We searched electronic databases (PubMed and OVID's EMBASE, Global Health, Medline and Health Management Information Consortium) to identify review articles of determinants of delivery-service use (not original articles). We combined search terms related to maternal health care (obstetric delivery, parturition, home childbirth, reproductive health services) with search terms related to service use (health service utilisation, accessibility, medically underserved area, rural health services, choice behaviour) as well as using a term combining both (maternal health services utilisation; detailed search strategy available on request). The two authors screened independently for relevance and only two review articles on determinants of delivery service use in low- or middle-income countries were found: Thaddeus and Maine [7 (link)], and Say and Raine [9 (link)]. All other hits were either not in fact review articles or focussed on different topics. We used the two systematic review articles to identify individual studies and main themes. Referenced articles as well as articles referencing these reviews were read and checked to ascertain further relevant literature. Both qualitative and quantitative studies investigating determinants of skilled attendance at delivery or of facility delivery were included. No constraints were placed on date or language. The overview of the types of determinants studied, their suggested pathways and the typical findings is comprehensive, however, the studies mentioned are exemplary but not exhaustive of the vast literature published on each determinant. Over 80 studies were read but not all are cited in this article.
The breadth of topic, its context-specificity, the lack of comprehensive index terms and the vast differences in methodology employed, rendered the option of doing our own systematic review of this literature in its entirety impractical. Systematic reviews of observational data are useful when trying to estimate an effect of interest that can be assumed to be independent of context (which is true for most biological effects) or when trying to explore heterogeneity that is thought to be due to a limited range of factors. It is only feasible when looking at a narrow range of clearly defined exposures. Our aim instead was to explore the range of what has been done in the field so far and give an overview of findings, rather than estimating any specific effect or even attempt a meta-analysis. While we could have restricted the review to a limited number of exposures, years or countries, this went contrary to our desire to work out the scope of what has been explored in the literature.
The breadth of topic, its context-specificity, the lack of comprehensive index terms and the vast differences in methodology employed, rendered the option of doing our own systematic review of this literature in its entirety impractical. Systematic reviews of observational data are useful when trying to estimate an effect of interest that can be assumed to be independent of context (which is true for most biological effects) or when trying to explore heterogeneity that is thought to be due to a limited range of factors. It is only feasible when looking at a narrow range of clearly defined exposures. Our aim instead was to explore the range of what has been done in the field so far and give an overview of findings, rather than estimating any specific effect or even attempt a meta-analysis. While we could have restricted the review to a limited number of exposures, years or countries, this went contrary to our desire to work out the scope of what has been explored in the literature.
Free full text: Click here
