used by experts to derive these ratios. However, the 95% confidence intervals for the MMR estimates have an imprecision of plus or minus 30% of the estimate, making it difficult to assess whether a country has actually reduced its maternal mortality ratio over a ten-year period (Stanton et al. 2000). Hence, the Commission on Information and Accountability for Women's and Children's Health pushed for real-time reporting of every maternal death (Commission on information & accountability for women's & children's health's 2011). The objective of such a maternal death surveillance and response system goes beyond the knowledge of national maternal mortality ratios. Such information can stimulate community and civil society awareness if the reports are published at an appropriate time and in an appropriate manner. It can also contribute to improving the quality of maternal health care thanks to a better understanding of the causes and circumstances of death and the implementation of appropriate corrective actions (Danel et al. 2011; Gilmore & Gebreyesus 2012).

In 1997 and 2003 in Morocco, two successive demographic and health surveys (DHS) estimated MMRs that were considered too high: 228 [125–307] per 100 000 live births for the period 1992–97 (Ministère de la Santé du Royaume du Maroc & Ligue des Etats arabes 1997) and 227 [186–267] per 100 000 live births for the period 1995–2003 (Ministère de la Santé du Royaume du Maroc et al. 2005). In December 2007, the National Commission on Maternal and Neonatal Mortality Reduction was established by the Minister of Health. The main stakeholders are universities, professional associations (obstetricians, neonatologists, anaesthetists, paediatricians and midwives), safe motherhood programmes and civil association representatives. The involvement and active participation of these stakeholders led to the definition of strategies to accelerate the reduction in maternal mortality in April 2008. Reduction in maternal mortality became one of the top priorities of the national plan 2008–12 set out by the Moroccan Minister of Health. The objective of the national plan 2008–12 was ambitious in setting the MMR at 50 for 100 000 live births by 2012. (Ministère de la Santé du Royaume du Maroc 2008). Three strategic goals were identified: reducing geographical and financial barriers to access obstetric and neonatal care; improving management of deliveries and obstetric complications; and improving governance and programme management. These goals were then subdivided into nine work streams and 28 action plans, all implemented in 2009 and 2010 (Ministère de la Santé du Royaume du Maroc 2010). One of the nine work streams was the systematic reporting and analysis of maternal deaths, by documenting the location, the cause and circumstances of each maternal death. It was anticipated that involving care providers in both the surveillance system and in designing the actions to be taken would help boost their commitment to the fight to reduce maternal mortality. The aim of this study is to describe the development of the maternal death surveillance system in Morocco and discuss the initial results.

Material and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Background

In Morocco in 2011, the proportion of pregnant women attending at least one antenatal care consultation was 77.2%, and the proportion of deliveries with a skilled attendant was 73.6% (92% in urban areas and 55% in rural areas) (Ministère de la Santé du Royaume du Maroc 2012). The maternal mortality ratio as estimated by the MMEIG was 100 (62–170) in 2010, which means a total number of deaths of 650, showing a regular annual decrease of 5.1% since 1990 (WHO et al. 2012).

Moroccan law provides that when a death occurs, family members are required to report it to the Hygiene Bureau (municipal in cities and communal in rural areas). The Hygiene Bureau then issues a death certificate consisting of two parts. One part is given to the family and is used as a burial permit (a strict condition for burial in cities, less strict in rural areas). The family has to complete its part of the form and provide it to the Civil Registration Office. The Hygiene Bureau officer assesses whether it is a death from natural causes or a violent death. Declaring the death “violent” leads to legal procedures; this is carried out only when a murder is suspected. The medical cause of death is rarely recorded.

In the late 1990s, the Ministry of Health developed a routine reporting system for maternal deaths covering all public health facilities. In 2009, 285 maternal deaths were identified and reported. However, this routine health information system focuses on deaths in maternity wards, does not cover private facilities and does not systematically collect data on home deaths. The Maternal Death Surveillance System (MDSS) was designed to bridge this gap and developed between June 2008 and May 2009 and implemented from June 2009 onwards.

Data collection

Identification of deaths of women in reproductive age

A system of mandatory reporting by local authorities of deaths of women aged 15 – 49 to local health services was implemented with the collaboration of the Ministry of Interior (executive letter 8 September 2008). In cities, the Municipal Hygiene Bureau registers all deaths and shares the information with the District Health Authority. In rural areas, the reporting of the death of a woman of reproductive age by the local authority (moqadem) to the health centre was recorded in a special register and immediately followed by a preliminary enquiry into the death to exclude obvious non-maternal deaths and clearly identify possible pregnancy-related deaths for further verbal autopsy investigation.

In parallel, the discharge register in all hospitals was reviewed daily to identify and record the deaths of woman of reproductive age (WRA) in a special register. In case of a WRA death, the hospital medical officer in charge of medical activities (or equivalent) made a preliminary enquiry. This enquiry had the objective of eliminating any deaths obviously not related to pregnancy and to inform the District Officer in charge of Health Programmes (DOHP) of all possible pregnancy-related deaths requiring further enquiry (Figure 1).

image

Figure 1. Pathway in the identification of maternal deaths.

Download figure to PowerPoint

Identification of maternal deaths

When a pregnancy-related death was identified in a health facility, the DOHP assembled all the components of the hospital file, interviewed the health personnel involved in caring for the deceased woman and completed a questionnaire entitled confidential audit (CA). The CA questionnaire was designed in Morocco and was based on British and French forms (CEMACH 2007) (INSERM & InVS 2006). The data collected described the socio-demographic characteristics of the women, medical information about their obstetric and medical history, the course of the pregnancy, the mode of delivery (birth), its follow-up and information about pathologies specific to the cause of death (records of haemorrhage, pre-eclampsia/eclampsia, infection, anaesthesia, resuscitation) and their management.

In the case of a pregnancy-related death at home, the DOHP or the person in charge of the Safe Motherhood programme at district level (trained in conducting verbal autopsies and supervised by the DOHP) conducted a verbal autopsy using a verbal autopsy (VA) form. Also, when a death occurred less than 24 h after admission to hospital, a further investigation was carried out at the home using a simplified VA form. The VA form is a Moroccan version of the questionnaire created by the MRC Farafenni Field Station (The Gambia) for their demographic surveillance system (WHO 2004a).

The maternal death definition of the 10th international classification of diseases (WHO 2004b) of direct and indirect obstetric causes was used in this analysis. We thus excluded from the analysis late maternal deaths (“deaths from direct or indirect obstetric causes more than 42 days but less than one year after the termination of pregnancy”) and accidental or incidental pregnancy-related deaths.

In each of the 71 health districts, the DOHP conducted the confidential enquiries in hospitals and supervised or led the verbal autopsies. DOHPs are medical doctors who have been trained in collecting hospital data and conducting verbal autopsies during a 2-day session held before the implementation of the MDSS. The team in charge of the MDSS at central level regularly contacted each of them to follow-up on the progress in collecting data and to supervise their work. Quarterly activity reports were transmitted from the districts to the MDSS central level team so that it was possible to monitor progress in the reporting of maternal death cases. The DOHP's role was crucial because she/he guaranteed the accuracy and completeness of the records before making the data anonymous and transmitting them to the National Expert Committee (NEC). After this stage, it was no longer possible to go back to the original source to obtain additional information.

Organisation of maternal death reviews

Each of the 16 regions of Morocco set up a regional committee to review the maternal deaths identified in their region. However, in 2010, the Minister, who committed to reducing the MMR to 50/100 000 live births by 2012, asked the NEC to present the results of the first year (2009) in October 2010. This did not leave enough time for regional committees to analyse their data, and it was decided to exceptionally organise the first MDR at national level with the NEC. Therefore, districts were asked to send all maternal death case files by the end of June so that the NEC could meet in July and produce its report by September 2010.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Reported pregnancy-related deaths

In 2009, the mandatory reporting of deaths of women aged 15–49 identified 3 814 deaths of which 2 317 (60.8%) were reported as home deaths by local authorities and 1497 (39.3%) on the basis of the surveillance system in health facilities. Reporting of deaths of women of reproductive age by local authorities varied between 16.5% in the Fes region and 71% in the Oriental region (mean 42.5%).

A preliminary enquiry to eliminate deaths that were obviously not pregnancy related was carried out in 1804 cases (78%) of the 2 317 home deaths of women of reproductive age reported by the local authorities. Four hundred and thirty-six pregnancy-related deaths were identified in the districts. This gives a proportion of 11.4% maternal deaths among deaths of women of reproductive age and a ratio of 73 pregnancy-related deaths per 100 000 live births. One hundred and sixteen of these 436 deaths occurred at home (26.6%) and 320 in health facilities (73.4%).

Of the 436 deaths identified as pregnancy related, 345 files (78%) were completed (255 files for women who died in hospital and 90 for women who died at home), but only 317 were sent to the NEC in time to be reviewed for this analysis. Four verbal autopsy (VA) files were eliminated (one death unrelated to pregnancy, one death in 2008 and two deaths were duplicates) leaving 313 files for audit. Ten additional deaths were excluded from the analysis: two deaths due to late obstetric complications (between 42 days and one year) and eight incidental deaths. Thus, the scope of the analysis concerns 303 maternal deaths due to obstetric causes.

Socio-demographic characteristics of the deceased women

At the time of death, nearly 50% of the women were between 25 and 35 years old (median 30 years). They had had on average two children before dying. Among the cases with documentation on antenatal consultations (ANC) (79% of all records), 56% had attended at least one ANC. Only 2.9% had attended an ANC at least four times. However, the large number of cases (20.8%) for which the exact number of ANCs was not registered prompts cautious interpretation of this figure. Further, although 53.6% of the women in the study lived in rural areas, the high proportion of cases for which we have no information about their origin (31%) does not allow us to confirm that the majority of dead women came from a rural setting.

Location of maternal deaths

The majority of maternal deaths (217/303; 71.6%) occurred in public hospitals, one death occurred in a basic health service and 12 (4.0%) during the transfer between health facilities. No deaths occurred in private hospitals, but one death occurred during the transfer from a private hospital to a referral public hospital. Only 72 deaths (23.8%) occurred at home or during transit from home to health facility (Table 1).

Table 1. Obstetric causes of maternal death by location
Location of deathCause of deathTotal
Direct (%)Indirect (%)Non-specified obstetric cause (%)
Health facility (hospital and basic HS)180 (83%)31 (14%)7 (3%)218
Transfer between facilities12 (100%)0012
At home39 (72%)8 (15%)7 (13%)54
On the way from home to a health centre14 (78%)1 (6%)3 (17%)18
Other01 (100%)01
Total245 (81%)41 (14%)17 (6%)303

Of the 215 women who died in hospital, and for which there is detailed information, 52.1% died in an intensive care unit, 37% in a maternity ward and 1.2% in various other units (operating theatre, emergency ward and other hospital units).

The majority of women (235; 77.6%) died after delivery and 63% within 24 h of delivery (142 cases of a total of 225 for whom detailed information was available). Fifty-two women (17.2%) died before delivery and 16 (5.3%) during childbirth (Figure 2). Among the 251 women who died during or after childbirth, 129 (51.4%) had a vaginal delivery, while 68 (27.1%) had a caesarean section, and for 54 (21.5%), no information was available.

image

Figure 2. Time of maternal death.

Download figure to PowerPoint

Medical causes of maternal deaths

Of the total number of 303 maternal deaths, 245 (80.9%) were direct obstetric deaths. Forty-one deaths (13.5%) were classified as indirectly caused. In 5.6% of cases, the cause of death could not be specified except for the fact that it was unequivocally obstetric. This lack of documentation of the cause of death is less common in health facilities (3%) than at home (13%) or on the way to a health centre (17%) (Table 1).

Haemorrhage is the first direct obstetric cause (33% of all deaths). Seven of the 12 deaths that occurred during a transfer between health facilities were due to haemorrhage. The second leading cause was pre-eclampsia/eclampsia (18%), followed by infections (8%) and ruptured uterus (7%). Six deaths from abortion were classified as haemorrhage (four cases) and infection (two cases). Not a single death from ectopic pregnancy was identified. Cardiac disease was the most frequent indirect cause of maternal death with 16/41 (39%) deaths (Table 2).

Table 2. Obstetric causes of maternal deaths (N = 303)
Direct cause245 (81%)
Haemorrhage99 (33%)
Ruptured uterus22 (7%)
Pre-eclampsia/eclampsia55 (18%)
Infection23 (8%)
Other direct cause46 (15%)
Indirect cause41 (14%)
Heart disease16 (5%)
Infectious disease6 (2%)
H1N1 influenza4 (1%)
Respiratory diseases4 (1%)
Epilepsy3 (1%)
Other7 (2%)
Unspecified1 (0%)
Obstetric death of unspecified cause17 (6%)

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

This study describes the results of the first year of the maternal death surveillance system recently implemented in Morocco. Its originality resides in the fact that while such systems are currently being developed in a few developing countries, their results and performances are rarely reported in a scientifically rigorous way. We identified nationwide maternal deaths surveillance reports for only two countries, Egypt and South Africa (Gipson et al. 2004; Roushdy 2007; National Committee of Confidential Enquiries into Maternal Deaths 2008; Bradshaw & Dorrington 2012). Among developing countries, only the Philippines and Mexico have fairly complete statistical records of maternal deaths (numbers and causes) thanks to a well-performing vital statistics registration system (WHO 2011). Most other countries rely on surveys.

In the first year of the Moroccan MDSS (2009), a total of 436 pregnancy-related deaths were identified using this new tool. Three hundred and twenty deaths (73%) occurred in hospitals or during the transfer between two facilities. This means that the MDSS identified 35 deaths more from public hospitals than the routine health information system, and 151 more deaths if the 116 home deaths are included. However, only 317 files (73%) were completed and sent in time for review. The majority of the 303 maternal deaths analysed were of direct obstetric cause (81%), and haemorrhage was the leading cause. 2009 was the first year of the MDSS in Morocco, and although it has contributed to a better understanding of the causes of maternal deaths, the system still has several limitations.

Low coverage of reported deaths of women aged 15–49

The MDSS relies on a Reproductive Age Mortality Study (WHO et al. 2010) approach which covers the whole country. After one year of activity, the first weakness of the programme appeared to be the incompleteness of data. According to the 2009–2010 national demographic survey (Haut - Commissariat au Plan du Royaume du Maroc 2011), the mortality rate of women aged 15–49 was 1.02 per 1000 which, applied to the 2009 Moroccan population, would mean a total of 8983 deaths. Through the MDSS, only 3814 deaths of women aged 15–49 were reported (42.5%).

The surveillance programme started mid-2009, and therefore, deaths between January and June 2009 were retrieved retrospectively; as a result, some may have been missed. The start of the surveillance system coincided with the threats of a large-scale outbreak of H1N1 influenza, which was given higher priority by the Ministry of Health. This diverted the DOHPs from full involvement in the new monitoring system for maternal deaths. Finally, the collaboration with the Ministry of Interior, responsible for the declaration of any death of women aged 15–49 to the local health service, was not optimal in rural areas. Moqadems (communal clerks) were either not informed that they should send a declaration to the health services or not committed to doing it. This probably explains the under-reporting of deaths of women aged 15–49.

Underreporting of pregnancy-related deaths

The identification of pregnancy-related deaths in hospitals has been carefully conducted, and we assume that the majority of deaths have been reported. It is important to notice that 35 more deaths (320 cases in total) were identified than through the routine health information system (285 cases). We were, however, surprised by the small number of pregnancy-related deaths identified at home. This may indicate that women only use hospitals in case of severe complications, or it may be the result of under-reporting. The survey conducted by the Haut-Commissariat au Plan estimated maternal mortality at 112 [95% CI 73–148] maternal deaths per 100 000 live births for the period 2009–10 (Haut-Commisariat au Plan du Royaume du Maroc 2011). Applied to the total number of births in Morocco, this would mean 667 maternal deaths [95% CI 435–882]. Using the MDSS, we identified 436 pregnancy-related deaths, a figure which is at the low end of the estimate. Moreover, only two cases of late maternal deaths were identified, which is surprisingly few. Finally, all pregnancy-related deaths for which a confidential enquiry had been conducted turned out to be maternal deaths, indicating a very high specificity in the identification of pregnancy-related deaths. In South Africa, deaths reported by confidential enquiries turned out to be underestimated, particularly deaths at home, and cannot be used to measure the ratio of maternal mortality (National Committee of Confidential Enquiries into Maternal Deaths 2008). They have, however, allowed a better understanding of the shortcomings in the quality of the health system and boosted the commitment of professionals to improve the care of mothers. In Egypt, although the maternal death surveillance is presented as a success, voices have been raised to ask for: more transparency in the MDSS report, more in-depth analysis to elucidate the relationship between women's educational level and mortality or between wealth and mortality, and to identify avoidable factors (Egyptian Initiative for Personal Rights 2010).

Of the 303 files analysed, the precise cause of maternal death could not be determined for 17 cases (5.6%) because of missing information or incomplete records. Seven of these cases occurred in hospital. One of the recommendations of the analysis of the records has been a better documentation of deaths in hospitals and the encouragement of autopsies when it is not possible to determine the cause of death on the basis of clinical data.

Relevance of a surveillance system for maternal deaths

What is the justification for creating an additional system instead of strengthening the existing civil registration system? The main advantages of specific maternal death surveillance systems with confidential enquiries are their completeness and representative nature (Ronsmans 2001). Efforts to identify all maternal deaths often lead to a more complete report about the deaths than those obtained through the normal civil registration process. For example, in Menoufia (Egypt), the Reproductive Age Mortality Study (RAMOS) identified three times more maternal deaths than from death certificates alone (Grubb et al. 1988). The method of identifying maternal deaths (for example, through the routine surveillance system or a special study, by verbal autopsy by a man or by a woman, by the use of open or closed questions) also influences the results in terms of numbers of maternal deaths diagnosed and by type (direct or indirect) of obstetric causes (Ronsmans et al. 1998). A specialised system can have its place, most importantly in countries where the usual system of recording deaths has not yet reached complete inclusion of the population and does not yet have the support of quality assurance mechanisms. In the UK, confidential enquiries began in 1952 and are still carried out, in an extended and evolving form. These enquiries complement the identification of deaths of the registry offices by validating the cause of death and by describing the avoidable and unavoidable circumstances which led to death (Drife 2005). In USA, it was estimated that 38% of maternal deaths were unreported on death certificates (Horon 2005). In France, maternal mortality based on official statistics was underestimated by 22% (Deneux-Tharaux et al. 2005). In Tunisia, a system of recording pregnancy-related deaths exists since 1999. It is focused on hospitals assuming that about all women deliver in a health facility, and hence, there should be a very small number of home deaths. The coverage was, however, low with only 67% of maternal death investigations completed in public hospitals (Dellagi et al. 2008). These observations show the limitations of maternal mortality statistics and suggest the need to carefully investigate pregnancy-related deaths.

In Morocco, the decision to apply the system from the start to the whole country instead of starting with pilot studies in one or two regions was justified by the political will to reduce quickly maternal mortality. The cost of the programme has not been fully quantified but can be estimated at around € 90 000–110 000 annually (between 206 and 252€ per pregnancy-related death). This amount includes the staff and travel to conduct the surveys at home, to establish hospital files and verbal autopsies and includes the time needed to manage the programme, to enter and analyse the data, as well as setting up expert meetings at regional and central level to determine the causes of deaths. The cost also covers some of the resources needed for the audit of deaths. The computerisation of data collection through a secured website is under implementation and should reduce the recurrent costs of the central management of the programme. Computerisation simplifies data entry. The average cost per death identified and attributed to maternal causes will, however, still exceed € 120 per pregnancy-related death. This figure is lower than the cost measured in three rural districts in India (US$ 362 or € 300/identified and validated death), where 62% of women die at home (Barnett et al. 2008). In the Indian study, key informants identified all births and deaths of women aged 15 – 49, and verbal autopsies were conducted to ascertain the maternal cause of death. In Morocco, births and deaths at home are identified by a civil servant from the Ministry of Interior, whose salary is not taken into account in the cost calculation. Moreover, the proportion of identified home deaths is smaller in Morocco (27%).

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

The transition from periodic demographic and health surveys to an epidemiologically based surveillance system of maternal deaths was a political choice made to improve the governance of the maternal health programme in Morocco. Such a system allowed for the identification of 12.3% more pregnancy-related deaths in health facilities than the routine information system. By enabling a better understanding of the causes and circumstances of maternal deaths, it also provided a basis for action. We estimate that the cost of this new system should be less than € 120 000 per year once fully implemented, an expense that in our opinion is fully worth it.

The short-term challenges faced by this new surveillance system are to improve coverage, notably in rural areas, and maintain staff commitment and performances at the field level. This will not happen without regional committees. In Morocco, regional accountability is the necessary condition to improve the completeness of the MDSS, both for the registration of deaths of women aged 15–49 and for maternal deaths. It should also improve the quality of data, enhance the commitment of the health professionals and local leaders and guarantee the implementation of recommendations. The medium and long-term objectives would be to implement a more rigorous civil registration system where the cause of death of a woman aged 15–49 is stated on the death certificate.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

The authors thank the District Officers in charge of Health Programmes (SIAAP) and the district teams in charge of the MDSS for their contribution to the implementation of this system. We are grateful to the Ministry of the Interior for its cooperation. Particular thanks go to Marilys Corbex (Maternal & Reproductive Health Unit, Institute of Tropical Medicine) for her useful comments and to Rachel Hammonds (International Health Policy Unit, Institute of Tropical Medicine) for her careful editing.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References
  • Barnett S, Nair N, Tripathy P, Borghi J, Rath S & Costello A (2008) A prospective key informant surveillance system to measure maternal mortality - findings from indigenous populations in Jharkhand and Orissa, India. BMC pregnancy and childbirth 8, 6.
  • Bradshaw D & Dorrington R (2012) Maternal mortality ratio – trends in the vital registration data. South African Journal of Obstetrics & Gynaecology 18, 3842.
  • Campbell OMR (1999) Measuring progress in safe motherhood programmes: uses and limitations of health outcome indicators. In Safe Motherhood Initiatives: critical issues. (eds M Berer & TK Sundari) Blackwell Science, Oxford, pp. 3142.
  • CEMACH (2007) Saving Mothers' Lives: reviewing maternal deaths to make motherhood safer- 2003-2005. The Seventh Report on Confidential Enquiries into Maternal Deaths in the United Kingdom. The Confidential Enquiry into Maternal and Child Health, London.
  • Commission on information and accountability for women's and children's health (2011) Keeping promises, measuring results. World Health Organization, Geneva, p -Available from: http://www.everywomaneverychild.org/images/content/files/accountability_commission/final_report/ Final_EN_Web.pdf.
  • Danel I, Graham WJ & Boerma T (2011) Maternal death surveillance and response. Bulletin of the World Health Organization 89, 779779A.
  • Dellagi RT, Belgacem I, Hamrouni M & Zouari B (2008) [Evaluation of the maternal mortality surveillance system in public facilities in Tunisia (1999-2004)]. Eastern Mediterranean health journal 14, 13801390.
  • Deneux-Tharaux C, Berg C, Bouvier-Colle MH et al. (2005) Underreporting of Pregnancy-Related Mortality in the United States and Europe. Obstetrics and gynecology 106, 684692.
  • Drife JO (2005) Why mothers die?. The journal of the Royal College of Physicians of Edinburgh 35, 332336.
  • Egyptian Initiative for Personal Rights (2010) Maternal Health. http://www.eipr.org. (accessed 1 March 2010).
  • Gilmore K & Gebreyesus TA (2012) What will it take to eliminate preventable maternal deaths? Lancet 380, 8788.
  • Gipson R, El Sharkably E & Issa AH (2004) Maternal Mortality in Egypt: from survey to surveillance. The 132nd Annual Meeting (November 6-10) of American Public Health Association.
  • Graham WJ & Hussein J (2006) Universal reporting of maternal mortality: an achievable goal?. International journal of gynaecology and obstetrics 94, 234242.
  • Graham WJ, Filippi VG & Ronsmans C (1996) Demonstrating programme impact on maternal mortality. Health Policy and Planning 11, 1620.
  • Grubb GS, Fortney JA, Saleh S et al. (1988) A comparison of two cause-of-death classification systems for deaths among women of reproductive age in Menoufia, Egypt. International Journal of Epidemiology 17, 385391.
  • Haut-Commisariat au Plan du Royaume du Maroc (2011) Enquête nationale démographique à passages répétés 2009-2010. Principaux résultats. Haut-Commissariat au Plan, Rabat.
  • Horon IL (2005) Underreporting of maternal deaths on death certificates and the magnitude of the problem of maternal mortality. American journal of public health 95, 478482.
  • INSERM & InVS (2006) Rapport du Comité national d'experts sur la mortalité maternelle (CNEMM) 1999-2001. INSERM et InVS, Paris, p-149.
  • Ministère de la Santé du Royaume du Maroc (2008) Plan d'action santé 2008-2012: réconcilier le citoyen avec son système de santé., Ministère de la Santé, Rabat. http://srvweb.sante.gov.ma/Ministere/Mission/strategie/Documents/plan_action_sante_2008_2012.pdf.
  • Ministère de la Santé du Royaume du Maroc (2010) Plan National 2008-2012 pour l'Accélération de la Réduction de la Mortalité Maternelle et Infantile. Etat d'avancement. Direction des Hôpitaux et des Soins Ambulatoires, Rabat.
  • Ministère de la Santé du Royaume du Maroc (2012) Enquête Nationale sur la Population et la Santé Familiale [EPSF 2011]. Direction de la Planification et des Resources Humaines. Ministère de la Santé du Maroc, Rabat.
  • Ministère de la Santé du Royaume du Maroc & Ligue des Etats arabes (1997) Enquête Nationale sur la Santé de la Mère et de l'Enfant (ENSME) 1997. Morocco, Rabat, 36.
  • Ministère de la Santé du Royaume du Maroc, ORC Macro, & Projet PAPFAM (2005) Enquête sur la Population et la Santé Familiale 2003-04 (EPSF). Morocco and Calverton USA and Cairo, Egypt, Rabat.
  • National Committee of Confidential Enquiries into Maternal Deaths (2008) Saving Mothers 2005-2007: Fourth Report on Confidential Enquiries into Maternal Deaths in South Africa. Government Printer, Pretoria.
  • Ronsmans C (2001) What is the evidence for the role of audits to improve the quality of obstetric care. In: Safe Motherhood Strategies: a review of the evidence. (eds Vincent De Brouwere & Wim Van Lerberghe) ITG Press, Antwerpen, pp. 207228.
  • Ronsmans C & Graham WJ (2006) Maternal mortality: who, when, where, and why. Lancet 368, 11891200.
  • Ronsmans C, Vanneste AM, Chakraborty J & Van Ginneken J (1998) A comparison of three verbal autopsy methods to ascertain levels and causes of maternal deaths in Matlab, Bangladesh. International Journal of Epidemiology 27, 660666.
  • Roushdy N (2007) Maternal Mortality surveillance System. Bangkok, September 6, 2007. USAID, http://www.esdproj.org/DocServer/CC8_Nahla_Roushdy.pdf (accessed 25 October 2012).
  • Stanton C, Abderrahim N & Hill K (2000) An assessment of DHS maternal mortality indicators. Studies in Family Planning 31, 111123.
  • WHO (2004a) Beyond the numbers. Reviewing maternal deaths and complications to make pregnancy safer. World Health Organization, Dept. of Reproductive Health and Research, Geneva, p-150.
  • WHO (2004b) ICD-10: international statistical classification of diseases and related health problems: tenth revision. 2nd edn. World Health Organization, Geneva.
  • WHO (2011) Monitoring maternal, newborn and child health: understanding key progress indicators World Health Organization, Geneva.
  • WHO, UNICEF, UNFPA, & The World Bank (2010) Trends in Maternal Mortality, 1990 to 2008. Estimates developed by WHO, UNICEF, UNFPA and The World Bank. World Health Organization, Geneva.
  • WHO, UNICEF, UNFPA, & The World Bank (2012) Trends in Maternal Mortality: 1990 to 2010. WHO, UNICEF, UNFPA and The World Bank estimates. World Health Organization, Geneva.