Sequential analysis for quality control in the neonatal intensive care unit.
OBJECTIVE: This article describes a novel application of a statistical technique for continuous quality assurance in the NICU. METHODS: We used prospective analysis of rates of survival to 28 days of life, without major IVH in ELBW infants in a single tertiary NICU, before and after the introduction of an evidence-based treatment protocol. By using the CUSUM function, each infant's results were sequentially plotted, and significant changes in outcomes were noted when the plot crossed predetermined boundary lines. RESULTS: Significant changes in outcomes were evident with this method sooner than traditional analyses on the basis of year-end or other arbitrary intervals. The introduction of the ELBW protocol was temporally associated with significant improvement in intact short-term survival. CONCLUSION: Sequential analysis techniques are useful tools for ongoing quality assurance; deviations in outcomes may be detected more quickly, which should assist in the identification of improvements or decrements in performance of the NICU.