Objective: We investigated the value of information on clinical features and intensity of treatment activity in the Intensive Care Unit (ICU) in predicting the need for further interventions after a patient is discharged from the Intensive Care Unit. Our aim was to assess if this could aid in making decisions about transfer to an Immediate Care Unit (sub-ICU).
Methods: We studied 39 patients with acute brain damage (traumatic or vascular causes). They ranged in age from 15 to 75 years and none had an associated spinal cord injury. The SAPS II, Glasgow Coma Scale (GCS), length of stay in ICU, duration of hospital stay and daily NEMS scores were recorded for each patient. We recorded the activities performed after transfer to an sICU, including complications that required active "life-saving" treatment. The role of each factor was assessed by using the odds ratio (OR), and with linear logistic regression.
Findings: 8 of the 39 patients developed a complication in the Sub-ICU. A linear logistic regression analysis demonstrated that the principal features having significant predictive value were:a) age, with an increase in risk of over 10 times for patients that were older than 50 (p=0.011);b) SAPS II scores > or = 50 points, with 24 times an increase in risk (p=0.002); and c) a GCS score < or =5 points, with an increase in risk of almost 7 times (p=0.024).
Interpretation: Complications in Sub-ICU are less likely in patients younger than 50 and who have SAPS II and GCS scores within predetermined limits. These indices can help in making decisions about discharge of a patient from ICU to Sub-ICU.
Copyright 2004 Springer-Verlag