Birth order is the most consistent predictor of poor birth outcomes in Kenya, where women having their first child are about twice as likely as others to deliver prematurely, have a baby who is smaller than average and require a cesarean section. Other key factors influencing the odds of having a premature birth are related to the extent to which women use prenatal care services, while women's nutritional status is one of the most important factors in their infant's size at birth; the odds of cesarean section, meanwhile, are affected by women's socioeconomic status and contraceptive practice. These are among the chief findings of an analysis of data from the 1993 Kenya Demographic and Health Survey.¹

The survey gathered information on women's background characteristics, as well as on pregnancies and births that occurred during the previous five years. A total of 5,295 births for which complete information was available were included in the analyses of factors associated with adverse birth outcomes. According to the women's accounts, 4% of these births were premature, 5% of the deliveries were by cesarean and 15% of the infants were smaller (i.e., weighed less) than average.

In bivariate analyses, the researchers identified a wide array of factors related to women's socioeconomic background, reproductive history, health care, nutritional status and biological characteristics that were associated with the risk of poor birth outcomes. They then used multilevel regression analysis to estimate the independent effects of these factors on women's odds of having a premature birth, a smaller-than-average baby or a cesarean section.

A woman's use of prenatal care and the quality of services played a large role in her risk of having a premature birth. Women who visited a prenatal care provider only once or twice had considerably higher odds of this outcome than those who made seven or more visits (odds ratio, 5.1). Furthermore, those who had at least one tetanus shot (which suggests good-quality care) had substantially reduced odds of delivering prematurely (0.3). The odds also were lower among women who delayed care until the second or third trimester (0.5 and 0.2, respectively) than among those who first saw a provider in the first trimester, a result that the researchers note probably reflects early initiation of care by women with pregnancy complications.

A number of other factors also were strongly associated with the risk of premature delivery. The odds of this outcome were significantly elevated among women having a first birth (2.3), those having a multiple birth (7.0) and those who were members of the Luo (as compared with the Kikuyu) ethnic group (7.1). Relative to residents of the Central Province, women who lived in Nairobi or Nyanza Province had sharply lower odds of delivering prematurely (0.3 and 0.2, respectively).

Some of the same factors affected women's likelihood of bearing a baby who, by their report, was smaller than average. The odds of this outcome were elevated among women having a first birth (1.8) or a multiple birth (3.1), and were reduced among those who had had at least one tetanus injection (0.6). Region of residence again played a role; women from the Western Province had increased odds of bearing a small infant (1.8). In addition, women who bore a girl were more likely than mothers of boys to say that the baby had been small (1.8), and those who scored low on a scale assessing women's weight for their height (an indication of poor nutrition) had elevated odds of this outcome (1.5).

For the final outcome examined, cesarean delivery, women having a first birth were again at greater risk than those who had given birth before (odds ratio, 2.2). The odds of this outcome were almost doubled (1.5-1.9) among women of high (as opposed to medium) socioeconomic status, those aged 30-34 (compared with those in their early 20s) and those who had ever used a modern contraceptive (as opposed to those who had never practiced family planning). Compared with women whose height was in the range of 150-160 cm, shorter women had elevated odds of cesarean delivery (2.5), and taller women had reduced odds (0.7). The increased risks among women of high socioeconomic status and those who had used modern methods of contraception, the authors note, are not surprising, since these women were more likely than others to have had access to--and to have used--appropriate health facilities.

At the district level, the probability of cesarean delivery varied significantly. On average, cesarean deliveries occurred less frequently than would be expected in districts in the Western, Nyanza and Coast Provinces, and more frequently than expected in Nairobi and the Central Province districts.

The investigators note that their data are limited, since they reflect women's perceptions of their birth outcomes, rather than objective measures. Nevertheless, the researchers conclude that the findings point up the need for integrated maternal health programs that include prenatal and delivery care, as well as nutrition interventions. Furthermore, the analysts stress, it is "crucial" that women having a first birth get appropriate care, given their increased odds of all three adverse outcomes studied. And since regional variations were apparent for each outcome, it is also important for programs to be "sensitive to regional disparities."--D. Hollander

REFERENCE

1. Magadi M, Madise N and Diamond I, Factors associated with unfavourable birth outcomes in Kenya, Journal of Biosocial Science, 2001, 33(2):199-225.