Delivery of care in high mortality hospital settings: a direct observational study examining 1848 h of neonatal nursing in Kenya.

Background: In resource-constrained countries, deploying better technologies is expected to improve neonatal care but little attention has been paid to the critical role of nurse staffing. We investigate nursing workloads and their relationship with care provision in recently upgraded Kenyan neonatal units. Methods: We conducted a cross-sectional analysis using data from direct bedside observations across 8 intermediate-level neonatal units (defined by the World Health organization (WHO) as neonatal units capable of providing specialised but not intensive neonatal care) in Kenya over a 6-week period between February and March 2022. We excluded babies who were so severely ill that they were at risk of imminent death or transfer to a critical care facility and babies with congenital anomalies and surgical conditions. We used a structured observation checklist that had undergone content and face-validity testing in Kenya to collect workload and nursing care provision data. We determined nursing hours per patient per 12-h shift to measure nursing workload (primary exposure variable) and used a composite measure, the nursing care index (NCI), to score the nursing care delivered to a baby (primary outcome variable). The relationship between nursing workload and nursing care provision was assessed using multilevel models. Findings: Across 8 hospitals spanning 1848 h of observation of 597 sick newborns, the median nursing time available for each newborn ranged from 19.2 to 72.0 min on a 12-h shift. Nurses delivered 32% of expected care, completely missed 32%, and informally delegated 36% of tasks. Unsupervised nursing students and mothers played prominent roles in the care of clinically unstable babies. An exploratory model combining qualified nurse and nursing student hours, that considerably expanded the range of the nursing workload metric, demonstrated a 3.1% increase in nurse-delivered care per additional 60 min of nursing time per baby per shift (β: 0.031, 95% CI: 0.019-0.043). Interpretation: The nursing time available to care for sick newborns in Kenya is inadequate resulting in suboptimal care quality. Improving care quality, reducing of newborn mortality and making effective use of new technologies may not be possible without increasing nurse staffing in resource-constrained countries. Funding: National Institute for Health and Care Research.