Autor: Jukka Räsänen, Seppo O.-V. Ranta, Markku Hynynen
Rok vydání: 2002
Předmět:
Zdroj: Journal of Clinical Monitoring and Computing. 17:53-60
ISSN: 1387-1307
Popis: Objective.Awareness with recall is a rare but seriouscomplication of general anaesthesia with an incidence ranging from0.1%–0.7%. In the absence of a reliabledepth-of-anaesthesia monitor, attempts have been made to predictawareness from intraoperative haemodynamic monitoring data, with littlesuccess. Artificial neural networks can sometimes detect relationshipsbetween input and output variables even when conventional methods fail.Therefore, we subjected standard intraoperative monitoring data to bothartificial neural models and conventional statistical methods in anattempt to predict awareness with recall. Methods.Anaesthesiarecords from 33 patients with awareness and 510 patients withoutawareness were collected. Summary data (mean, maximum, and minimum) ofend-tidal carbon dioxide concentration, arterial blood oxygensaturation, systolic and diastolic blood pressure, and heart rate werecalculated for each patient. These data were subjected to an analysis byartificial neural networks and by Poisson regression. Results.The two best neural models both had sensitivity and specificity of23% and 98%, respectively. The models have highspecificity, and in view of the low incidence of awareness, a highnegative predictive value. The prediction probabilitiesPk (SE) for the best neural models were 0.66 (0.08)and 0.60 (0.10), respectively. In the Poisson regression, there weresignificant differences in systolic and diastolic blood pressures andheart rate between patients with and without awareness. Conclusions.A prediction indicating awareness by the network is very suggestiveof true awareness and recall. Blood pressure and heart rate aresignificantly higher on average in patients with awareness than inpatients without. In an individual patient, however, none of ourartificial neural models can detect awareness sufficiently reliably.
Databáze: OpenAIRE
Externí odkaz: