Menu
Menu
Menu
Menu
Background: In sub-Saharan Africa, the literature on end of life is limited and focuses on place of death as an indicator of access and utilization of health-care resources. Little is known about population mobility at the end of life. Objective: To document the magnitude, motivations and associated factors of short-term mobility before death among adults over 15 years of age in Burkina Faso and Senegal. Methods: The study was based on deaths of adult residents reported in three Health and Demographic Surveillance System (HDSS) sites in urban (Ouagadougou) and semi-rural areas (Kaya) of Burkina Faso, and rural areas of Senegal (Mlomp). After excluding deaths from external causes, the analysis covered, respectively, 536 and 695 deaths recorded during the period 2012–2015 in Ouagadougou and Kaya. The period was extended to 2000–2015 in Mlomp, with a sample of 708 deaths. Binary logistic regressions were used to examine the effects of socio-demographic characteristics on place of death (health facility or not) and location of death (within or outside the HDSS). Results: In Mlomp, Kaya and Ouagadougou, respectively 20.6%, 5.3% and 5.9% of adults died outside the HDSS site. In Mlomp and Kaya, these deaths were more likely to occur in a health facility than deaths that occurred within the site. The reverse situation was found in Ouagadougou. Age is the strongest determinant of mobility before death in Mlomp and Kaya. In Mlomp, young adults (15–39) were 10 times more likely to die outside the site than adults in the 60–79 age group. In Ouagadougou, non-natives were three times more likely to die outside the city than natives. Conclusions: At the end of life, some rural residents move to urban areas for medical treatment while some urban dwellers return to their village for supportive care. These movements of dying individuals may affect the estimation of urban/rural mortality differentials.
In the context of SSA, data collected in censuses and standard household-based surveys such as the Demographic and Health Surveys (DHS) do not allow a detailed analysis of rural–urban migration in relation to adult health [27,28]. First, these surveys do not collect often information on migration flows between urban and rural areas. Second, the focus is on child and maternal health rather than on adult health. Finally, the few data available on migration concern long-term mobility leading to change in a place of residence. So, capturing short-term mobility and its relation to health is a particular challenge. The data used here come from three HDSSs of Burkina Faso and Senegal. They will serve as case studies to explore mobility before death in different West African contexts. The first one is the Ouagadougou HDSS in Burkina Faso, which is the only fully urban HDSS of West Africa. Second, the Kaya HDSS will allow an analysis of mobility before death in a semi-rural context of Burkina Faso. The Kaya HDSS is covered by a regional hospital. Finally, the Mlomp HDSS located in rural Senegal offers a contrasting context compared to that of Kaya because the site is covered by a primary health-care facility only. Furthermore, in this HDSS site in particular, information is collected on whether mobility is related to health-care-seeking. Such data allow a better understanding of the reasons for mobility before death in rural areas. While other HDSSs sites exist in West Africa, only data from these three sites were accessible and/or suitable for the analysis conducted here. The three HDSS sites are members of the International Network for Demographic Evaluation of Populations and Their Health (INDEPTH) and share similarities in terms of methodology [29]. Following an initial census in the area under surveillance, fieldworkers conduct regular household update rounds, and register vital events (births and deaths, migrations and marriages). In case of death, a verbal autopsy (VA) questionnaire is completed with the next of kin to determine the circumstances that led to the death, including the history of the illness and the specific symptoms that preceded death. VA data can be interpreted by physicians or using computer-based methods. In the first method, each VA is reviewed separately by two distinct physicians to determine the probable cause of death. In case of disagreement, the VA is reviewed by a third physician who assigns a consensus diagnosis. Otherwise, the cause of death is categorized as ‘indeterminate’ [30]. The most widely used automated method on INDEPTH sites is the one based on Bayes’ theorem implemented in the InterVA-4 software [31]. The Ouagadougou HDSS was established in late 2008 in five neighbourhoods at the northern periphery of the capital city of Burkina Faso [32]. Two of them (Kilwin and Tanghin) are formal neighbourhoods with full access to public services, while the other three (Nonghin, Polesgo and Nioko 2) are informal settlements (like slums) without access to such services [32]. People living in the Ouagadougou HDSS are mostly from the Mossi ethnic group (90%), which is currently the majority ethnic group in the country. More than half of active adults work in the commerce and construction sectors [33]. The population under surveillance in the Ouagadougou HDSS totalled about 90,000 residents in 2015 and periodic household update rounds are conducted with an average periodicity of 10 months. VA data are interpreted by the InterVA-4 software to determine the probable causes of death. Health-care provision in the city of Ouagadougou is better than in any other location in Burkina Faso, with a private sector representing two-thirds of all care services. In addition, large teaching hospitals offering the country’s highest standards of care are located in the city [34]. The Kaya HDSS was established in late 2007 in the North Central region of Burkina Faso, 100 kilometres from the capital city, Ouagadougou. The site covers the town of Kaya and 18 villages [35]. The follow-up population was estimated at 70,000 inhabitants in 2015. This population lives in semi-urban (70%) and rural (30%) areas. The site is easily accessible from Ouagadougou and is covered by seven health facilities and one regional hospital. Residents are mostly from the Mossi ethnic group and are of Muslim faith. Only half of the population have been to school, and the main economic activities are small-scale agriculture and livestock breeding. In recent years, gold mining in the neighbouring villages of the HDSS has grown in scale. Although the site covers the town of Kaya, health indicators and fertility levels are typical of a rural area of Burkina Faso. Life expectancy was estimated at 54 years in 2013 and the total fertility rate was estimated at 7 children per woman. Households are visited every six months. In case of death, causes of deaths are certified by physicians based on information available in VA questionnaires. During the period considered in this study (2012–2015), a large share of VA questionnaires was not completed and available ones were not yet diagnosed by physicians. However, lay reporting of causes of death was available. The Mlomp HDSS was set up in 1985 in the Southwest Senegal in the administrative region of Ziguinchor, nearly 500 kilometres from Dakar, the capital city [36]. The site covers 11 villages. The population under surveillance belongs to the Diola ethnic group and is mostly animist or Catholic. Rice cultivation is the main activity in the area but many adults migrate during the dry season, with men leaving to find work in wine palm harvesting and fishing in other regions. Young women are often employed as domestic servants in Dakar or in Banjul (the capital city of Gambia) before they get married. The educational level is relatively high in Mlomp with respect to other rural areas of Senegal. In the 2000s, while only a minority of women aged 15–49 years in rural Senegal as a whole had attended school, around half done so in Mlomp. Health indicators are also encouraging, thanks to a very dynamic private health centre opened in 1961 by French Catholic nurses. However, to see a physician, patients must be referred to the local hospital at Oussouye, 10 kilometres from Mlomp. Advanced medical care including surgery is only available in the larger regional hospital at Ziguinchor, 50 kilometres from Mlomp. The follow-up population was estimated at 9000 inhabitants in 2015 and vital events are updated on an annual basis. In case of death, physicians interpret the completed VA questionnaires to assign a probable cause of death. Two outcome variables were considered in this analysis. First, place of death was grouped into two main categories: health facility versus non-health facility. We did not make any distinction between the types of health facility; for example, public versus private. The ‘non-health facility’ category mainly included deaths that occurred at home. Deaths that took place elsewhere or for which information was not available represent 5.1% of deaths in Ouagadougou, 2.4% in Kaya and 2.1% in Mlomp. Second, mobility before death was defined using information on location of death, i.e. if the death has occurred within or outside the HDSS. For the particular case of the Ouagadougou HDSS, to ignore mobility within the city, location of death was defined on the basis of the entire city. Individuals were classified into two categories: those who died in Ouagadougou and those who died elsewhere. Furthermore, for deaths at Mlomp, it was possible to know if the deceased person had left the site to seek health care or not. Independent variables included sex, age group at death, education, marital status, birthplace and group of causes of death. We categorized the different variables in this way to ensure comparability across the three sites. Age of death was divided into four categories, 15–39, 40–59, 60–79, and 80 years and more. For education, two main categories were considered: individuals with no schooling, and those with at least one year of schooling. We defined three marital-status categories: married, single, divorced/widowed. In some cases, data on education and marital status were missing. These cases were coded as ‘unknown’. Place of birth was taken into account only in the Ouagadougou HDSS, and two categories were defined: native of Ouagadougou and non-native. Finally, causes of death recorded in the sites of Ouagadougou and Mlomp were aggregated into three main categories, excluding deaths from external causes: malaria, HIV/AIDS, respiratory infections and other infections were classified as communicable diseases; diseases such as neoplasms, diabetes, stroke and other chronic diseases were grouped as ‘non-communicable diseases’; and indeterminate causes of deaths and deaths for which there was no VA were classified as ‘ill-defined’. In this study, we analyse adult deaths after 15 years of age. The analysis covers the period 2012–2015 in Ouagadougou and Kaya. Since the population of Mlomp is much smaller, the analysis was extended to cover deaths over the period 2000–2015. In Ouagadougou and Kaya, a six-months criterion is used to define residency in the HDSS, i.e. individuals are excluded from the follow-up after six months of absence. This is not the case in Mlomp, due to a high volume of circular migrations. Individuals are excluded from the follow-up only after two successive years of absence. In order to approximate the same residence criteria in Mlomp as in the two other sites, the date of the most recent presence of the residents who died was compared to the date of their death. When the precise date of departure from the village was missing, it was estimated on the basis of information on the person’s presence or absence during the dry and the rainy seasons recorded in the two last follow-up surveys. The deaths of individuals who reported as absent from the HDSS area more than six months are then excluded from the analysis. As the analysis aims to highlight mobility before death for health reasons, deaths from external causes were excluded (8.4% of deaths in Ouagadougou, 4.5% in Kaya and 9.5% in Mlomp). In the Kaya site, causes of death based on VAs were not available for the period considered, so lay reporting of causes of deaths (disease, accident, suicide, murder, pregnancy-related deaths), by the relatives of the deceased person, was used to exclude deaths from external causes. To sum up, the analysis included 536 eligible deaths in Ouagadougou, 695 in Kaya and 708 in Mlomp. In Mlomp, out of 809 deaths, 101 were discarded because the persons had been away from the site more than six months before their death. Two sets of analyses were performed for each site. In a first step, in order to investigate the reasons for ultimate mobility, the net effect of the location of death on the place of death (in health structure or not) was assessed using a binary logistic regression. Covariates included sex, age group at death, marital status, education, and group of causes of death. In the second analysis, location of death was the outcome variable to determine factors associated with mobility before death. Its association with independent variables (sex, age group at death, marital status, education, and group of causes of death) was tested again using a binary logistic regression. For the particular case of Ouagadougou, place of birth was also included in the model to examine the effects of migration status (native or not) on location of death. All analyses were performed using STATA software, version 14.
N/A
