A Longitudinal Study of Maternal Morbidity and Mortality in a Remote Region of Northern Tanzania

Brooke R Johnson, Ipas
Moke Magoma, St. Elizabeth's Hospital, Tanzania
Paul W. Leslie, University of North Carolina at Chapel Hill
Trude Bennett, University of North Carolina at Chapel Hill

1 Large numbers of people in East Africa live in dispersed settlements in remote rural areas. Provision of reproductive health services to such communities is difficult, and it is often difficult even to carry out studies on representative samples in order to determine the nature and prevalence of the most important reproductive health challenges. A multidisciplinary research team is conducting a comparative, longitudinal survey on maternal morbidity and mortality in 55 dispersed, small, isolated pastoralist communities in Ngorongoro District, located in northern Tanzania. The study is focused on pregnancy-related morbidity/mortality (including abortion, stillbirths, and complications of delivery), the patterns and prevalence of syphilis and HIV, and other sexual and reproductive morbidities such as injuries and complications from gender-based violence and genital cutting. The study area is a relatively remote mixed pastoral/agricultural community setting with a maternal mortality ratio that is estimated to be among the highest in the world (950/100,000 deliveries), nearly twofold the overall rate for Tanzania. Leading causes of maternal mortality at local hospitals include hemorrhage during pregnancy, labor, or delivery, puerperal sepsis or infection, and anemia. Mortality represents only a small portion of the totality of pregnancy-related morbidity, the extent and nature of which are virtually unknown in this setting. The study, funded by the Rockefeller Foundation, adds a significant reproductive health component to an ongoing project, funded by the National Science Foundation and A.W. Mellon Foundation, on population, migration, and land-use diversification among Maasai pastoralists. The maternal morbidity/mortality study is a collaborative effort that includes researchers from Tanzania, Ipas, and the University of North Carolina at Chapel Hill. The project makes use of an extensive system of mobile outreach clinics run by the district hospital in Wasso and sub-district hospital in Endulen. The population of Ngorongoro District is estimated to be 106,730 of which 70% are Maasai, 20% Watemi, and 10% other ethnic groups. The district has one of the lowest literacy rates in the country, estimated at 17% in the 1988 national census.1 The Maasai of Ngorongoro District are traditionally nomadic pastoralists. However, pressures from population growth, exacerbated by alienation of lands traditionally used for grazing in order to establish national parks and other conservation zones, has in recent decades led many Maasai to adopt agriculture as a supplement to their herding. Also, in the past four or five years, another form of economic diversification has become important: increasing numbers of young men are leaving Ngorongoro District to find employment in urban 2 areas such as Mwanza. In the majority of cases, the migration is temporary – the men return with cash and/or cattle to increase the family herds and provide an economic foundation for establishing their own families. There is widespread concern in the medical community that HIV and other STIs are reaching epidemic proportions as a result of migration, especially in the context of a polygamous society. This concern is supported by an HIV+ rate of 19% among blood donors at Endulen Hospital. Despite the recent economic changes, most Ngorongoro Maasai value their traditional way of life and many adhere to cultural beliefs and practices likely to reduce their utilization of modern maternal health care. It is estimated that approximately one- fifth of pregnant women in the district will develop serious complications such as obstructed labor, placenta abruptio, pre-eclampsia, severe anemia, and pre-term pre- labor rupture of membranes.2 The prevalence and impact of unsafe abortion are unknown. Study objectives Specific objectives of this study are to: · Develop effective processes to document deliveries, maternal morbidity and mortality in a remote, dispersed, pastoral community setting; · Document the prevalence and patterns of morbidity and mortality associated with pregnancy (including spontaneous and induced abortion, stillbirths, and delivery complications), HIV/AIDS and syphilis, and injury due to gender-based violence and other causes; · Assess women’s knowledge and perceptions of sexual/reproductive health risks and morbidity; and · Investigate relationships between labor migration and women’s health by comparing morbidity/mortality of women whose sexual partners work exclusively within Ngorongoro District with that of women whose partners have migrated to work in urban areas outside the district. We hypothesize that the prevalence of syphilis, HIV, and other morbidities will be greater among women whose partners are now or have recently engaged in labor migration to urban areas. Preliminary results from related research now being carried out by one of the principal investigators (Leslie) indicate that there is much variation among communities in the district with regard to proportions of men involved in labor migration. There is a particularly strong contrast between the division comprising the Ngorongoro Conservation Area (NCA) and the divisions outside the NCA (Loliondo and Sale), a contrast arising from the severe limitations placed on agriculture within the conservation area. Consequently, we further hypothesize that there will be significant variation among the communities within the district in the prevalence of syphilis, HIV and secondary infertility associated with STIs, and in associated maternal morbidity. Differences between the spatial distribution of HIV infection and the distribution of syphilis may reflect the relatively recent introduction of HIV to the district and the greater opportunities for transmission of syphilis and other STIs in recent decades. The project team works closely with local village health workers to identify and recruit pregnant women and to prevent significant levels of loss-to-follow-up. The village health workers are members of the communities in which they work, are well known and 3 respected by the community, and have done an excellent job generating interest among women in this study. We expect to achieve >90% follow-up of the women who have pregnancies in the one-year recruitment period. Data and research methods To date, over 600 pregnant wo men have been recruited into the study, interviewed, and screened for syphilis and HIV. During the coming year, the research team will recruit approximately 2000 additional pregnant women. The study cohort will consist of all pregnant women identified through pregnancy testing or attendance at one of 55 mobile outreach antenatal clinic sites conducted by health-care teams from Wasso and Endulen Hospitals. Recruitment will be ongoing for a 12- month period that began in late July 2002; follow- up interviews will be conducted with each participant at one and nine months post-delivery. Women will be stratified by mobile outreach clinic site and within each site by where her partner works (either inside or outside of Ngorongoro District). Preliminary data will be presented on causes and patterns of maternal morbidity and mortality, including results from syphilis and HIV screening, descriptions of pregnancy and delivery outcomes, and patterns of partner migration for the generation of nonpastoral income. The research team will employ a range of qualitative and quantitative methods including, structured and semi-structured individual and group interviews; specimen collection for pregnancy testing, syphilis testing, and anonymous HIV testing; and observation. A series of approximately 30 focus- group discussions with Maasai and Watemi women and men will be held at different stages of the longitudinal study in order to explore sociocultural factors associated with severe maternal morbidity and mortality. The first set of focus groups will be conducted to develop the cultural context for the longitudinal study. The design of interviews with women and (separately) with their partners will be informed by assessment of information including the following topics: How do women and their partners recognize signs and symptoms of pregnancy; what are marital taboos on sexual contact during pregnancy; what are perceived benefits of antenatal care; what is the optimal timing for initiation and cessation of antenatal care; what are perceived as normal conditions occurring during pregnancy (e.g., nausea, bleeding, shortness of breath); what are perceived as abnormal or problematic conditions during pregnancy; for what problems should medical help be sought; what might be disadvantages of medical care during pregnancy; what traditional remedies are used to treat pregnancy-related conditions; what preventive, curative, or palliative practices are desirable during childbirth, and what purpose do they serve; what reasons exist for avoidance of delivery in hospital or with medical attendance. Midway through the study period, a second set of focus-group discussions will be held for purposes of clarification and gathering additional information on the same subjects. Observations from data collection in the interviews and medical record review will suggest other topics for group discussion. Pregnant women and their partners will be asked about their immediate experiences during the current stage of pregnancy. Issues related to STIs, domestic violence, and other relevant issues will be explored as 4 appropriate. Women's and men's experiences and expectations of childbirth and the postpartum period will be explored in these groups. At the end of the study period, a final round of focus groups will serve the purpose of confirming sociocultural understandings, encouraging participants to offer additional observations of their own on related topics, and probing topics that might remain ambiguous or suggestive of hypotheses about causes of severe pregnancy-related morbidity and mortality. Expected findings Expected study results include the following: · Increased understanding of sexual/reproductive attitudes and behaviors, especially among the Maasai and Watemi populations of Ngorongoro District, Tanzania; · Increased understanding of health-seeking behaviors and outcomes; · Documentation of the patterns and causes of a range of maternal morbidities, including abortion, pregnancy and delivery-related complications, syphilis, HIV, and domestic viole nce during pregnancy; · Increased understanding and documentation of the patterns and causes of maternal mortality; · Increased understanding of the effect of migration on sexual/reproductive morbidity, including transmission of HIV; · Thorough documentation of the reproductive health services provided by the mobile outreach clinics and recommendations for service-delivery improvement; and · Increased understanding of the role and capacities of village health workers. There are potentially many challenges to working in Ngorongoro District, due to the remote and dispersed nature of the human population. However, according to hospital statistics the mobile outreach clinics reach the vast majority of antenatal women who attend services with great regularity. There is also exceptionally high attendance at vaccination clinics at one and nine-months post-delivery. However, the majority (up to 75%) of women deliver at home. The excellent network of mobile outreach clinics provides a unique opportunity to collect and report data from a remote, dispersed pastoral community setting. The mobile clinics will also provide an exceptional opportunity to provide sexual and reproductive health interventions recommended by the study investigators as a result of their findings. References 1Bureau of Statistics, DSM. (1990). Population Census Regional Profile Arusha1988. 2Bergstrom S. (1994a). Maternal health: a priority in reproductive health. In: Lankinen KS, Bergstrom S, Makinela PH, Petromaa M., eds. Health and disease in developing countries. London: MacMillan Education Ltd.