Statistical methods for analyzing complex survey data: an analysis of ITN and IPTp usage in Kenya
Introduction: Malaria still remains one of the leading causes of mortality and morbidity in Kenya despite major interventions such as ITNs (Insecticide Treated Nets) and IPTp (Intermittent Preventative Treatment of malaria in pregnancy) for expectant mothers in malaria endemic regions of Kenya. Methods: The purpose of the study was to explore IPTp and ITN usage among the expectant women in Kenya using data from the KDHS of2008. It examined two variables; one related to ITN usage and the other related to IPTp usage. The study did both logistic and survey logistic regressions on the two variables against a number of selected covariates. The study also went further to do a generalized ordered survey logistic regression on a newly generated variable which combined both lIN and IPTp against the same selected covariates. Major significant findings: The region where a woman comes from is very significant in establishing whether she uses ITNs, IPTp or both. Western, Nyanza and Coast regions have the highest usage of ITN andlor IPTp. However, both ITN and IPTp usage was generally low in all the regions. The target, especially for IPTp, remains largely underachieved in the country. The age of a woman is also quite a significant covariate in IPTp and ITN usage. Partner's level of education is a significant covariate in explaining ITN usage. Ethnicity is the least significant of all the covariates dealing with ITN and IPTp usage. Conclusion Compliance is a major hindrance for IPTp usage among expectant women. Many women are still not going for antimalarial doses during antenatal clinic visits. ITN usage gives better results than IPTp. However, the usage of ITN together with IPTp, which is the ideal recommendation, still suffers a low uptake in Kenya.