Abstract
The causes of maternal death are well known, and are largely preventable if skilled health care is received promptly. Complex interactions between geographic and socio-cultural factors affect access to, and remoteness from, health care but research on this topic rarely integrates spatial and social sciences. In this study, modeling of travel time was integrated with social science research to refine our understanding of remoteness from health care. Travel time to health facilities offering emergency obstetric care (EmOC) and population distribution were modelled for a district in eastern Indonesia. As an index of remoteness, the proportion of the population more than two hours estimated travel time from EmOC was calculated. For the best case scenario (transport by ambulance in the dry season), modelling estimated more than 10,000 fertile aged women were more than two hours from EmOC. Maternal mortality ratios were positively correlated with the remoteness index, however there was considerable variation around this relationship. In a companion study, ethnographic research in a subdistrict with relatively good access to health care and high maternal mortality identified factors influencing access to EmOC, including some that had not been incorporated into the travel time model. Ethnographic research provided information about actual travel involved in requesting and reaching EmOC. Modeled travel time could be improved by incorporating time to deliver request for care. Further integration of social and spatial methods and the development of more dynamic travel time models are needed to develop programs and policies to address these multiple factors to improve maternal health outcomes.