Factors predicting discharge from intensive care: a Canadian experience.

This study prospectively analyzed 481 admissions to a multidisciplinary intensive care unit in order to determine factors which may prove helpful in predicting outcome from an intensive care admission. Severity of illness was assessed by the admission acute physiology score and daily therapeutic intervention scoring system. Age, sex, diagnosis upon admission, nature of the admission (medical vs. surgical, emergency vs. elective) were also studied. Our results indicate that the admission acute physiology score during the first three days of the admission and thereafter, the daily therapeutic intervention scores from the previous day are the most reliable predictors of outcome from intensive care. Age was related to outcome only when it was associated with emergency and medical type admissions. A linear logistic regression analysis was used to construct a model predicting mortality within the intensive care unit based on acute physiology score and nature of the admission.