NEONATE: Decision Support in the Neonatal Intensive Care Unit – A Preliminary Report

Abstract

Abstract. The aim of the NEONATE project is to investigate sub-optimal decision making in the neonatal intensive care unit and to implement decision support tools which will draw the attention of nursing and clinical staff to situations where specific actions should be taken or avoided. We have collected over 400 patient-hours of data on 31 separate babies, including physiological parameters sampled every second, observations made by a research nurse of all the actions performed on the baby with an accuracy of a few seconds, occasional descriptions of the appearance, mobility, sleep patterns, etc of the baby. We describe our attempts to use this data to discover examples of sub-optimal behaviour.

The Role of Expertise Research and Human Factors in Capturing, Explaining, and Producing Superior Performance

“…These observations, as well as informal feedback interviews with the same staff, suggested modifications to the settings of monitoring equipment displays, along with a number of training interventions to improve signal interpretations. Subsequent trials by a companion research group (Hunter et al 2003) demonstrated the utility of designing neonatal monitoring software according to actual observations of staff usage.…”

A self-tuning adaptive trend extraction method for process monitoring and diagnosis

“…Both versions of the trend extraction methods, FTE and ATE, were applied on a data base of 15 recordings from 15 different babies, lasting 24 h each. The data came from the Neonate project [34]. Physiological variables were collected automatically by the Badger 2 system [35] at a rate of one sample per second.…”

Automatic generation of textual summaries from neonatal intensive care data

“…The data used by BT-45 came from the Neonate project [39]. Physiological variables were collected automatically by the Badger 2 system [13] at a rate of one sample per second.…”