Menu
Menu
Menu
Menu
Background. Current estimates of maternal mortality ratios in Kenya are at least as high as 560 deaths per 100,000 live births. Given the pervasive poverty and lack of quality health services in slum areas, the maternal mortality situation in this setting can only be expected to be worse. With a functioning health care system, most maternal deaths are avoidable if complications are identified early. A major challenge to effective monitoring of maternal mortality in developing countries is the lack of reliable data since vital registration systems are either non-existent or under-utilized. In this paper, we estimated the burden and identified causes of maternal mortality in two slums of Nairobi City, Kenya. Methods: We used data from verbal autopsy interviews conducted on nearly all female deaths aged 15-49 years between January 2003 and December 2005 in two slum communities covered by the Nairobi Urban Health and Demographic Surveillance System (NUHDSS). In describing the distribution of maternal deaths by cause, we examined maternal and late maternal deaths according to the ICD-10 classification. Additionally we used data from a survey of health care facilities that serve residents living in the surveillance areas for 2004-2005 to examine causes of maternal death. Results: The maternal mortality ratio for the two Nairobi slums, for the period January 2003 to December 2005, was 706 maternal deaths per 100,000 live births. The major causes of maternal death were: abortion complications, hemorrhage, sepsis, eclampsia, and ruptured uterus. Only 21% of the 29 maternal deaths delivered or aborted with assistance of a health professional. The verbal autopsy tool seems to capture more abortion related deaths compared to health care facility records. Additionally, there were 22 late maternal deaths (maternal deaths between 42 days and one year of pregnancy termination) most of which were due to HIV/AIDS and anemia. Conclusion: Maternal mortality ratio is high in the slum population of Nairobi City. The Demographic Surveillance System and verbal autopsy tool may provide the much needed data on maternal mortality and its causes in developing countries. There is urgent need to address the burden of unwanted pregnancies and unsafe abortions among the urban poor. There is also need to strengthen access to HIV services alongside maternal health services since HIV/AIDS is becoming a major indirect cause of maternal deaths.
The study was conducted in two slums in which the African Population and Health Research Center (APHRC) is implementing the longitudinal Nairobi Urban Health and Demographic Surveillance System (NUHDSS) since August 2002. The surveillance system monitors vital events such as births, deaths, and migration on over 58,000 individuals in two slum communities, Korogocho and Viwandani. The two slums are located about 5–10 km from the city centre and occupy an area less than one square kilometer in size. The main means of transport is by commuter mini-buses which drop and pick-up passengers at the periphery of the slums but do not go inside the slum settlements as there are no paved roads and there is heavy human traffic on the available paths. The informal nature of slums underscores their non-permanence with lack of public infrastructure and social services. There are very few public health facilities serving the two slum communities, and these are mainly located on the outskirts of the slums and are therefore inaccessible at night due to security concerns. The residents are from over 15 ethnic backgrounds with the majority being Kikuyu (28%), Luhya (24%), Kamba (21%) and Luo (15%). In Viwandani, the population mainly comprises labour migrants working in the neighboring industrial area, while the Korogocho population consists mainly of long-term settlers working in the informal sector. We used data from verbal autopsies conducted on nearly all female deaths aged 15–49 years between January 1, 2003 and December 31, 2005 in the NUHDSS. We also used data from a health care facility survey conducted in 2006 to assess maternal health experiences as captured by the health management information system (HMIS) in health care facilities during 2004–2005. The health care facilities were identified from reports provided by women who participated in the household survey component of the project. As part of the NUHDSS routine procedures, verbal autopsy interviews are conducted using a questionnaire adapted from the verbal autopsy tool developed by the World Health Organization [20]. All deaths in the two slum communities are captured through a death registration form completed by a field worker. A detailed verbal autopsy interview is then conducted by a field supervisor trained to conduct verbal autopsy interviews. All verbal autopsy interviewers must have a minimum of 12 years of formal education and are familiar with the slum setting. They are trained on the verbal autopsy procedures for at least one week and retrained at the beginning of each data collection round. Interviews are conducted after making an appointment with the bereaved household normally after the funeral but within approximately 6 weeks of registering the death. Respondents are typically members of the household who cared for the deceased prior to death or have good knowledge of the symptoms or events that led to death and they must consent to be interviewed. Three physicians independently review the completed verbal autopsy forms and assign cause of death using the tenth revision of the International Classification of Diseases (ICD-10) [21]. If two or more concur, the result is then taken as the probable cause of death. Where agreement is not reached, the three physicians meet and discuss the case in order to reach a consensus. If consensus is not reached, the cause of death is coded as unknown. In this study, women who died between January 1, 2003 and December 31, 2005 were identified from the NUHDSS database. Three physicians (a medical epidemiologist and two obstetricians) independently reviewed the verbal autopsy records in order to ascertain the cause of death. The purpose of having the verbal autopsies reviewed again by two obstetricians was to ensure that any maternal deaths that could have been missed by the routine DSS coding were captured. The ICD-10 definition of maternal death, “the death of a woman while pregnant or within 42 days of termination of pregnancy, irrespective of the duration and site of the pregnancy, from any cause related to or aggravated by the pregnancy or its management but not from accidental or incidental causes”, was used. Further, we considered late maternal deaths defined by ICD-10 definition as “death of a woman from direct or indirect obstetric causes more than 42 days but less than one year after termination of pregnancy” [21]. We categorized maternal deaths into direct and indirect causes. Direct obstetric deaths are defined as maternal deaths resulting from obstetric complications of the pregnant state (pregnancy, labor, and the puerperium), interventions, omissions, incorrect treatment or a combination of any of the above while indirect obstetric deaths are those resulting from previous existing disease or disease that developed during pregnancy and which was not due to direct obstetric causes, but was aggravated by physiologic effects of pregnancy [21]. A health care facility survey was conducted in 2006 and targeted facilities that are commonly used by pregnant women living in the two slum communities for obstetric care. A total of 25 health facilities where women delivered between 2003 and 2005 were identified. Selection of health facilities was based on information provided by women who had had a pregnancy outcome between 2003 and 2005 and had been interviewed in the household survey which was part of the larger maternal health project. Some of the health facilities assessed were located in the two slums while the rest were in other parts of Nairobi. We sought ethical approval from the Kenya Medical Research Institute (KEMRI) Ethical Review Committee, which is one of the Institutional Review Boards authorized to give ethical approval for research in Kenya. We also obtained permission from the Ministry of Health and from the Medical Officer of Health in-charge of the Nairobi City Council before visiting the health care facilities. Appointments were made with the respective health care facility personnel to explain the details of the survey after which consent was sought to carry out the interview. Structured interviews were carried out by one clinical officer who underwent three-day training for this exercise. Data on causes of deaths for 2004 and 2005 were extracted from the medical records. Descriptive statistics were used to describe the maternal mortality levels and causes of maternal death in the two slums.
N/A
