BACKGROUND OF THE STUDY

Approximately 1000 women die each day worldwide from pregnancy related causes, 99% of them in developing countries and more than 50% in sub-Saharan Africa with most deaths concentrated around the time of delivery. An estimated 2.65 million stillbirths occurred in 2008 worldwide while 3 million new-borns do not survive the first month of life worldwide annually . Skilled assistance during childbirth, readily accessible appropriate care in case of complications and effective postnatal care within the first 24 hours of delivery are strategies that can improve perinatal outcomes for mothers and babies. A key strategy to reducing maternal and neonatal deaths is the ‘health-centre intrapartum care strategy’, where qualified skilled workers manage labour, effectively manage complications and are supported with effective referral systems for specialised care when needed, and an effective postnatal care package.

A significant proportion of mothers in developing countries still deliver at home unattended by skilled health workers. In diverse contexts, individual factors including maternal age, parity, education and marital status, household factors including family size, household wealth, and community factors including socioeconomic status, community health infrastructure, region, rural/urban residence, available health facilities, and distance to health facilities determine place of delivery and these factors interact in diverse ways in each context to determine place of delivery]. Eijk et al. looked at antenatal care and delivery care among women in Western Kenya and demonstrated that older women, high parity, lower socioeconomic status, low education levels and more than an hour walking distance were associated with delivery outside health facilities. Studying poor urban dwellers in Nairobi, Fosto et al. found from bivariate analyses that wealth, education, parity, place of residence were associated with place of delivery. Ochako has previously demonstrated that these factors together with marital status and age at birth of last child determined use and timing of first Antenatal Care (ANC) visit and type of delivery. There are also wide variations in the reasons women give for delivering at home between and within countries . For Kenya, recent studies looking at the degree of effect of such factors are lacking.

In Kenya, maternal mortality rate has not reduced over recent years, and may even have increased from an estimated 380/100000 live births in 1990 to 530/100000 live births in 2008 . Although a number of factors may have contributed to this, including improved identification of maternal deaths, health facility delivery remained low at 44% and 42.6% in the early 1990s and in 2008 respectively [18, 19]. Recent evidence on determinants of place of delivery in Kenyan utilising a nationally representative data and controlling for all factors is lacking, yet understanding the influences on place of delivery in Kenya is crucial to identifying key priority areas for policy and practise to increase the prevalence of skilled assisted deliveries.

We have used data from the 2008/2009 Kenya Demographic and Health Survey (KDHS) and linked them with a 2008 Kenyan Health Facility Database, that provides Global Positioning System (GPS) coordinates for distance analysis, to describe the factors that influence where women deliver in Kenya, and the reasons that women give for delivering at home.