Avtor/Urednik		Stankovski, V; Smoth, AE; Arnež, ZM; Veselko, M; Lopez, JA; Young, IR
Naslov		A decision tree model for length of stay in hospital
Tip		članek
Vir		IFMBE Proc
Vol. in št.		Letnik 2
Leto izdaje		2002
Obseg		str. 258-9
Jezik		eng
Abstrakt		The aim of this study is to model the length of stay of the patient in a trauma unit i.e. bed occupancy, for all those with lower limb fractures. For this purpose, we use a recently developed modified methodology for medical research. The modelling is based on an audit database of 747 cases and the Quinlan's Se5S machine learning environment. The optimal decision tree model is presented and interpreted. The results show that the key parameters influencing the length of stay in hospital are: the level of training of the doctor involved, the age of the patient at the time of the accident, as well as the Accident Severity Indea.
Deskriptorji		LENGTH OF STAY DECISION TREES TRAUMA SEVERITY INDICES LEG AGE FACTORS SEX FACTORS SMOKING ALCOHOL DRINKING