Choosing where to deliver: decision criteria among women with low-risk pregnancies in France

As in the rest of Europe, the supply of maternity hospitals has progressively decreased over the past few decades in France. An understanding of user choice criteria is important to help health planners reorganize obstetrical services and to predict changes in utilization patterns in response to supply changes. The objectives were to understand the criteria that women use to select their maternity hospital in France and to analyse the relation to individual and community characteristics. A survey of 536 recently delivered women with low-risk pregnancies explored the factors motivating user choice in three territories with distinct geographical and health service supply characteristics: four districts in Burgundy, two districts in Pays de la Loire, and the district of Seine-Saint-Denis in Ile-de-France. Women were asked to select a principal choice criterion. Their responses were grouped into categories: Accessibility/proximity, reputation of the establishment among users, advice of treating physician, technical quality and cost. Accessibility and proximity were the most selected criteria (33%), followed by the reputation of the maternity (29%), technical quality (15%) and advice of treating physician (13%). Age, parity and education influenced choice criteria. After controlling for individual determinants, region of residence was highly related to choice criteria; women living in Burgundy were more likely to select an establishment based on proximity, in Seine-Saint-Denis women were more likely to follow the advice of their physician, and in Pays de la Loire, more likely to base their decisions on the reputation of the establishment. The association between choice criteria and community characteristics could explain the failure of previous models to predict behaviour in different contexts. It is important to carry out local surveys of user perceptions before restructuring in order to take into consideration women's opinions on their future places of delivery and to refine geographic models.