A multi-level study of reproductive outcomes in Kenya: Quality of care and community factors.

摘要

The purpose of this study is to examine how the contextual factors interact with individual characteristics to explain the number of children ever born to women and their contraceptive use in Kenya in 1993 and 1998. The present study provides a new perspective and findings on this debate by investigating individual-level fertility outcomes in one country, Kenya, utilizing multi-level models applied to a unique data set integrating community-level development and family planning service quality of care indicators with nationally representative Demographic and Health Survey data from 1993 and 1998 and clinic-level data from 1989 and 1999 to test both direct and interactive effects of community context on changing fertility behaviors and outcomes.;The results of most models examined in this study illustrate the primary importance of an individual woman's characteristics, particularly her age at marriage and educational attainment, in influencing the number of children ever born to her and whether she ever used modern contraceptives. The level of infrastructural development in a district was the only community development factor to have a very significant and consistent inverse direct effect on the number of children ever born to a woman.;Although the cross-level interaction considering the effect of a woman ever using modern contraceptives and an increased level of infrastructural development in a district has a significant positive effect, the overall effect of an increased level of infrastructural development ultimately further reduced the number of children ever born to a woman. Most of the other community development factors have no significant main effect on a woman's decision to ever use modern contraceptives, especially in the later time period.;Most of the cross-level interactions with the level of infrastructural development in a district and a woman's educational attainment had a non-significant combined effect in both time points as well. Similarly, the measures of quality of care at service delivery points on their own have a limited effect on the number of children ever born to a woman and almost no effect on her decision to ever use modern contraceptives. However, the inclusion of cross-level interactions of the quality of care measures and whether a woman ever used modern contraceptives illustrated significant effects on the number of children ever born to a woman. The cross-level interactions of the choice of oral contraceptives at service delivery points in a district and a woman's educational attainment also illustrated the importance of family planning context on a woman's decision to ever use modern contraceptives.;The significance of a woman's individual characteristics is not surprising given the dependent variables are measured at the individual level. Thus, although there are only a few significant variables at the contextual level it is important to discuss how they relate to the ongoing debate of what factors influence a fertility decline. The subsequent section will discuss all the main effects and cross-level interactions in the context of current demographic scholarship.