767 HOUSE STAFF EDUCATION THROUGH COMPUTER ASSISTED MANAGEMENT(CAM) OF RESPIRATORY FAILURE:

Despite their generalized availability, acceptance of computers into Neonatal units(NICU) has been slow and their impact on care minimal. To facilitate staff education and standardize ventilator management we adapted an algorithm on mechanical ventilation into a user-friendly basic computer program. To determine its effectiveness we compared incidence of resolution of arterial blood gas(ABG) derangements in 2 groups of neonates with respiratory distress syndrome: 10 infants managed by house staff assisted by consultation with computer(CAM) and 10 infants managed by spontaneous staff actions(Spont) before introduction of computer into NICU. ABG derangements(PaO2<50 or >80, PaCO2<35 or >50, or combinations thereof) were assessed and improvement defined by correction of the ABG abnormality following a ventilator change. Staff followed computer recommendations only 48/75 (64%); however, following computer recommendation, ABG improved 46/48(96%) compared with alternate decisions (19/27, 70%,p<0.005). Overall, ABG improvement occurred more frequently with CAM(65/75, 87%) than with Spont(37/57, 65%, p<0.005). Additionally unnecessary increases in ventilatory settings were observed on 7 occasions with Spont and never with CAM. We conclude that normalization of ABG and subsequent weaning from the ventilator is more effectively accomplished with CAM. This validation that CAM can be of benefit to both the staff and their patients will hasten acceptance of computers into the NICU. A properly applied computer program may assist a large rotating house staff in understanding mechanical ventilation resulting in more effective care.