Institute for Health Metrics and Evaluation, Department of Global Health, University of Washington, 2301 5th Avenue #600, Seattle, WA 98121, USA.
Soc Sci Med. 2013 Sep;93:147-54. doi: 10.1016/j.socscimed.2013.06.012. Epub 2013 Jun 20.
Rates of neonatal and maternal mortality are high in Ghana. In-facility delivery and other maternal services could reduce this burden, yet utilization rates of key maternal services are relatively low, especially in rural areas. We tested a theoretical implication that travel time negatively affects the use of in-facility delivery and other maternal services. Empirically, we used geospatial techniques to estimate travel times between populations and health facilities. To account for uncertainty in Ghana Demographic and Health Survey cluster locations, we adopted a novel approach of treating the location selection as an imputation problem. We estimated a multilevel random-intercept logistic regression model. For rural households, we found that travel time had a significant effect on the likelihood of in-facility delivery and antenatal care visits, holding constant education, wealth, maternal age, facility capacity, female autonomy, and the season of birth. In contrast, a facility's capacity to provide sophisticated maternity care had no detectable effect on utilization. As the Ghanaian health network expands, our results suggest that increasing the availability of basic obstetric services and improving transport infrastructure may be important interventions.
加纳的新生儿和产妇死亡率居高不下。在机构内分娩和其他产妇服务可以减轻这一负担，但关键产妇服务的利用率相对较低，尤其是在农村地区。我们验证了一个理论假设，即旅行时间会对在机构内分娩和其他产妇服务的使用产生负面影响。从经验上看，我们使用地理空间技术来估计人口和卫生设施之间的旅行时间。为了考虑加纳人口与健康调查群集位置的不确定性，我们采用了一种新颖的方法，将位置选择视为一种插补问题。我们估计了一个多层次的随机截距逻辑回归模型。对于农村家庭，我们发现，在控制教育、财富、产妇年龄、设施能力、女性自主权和出生季节的情况下，旅行时间对在机构内分娩和产前护理就诊的可能性有显著影响。相比之下，医疗机构提供复杂产妇护理的能力对利用率没有明显影响。随着加纳卫生网络的扩大，我们的研究结果表明，增加基本产科服务的可及性和改善交通基础设施可能是重要的干预措施。
