Determinants of childbirth care quality along the care continuum in limited resource settings: A structural equation modeling analysis of cross-sectional data from Burkina Faso and Côte d’Ivoire

INTRODUCTION: Despite the important increase in in-facility births, perinatal mortality rates have remained high and slow to decrease in many developing countries. This situation is attributed to poor childbirth care quality. The reason why women delivering in health facilities do not always receive care of an adequate standard is unclear. We assessed the determinants of childbirth care quality along the care continuum by means of different approaches. METHODS: A health facility-based cross-sectional study with a direct observation of health care workers’ practices while caring for mother–newborn pairs was carried out in Burkina Faso and Côte d’Ivoire. The performance of a set of essential best practices (EBPs) was assessed in each birth event at the admission, prepushing and immediate postpartum stages. A quality score, in the form of the additive sum of EBPs effectively delivered, was computed for each stage. We used negative binomial regression models and a structural equation modeling analysis to assess the determinants of care quality at each stage and the relationships of the quality delivered at the different stages, respectively. RESULTS: A total of 532 and 627 mother–newborn pairs were evaluated in Burkina Faso and Côte d’Ivoire, respectively. In both countries, delivery care quality varied significantly at all stages between health districts. Predelivery care quality was consistently higher in referral hospitals than in primary health care facilities (incident rate ratio (IRR) = 1.02, p < 0.05, and IRR = 1.10, p < 0.05, respectively, for Burkina Faso and Côte d’Ivoire). Quality at admission was poorer among nurses than among midwives in Burkina Faso (IRR = 0.81, p < 0.001). Quality at the admission and predelivery stages was positively correlated with immediate postpartum care quality (β = 0.48, p < 0.001, and β = 0.29, p < 0.001, respectively). CONCLUSION: Quality improvement strategies must target both providers and health facilities, and different inputs are needed depending on the stage in the care continuum.