Identifying optimal labor and delivery nurse staffing: The case of cesarean births and nursing hours

Abstract

BackgroundNumerous studies have identified a relationship between nurse staffing and adverse patient outcomes in medical / surgical patient populations. However, little is known about the impact of labor and delivery (L&D) nurse staffing and adverse birth outcomes, such as unintended cesarean delivery, in low-risk term-gestation women. PurposeWe examined nurse staffing patterns on the likelihood of cesarean sections (C-sections) among low- risk, full gestation births and provided a testing framework to distinguish optimal from ineffective levels of nurse staffing. MethodsThis retrospective descriptive study used hours of productive nursing time per delivery as the treatment variable to determine direct nursing time per delivery and its impact on the likelihood of a C-section. For comparisons, we also assessed the likelihood of augmentations and of inductions, as well as the number of neonatal intensive care unit (NICU) hours per birth. We limited our sample to those births between 37 and 42 weeks of gestation. Two complimentary models (the quadratic and piecewise regressions) distinguishing optimal staffing patterns from ineffective staffing patterns were developed. The study was implemented in eleven hospitals that are part of a large, integrated healthcare system in the Southwest. DiscussionWhile a simple linear regression of the likelihood of a C-section on nursing hours per delivery indicated no statistically distinguishable effect, our ‘optimal staffing’ model indicated that nurse staffing hours employed by using a large sample of hospitals were actually minimizing C-sections (robustness checks are provided using similar model comparisons for the likelihood of augmentation and induction, and NICU hours). Where the optimal staffing models did not appear to be effective for augmentations, inductions, and NICU hours, we found significant differences between facilities (i.e., significant fixed effects for hospitals). In all specifications, we also controlled for weeks of gestation, race, sex of the child, and mother's age.