BMC Medical Informatics and Decision Making

Table 4 Prediction accuracy of organ dysfunctions risk at 24th hour and 48th hour

From: Dynamic Bayesian network for predicting physiological changes, organ dysfunctions and mortality risk in critical trauma patients

Organ dysfunction	24th hour			48th hour
	Sensitivity	Specificity	Accuracy	Sensitivity	Specificity	Accuracy
MIMIC-III
Respiratory	0.920	0.784	0.881	0.951	0.763	0.900
Renal	0.735	0.985	0.953	0.821	0.989	0.967
Hepatic	0.767	0.950	0.922	0.751	0.960	0.924
Cardiovascular	0.724	0.914	0.873	0.676	0.872	0.813
Hematologic	0.742	0.961	0.939	0.631	0.985	0.943
Neurologic	0.719	0.947	0.859	0.679	0.777	0.725
ICU-CH
Respiratory	0.852	0.739	0.781	0.951	0.634	0.796
Renal	0.715	0.984	0.950	0.784	0.990	0.965
Hepatic	0.707	0.983	0.970	0.690	0.989	0.967
Cardiovascular	0.761	0.961	0.938	0.632	0.946	0.889
Hematologic	0.773	0.991	0.979	0.711	0.993	0.977
Neurologic	0.775	0.955	0.903	0.764	0.858	0.824

Sensitivity = (true positives)/(true positives + false negatives), specificity = (true negatives)/(true negatives + false positives), accuracy = (true positives + true negatives)/total

Back to article page

ISSN: 1472-6947

Contact us

General enquiries: journalsubmissions@springernature.com