In rural Mali, where few women obtain maternal health care, use of prenatal and delivery services is influenced by both individual and community factors, according to a cross-sectional study of women who had recently given birth.1 The odds of receiving prenatal care in the first trimester, four or more prenatal visits, attended delivery and institutional delivery are all negatively related to the number of personal barriers to care and positively related to the level of household wealth, whereas women's odds of receiving care are elevated if their neighbors have medium or high levels of use of prenatal care. Overall, although factors associated with types of care vary, individual factors explain more of the variation in receipt of care than do community factors.
The study used data from the 2001 Mali Demographic and Health Survey on rural women aged 15–49 who had had a live birth in the preceding five years. Characteristics of the women's communities were obtained from the DHS community and household questionnaires; characteristics of the women themselves, including prenatal and delivery care for their most recent live birth, were obtained from the women's questionnaire.
Analyses were based on 6,178 live births to women living in 264 rural communities. Overall, 15% of the births had benefited from prenatal care in the first trimester; 22%, from four or more prenatal visits; 29%, from attendance by a trained medical provider; and 26%, from delivery in a medical facility.
The births took place in communities that had, on average, only one medical facility within five kilometers; only a tenth occurred within five kilometers of a facility offering emergency obstetric care. For seven in 10, there was no emergency auto transportation available and the nearest public transportation was at least 15 minutes away. Nearly four in 10 births occurred at least 10 kilometers from the nearest source of prenatal care or delivery care. One-half took place in communities where the level of education was low; one-seventh, in communities with a high proportion of poor households; and more than one-third, in communities with low uptake of prenatal care.
Fully 90% of births were to women who did not have any education. The current birth was their fifth, on average. More than eight in 10 women were long-term residents, having lived in their community for at least five years. Half had not received any prenatal care, slightly more than a third had received care but were not counseled about pregnancy complications, and the rest received care with counseling.
In bivariate analyses, personal barriers had a consistent relationship to receipt of medical care. Regardless of the barrier—whether it was not knowing where to go for care, having to get permission to go, difficulty obtaining money for treatment, distance to a medical facility, finding transportation, not wanting to go alone or concern about not being able to have a female provider—women who reported that the barrier was a "big problem" were significantly less likely to have received each of the four types of maternal health care than were their counterparts who reported that the barrier was "no problem."
In terms of community factors, multivariate analyses showed that women's odds of receiving prenatal care in the first trimester rose with the number of health facilities within five kilometers (odds ratio, 1.1) and were higher in communities having a medium or high level of prenatal care uptake rather than a low level (2.5–3.9). Women's odds of receiving at least four prenatal visits were elevated if the nearest public transportation was less than 15 minutes away instead of more (1.4) and if their neighbors' uptake of prenatal care was medium or high rather than low (3.5–6.3); the odds were reduced if their community had a high rather than low level of education (0.4) and a medium rather than low percentage of women of the same ethnicity (0.7).
Women's odds of having a delivery attended by trained medical personnel were elevated if their neighbors' uptake of prenatal care was medium or high (odds ratio, 1.8–5.4); the odds were reduced if the nearest source of delivery care was 1–29 kilometers away instead of within the community (0.3–0.5). Finally, women were more likely to deliver in a medical facility if their neighbors' uptake of prenatal care was medium or high (1.8–8.4), whereas they were less likely to do so if the nearest source of delivery care was 5–29 kilometers away (0.4–0.5).
In terms of individual-level factors, multivariate analyses showed that women's odds of obtaining prenatal care in the first trimester and their odds of making four or more prenatal visits were positively associated with household wealth (odds ratios, 1.4 and 1.5, respectively), but negatively associated with the number of personal barriers to receiving medical care (0.9 each). The odds were elevated if women had any education instead of none (1.6 and 1.5) and if they were short-term residents and had previously lived in an urban area, as compared with short-term residents who had previously lived in a rural area (1.6 and 1.9). For prenatal care in the first trimester, the odds were negatively associated with the number of children in their household younger than five (0.9) and with birth order (0.9).
Women's odds of having a delivery attended by trained medical personnel rose with household wealth (odds ratio, 1.6); they were elevated if the women were short-term residents and had previously lived in an urban area (2.0), and if they had received prenatal care along with counseling about pregnancy complications, as compared with prenatal care without counseling (1.4). However, the odds were negatively associated with number of personal barriers to medical care (0.9) and were reduced if women had not received any prenatal care (0.2). The same factors were associated with the odds of having an institutional delivery.
Final analyses looked at how well the measured factors explained why women's odds of receiving maternal health care differed. Individual factors explained a larger proportion of the variation than community factors across all four measures of maternal health care use; however, the contribution of individual factors was greater for prenatal care in the first trimester (ratio of community-level variance to total variance, 0.1) and for four or more prenatal visits (0.2) than for delivery attendance by a trained medical provider (0.4) or delivery in a medical facility (0.4). Even after individual and community factors were taken into account, differences in use between communities remained significant, suggesting that unmeasured factors were also at play.
The author suggests that possible strategies for increasing use of maternal health care in this population might include improving public transportation, adding more medical facilities, encouraging women living in remote areas to move closer to medical facilities when their due date approaches, ensuring that the poor have access to health care, empowering women socially and economically, teaching maternal health in schools and adult programs, and engaging women's partners and families in discussions about appropriate care during pregnancy and delivery. "As barriers to the utilization of maternal health services are multilevel, comprehensive (and multisectoral) programs are needed in rural communities…," she concludes.—S. London
1. Gage AJ, Barriers to the utilization of maternal health care in rural Mali, Social Science & Medicine, 2007, 65(8):1666–1682.