BEYOND THE NUMBERS: CONFIDENTIAL ENQUIRIES INTO MATERNAL DEATHS IN ACCRA-GHANA Researcher: Afisah Yakubu Zakariah Universite Libre de Bruxelles- School of Public Health Reproductive Health and Epidemiology of Perinatal Health Unit Brussels-Belgium Afiyakzak@yahoo.com Promoter: Professor Sophie Alexander-Karlin Universite Libre de Bruxelles- School of Public Health Reproductive Health and Epidemiology of Perinatal Health Unit Brussels-Belgium salexand@ulb.ac.be Co-Promoter: Professor Jos van Roosmalen Leiden University Medical Centre Department of Obstetrics and Gynaecology Leiden-The Netherlands J.J.M.van_Roosmalen@lumc.nl "A thesis submitted for the Degree of Doctor of Philosophy in Public Health Sciences Universite Libre de Bruxelles-School of Public Health Reproductive Health and Epidemiology of Perinatal Health Unit. November, 2008. EXECUTIVE SUMMARY Background Maternal mortality remains a severe problem in many parts of the world, despite efforts to reach MDG 5. Assessing progress towards this goal is difficult because maternal mortality is difficult to measure and the information available at country level does not generally permit the establishment of good baseline data. Countries with high maternal mortality ratios neither have adequate vital registration systems nor adequate resources to carry out surveys. Only few low-income countries have been able to establish a comprehensive reporting system and even where such vital registration systems are in place, maternal deaths are often underreported or misclassified as non-maternal even in large well developed cities. Ghana belongs to the group of low-income countries with high maternal mortality ratios (point estimate 560, lower bound 200 and upper bound 1300) per 100,000 live births and inadequate data on maternal deaths. Previous studies have demonstrated that most of these deaths could be prevented with existing effective practices. In this dissertation, we looked at the registration system of births and deaths in the Greater Accra Region of Ghana. We assessed completeness of registration of maternal deaths and data quality. We also looked at the degree of underreporting of maternal deaths, assessed causes of maternal deaths and substandard care of these cases through a confidential enquiry. This enabled us to identify problems associated with measuring of maternal mortality in Ghana and the standard of care of the cases. Through our findings we were able to make recommendations to achieve MDG 5 in the country by 2015 if implemented. Other maternal and child health (MCH) interventions were also looked as working to improve MCH is a continuum, and no aspect should be neglected. The first relates to seeking evidence based practice in presence of potentially complicated obstetrical conditions like premature rupture of membranes and the second pertains to preventive activities in MCH and concentrates on the results of tetanus immunisation of women in their reproductive age in the Northern Region of Ghana. Objectives 1. To assess the precision (quality) of civil registration of maternal deaths in Ghana 2. To identify the magnitude of maternal deaths using multiple sources and assess the degree of underreporting of these deaths 3. To identify the root causes and related factors of the identified maternal deaths as well as classify the deaths 4. To assess the level of substandard care in maternal deaths in order to identify weakness in the Ghanaian maternity care system and propose locally relevant and affordable solutions to improve maternity care in Ghana 5. To discuss the results of the research with the Decision makers and health care providers leading to an adoption of the method at national level and a reduction of maternal mortality in Ghana as a whole 6. To assess risk factors and management of premature rupture of membranes and make recommendations that will address avoidable complications 7. To assess challenges of other maternal and child health interventions like Tetanus Toxoid immunization in women of reproductive age group and formulate recommendations to increase the coverage in this group. Organization of dissertation This dissertation contains eight chapters. Chapter one is on general introduction, which gives an overview of the problems associated with the reduction of maternal mortality by literature review. Chapter two highlights the profile of Ghana and the Health system Chapter three is on maternal mortality in the Greater Accra region of Ghana for the year 2000. In this chapter, we assessed completeness of registration and data quality as well as the sensitivity and specificity of the civil registration system to capture maternal deaths in Ghana. Chapter four discusses Reproductive Age Mortality Survey (RAMOS) conducted in Accra city (Metropolis) for the year 2002. It showed how effective the method is in identifying maternal deaths using multiple sources approach. In this study, we aimed at identifying the magnitude of maternal deaths in the city and the degree of underreporting during the study period. Chapter five elaborates the main content of this thesis, which is on confidential enquiries into maternal deaths for the year 2002. This chapter deals solely with the seventy three cases of maternal deaths that were identified during the study period (1st January 2002-31st December 2002) with adequate information surrounding the deaths. We identified the root causes and related factors of the identified maternal deaths as well as their classification. We also assessed the level of substandard care in maternal deaths, identified weaknesses in the Ghanaian maternity care system and proposed locally relevant and affordable solutions to improve maternity care in Ghana. Chapter six is on controversy of management of premature rupture of membranes at term. We reviewed systematically the existing evidence on expectant management of PROM and prompt delivery after a premature rupture of membranes at term and gave an answer to this controversy. Chapter seven assessed challenges of other maternal and child health interventions like Tetanus Toxoid immunization in women of reproductive age group and formulated recommendations to increase the coverage in this group. Chapter eight gives the general discussion, conclusion and recommendations. Main results and conclusion The study on the assessment of completeness of registration and data quality of maternal deaths in the Greater Accra Region of Ghana found the process of registration of maternal deaths in the Region to be incomplete even though there is a system of registration of deaths in Ghana. We identified the major causes of under registration of maternal deaths during the study as omission of pregnancy status in the registers. Pregnancy status is not systematically noted in the registers and the death certificate forms. Other contributing factors to the under registration of maternal deaths were erroneous classification of maternal deaths as non-maternal and imprecision in the death certification and registration. The study on Reproductive Age Mortality Survey (RAMOS) in Accra city (Metropolis) for the year 2002 has almost doubled the officially reported figures, which showed that the method is effective in identifying maternal deaths using multiple sources approach. The findings of the confidential enquiries into maternal deaths for the year 2002 showed that the proportion of maternal deaths associated with substandard care factors at various levels is high in the Metropolis even though the city has the best medical services in the country. The study on the controversy of management of premature rupture of membranes at term shows that prompt delivery after the membranes are ruptured averts many complications both for the mother and the baby. It is presented as an illustration of how evidence based guidelines can be instrumental in challenging the maternal mortality problem In the study on challenges of maternal and child health interventions like Tetanus Toxoid immunization in women of reproductive age, we found out that the tetanus toxoid vaccine was perceived as an anti fertility drug by those who refused to be vaccinated. Among other reasons for refusal was complete authority over wives by their husbands and therefore not allowing them to be vaccinated. It also is an illustration of the role of good organisation of care in meeting MDGs 4 and 5. We finally conclude that, even in environments with limited resources, most maternal deaths can be prevented if appropriate measures are put in place to prevent their occurrence. Key words: Maternal mortality, registration, underreporting, RAMOS, confidential enquiries, sub-standard care, PROM.

Maternal mortality remains a severe problem in many parts of the world, despite efforts to reach MDG 5. Assessing progress towards this goal is difficult because maternal mortality is difficult to measure and the information available at country level does not generally permit the establishment of good baseline data. Countries with high maternal mortality ratios neither have adequate vital registration systems nor adequate resources to carry out surveys. Only few low-income countries have been able to establish a comprehensive reporting system and even where such vital registration systems are in place, maternal deaths are often underreported or misclassified as non-maternal even in large well developed cities. <p>Ghana belongs to the group of low-income countries with high maternal mortality ratios (point estimate 560, lower bound 200 and upper bound 1300) per 100,000 live births and inadequate data on maternal deaths. Previous studies have demonstrated that most of these deaths could be prevented with existing effective practices.<p><p>In this dissertation, we looked at the registration system of births and deaths in the Greater Accra Region of Ghana. We assessed completeness of registration of maternal deaths and data quality. We also looked at the degree of underreporting of maternal deaths, assessed causes of maternal deaths and substandard care of these cases through a confidential enquiry. This enabled us to identify problems associated with measuring of maternal mortality in Ghana and the standard of care of the cases. Through our findings we were able to make recommendations to achieve MDG 5 in the country by 2015 if implemented. Other maternal and child health (MCH) interventions were also looked as working to improve MCH is a continuum, and no aspect should be neglected. The first relates to seeking evidence based practice in presence of potentially complicated obstetrical conditions like premature rupture of membranes and the second pertains to preventive activities in MCH and concentrates on the results of tetanus immunisation of women in their reproductive age in the Northern Region of Ghana.<p><p>Objectives<p>1.\ / Doctorat en Sciences de la santé publique / info:eu-repo/semantics/nonPublished