Evaluating the effects of increasing nursing numbers on quality of newborn care in understaffed neonatal units in Kenya: a prospective intervention study

BACKGROUND: Newborn units in resource-constrained low-middle-income countries (LMICs) often have high neonatal mortality rates. Programmes to improve care quality often accept understaffing that directly affects care in these settings as a norm, and the effects of improving staff numbers are not studied. To address a major evidence gap, we examined the effects on quality of care of improving nurse staffing in four intermediate-level Kenyan newborn units. METHODS: We introduced three additional nurses to each of four newborn units. We measured nursing care provision using direct bedside observations with a validated structured checklist before and 6 months after intervention. Our primary outcome, changes in nurse-delivered care, was examined using descriptive analysis and multilevel modelling to adjust for confounding. We also examined the pattern of nursing care delivery and intervention fidelity. RESULTS: We observed a total of 1872 hours of care, over 156 nursing shifts for 290 and 300 babies before and after our intervention, respectively, across our four neonatal units. Our intervention increased the nursing hours per patient per shift observed from 34 to 43 min associated with a 4.7% increase in our primary outcome, nurse-delivered care and an 8.4% increase in delivery of 10 tasks nurses prioritise (adjusted B-coefficient 0.047 (95% CI 0.028 to 0.066) and B-coefficient 0.084 (95% CI 0.053 to 0.115), respectively). Intervention strength was reduced by changes in existing nurses' deployment and an increase in workload. CONCLUSIONS: In very high workload settings in LMICs where nurses can only deliver a fraction of nursing care, staffing increases improve care delivery more obviously for high-priority tasks. These findings provide much needed evidence that increasing neonatal nurse staffing in under-resourced newborn units improves care quality.