Abstract
BACKGROUND: The importance of antenatal care (ANC) for improving perinatal outcomes is well established. However access to ANC in Kenya has hardly changed in the past 20 years. This study aims to identify the determinants of attending ANC and the association between attendance and behavioural and perinatal outcomes (live births and healthy birthweight) for women in the Kwale region of Kenya. METHOD: A Cohort survey of 1,562 perinatal outcomes (response rate 100%) during 2004-05 in the catchment areas for five Ministry of Health dispensaries in two divisions of the Kwale region. The associations between background and behavioural decisions on ANC attendance and perinatal outcomes were explored using univariate analysis and multivariate logistic regression models with backwards-stepwise elimination. The outputs from these analyses were reported as odds ratios (OR) with 95% confidence intervals (CI). RESULTS: Only 32% (506/1,562) of women reported having any ANC. Women with secondary education or above (adjusted OR 1.83; 95% CI 1.06-3.15) were more likely to attend for ANC, while those living further than 5 km from a dispensary were less likely to attend (OR 0.29; 95% CI 0.22-0.39). Paradoxically, however, the number of ANC visits increased with distance from the dispensary (OR 1.46; 95% CI 1.33-1.60). Women attending ANC at least twice were more likely to have a live birth (vs. stillbirth) in both multivariate models. Women attending for two ANC visits (but not more than two) were more likely to have a healthy weight baby (OR 4.39; 95% CI 1.36-14.15). CONCLUSION: The low attendance for ANC, combined with a positive relationship between attendance and perinatal outcomes for the women in the Kwale region highlight the need for further research to understand reasons for attendance and non-attendance and also for strategies to be put in place to improve attendance for ANC.