Computerized weaning of childhood respiratory failure

Acute respiratory failure (ARF) is a major reason for admission to a pediatric intensive care unit. Endotracheal intubation and mechanical ventilation is a method of life support for ARF. Mechanical ventilation is intended to be a temporary measure with an ultimate goal of extubation and a return by the patient to natural breathing. Weaning is the process of slowly removing ventilator support in order to achieve this goal. A standardized process for weaning children from mechanical ventilation does not exist. This dissertation describes how use of a computerized pediatric ventilator weaning protocol is safe and efficient. Standardization of process reduces variation in clinical practice and results in improved outcomes. Protocols provide standardized medical decision making. Computerized protocols are the most robust form of protocols and can assist clinicians at the point of care by providing patient-specific instructions. A paper-based protocol to wean children with ARF was developed by a group of pediatric critical care clinicians. Protocol knowledge content was derived from outcomes in the medical literature and local practice. The protocol consisted of eight flow charts executed in series. Sixteen data elements were required to execute the entire protocol. Within a flow chart, time-sensitive, patient-specific instructions were provided. Following development and validation, the paper-based pediatric ventilator weaning protocol was computerized. Each flow chart was broken down into specific states corresponding to computerized if-then-else" logic statements. A commercially available rules engine was integrated with a web-browser based decision-support infrastructure to continually monitor patient specific data elements and deliver explicit instructions to clinicians. Upon completion of the computerized pediatric ventilator weaning protocol a prospective randomized controlled trial was conducted to compare the protocol to a typical-physician approach to weaning children from mechanical ventilation. Children in the protocol group were weaned significantly faster than children in the control group (8 hours vs. 28.5 hours p < 0.002). No complications of mechanical ventilation occurred in either group. These studies demonstrate that a pediatric ventilator weaning protocol can be developed, computerized and implemented to safely and efficiently assist clinicians in weaning children from mechanical ventilation. Standardizing processes for the delivery of healthcare can improve outcomes.