METHODS: The 2000 round of demographic and health survey (DHS) data sets for the countries are used. The study population are women aged 15-49 who have ever been exposed to the risk of pregnancy and interviewed during the surveys. Univariate analysis is done to examine the distribution of the women. Bivariate analysis of selected background, contraceptive, sexual behaviour and fertility variables by having ever had a terminated pregnancy are also carried out. Lastly, multivariate logistic regression models are fitted to examine the magnitude of associated variables while controlling for others.

RESULTS: the associated factors vary from country to country. However, age and marital status are associated with and are critical predictors of having ever had a terminated pregnancy in all the countries. For example, the odds of having ever had a terminated pregnancy is about 54 and 32 times (for currently and formerly married women, respectively) that of those that have never married in Rwanda.

CONCLUSION: This study has come up with findings that provoke more thinking and research on the topic of pregnancy termination and to reproductive health in general. This is because ‘ever had a terminated pregnancy’ as used in the DHS context made no distinction between the three different pathways through which a pregnancy may not result into a live birth.