Objective
To assess the accuracy of 4 mathematical equations used to estimate creatinine clearance versus the 24-h creatinine clearance in ICU patients.Design
Prospective study of renal function prediction.Setting
The general adult ICUs of 3 metropolitan hospitals.Patients
199 critically ill patients with indwelling foley catheters.Intervention and measurements
Routine 24 h creatinine clearances were evaluated only in patients whose urine volume recorded by the nurses was within 10% of the laboratory's measured volume. Four mathematical equations utilizing age, sex, body weight, height, and plasma creatinine were used as a comparison. There was no difference in estimated creatinine clearance by 3 published methods when the 24 h creatinine clearance exceeded 100 ml/min. When the 24 h creatinine clearance was less than 100 ml/min, however, one prediction equation adjusted for lean body weight (LBW), was the most accurate. This equation accurately predicted creatinine clearance in the range of 30–100 ml/min and slightly overestimated creatinine clearance at 0–30 ml/min (p<0.01, anova="" all="">0.01,>p<0.05 fisher="" and="" scheffé="" post-hoc="" tests)="" with="" a="" mean="" difference="" ±95%="" confidence="" interval="" of="" –5±3.1="">0.05>Conclusion
An initial rapid estimate of creatinine clearance in critically ill ICU patients with reduced renal function can be determined by an equation adjusted for LBW.Key words Creatinine clearance - Cockcroft and Gault - Prediction equations - Intensive care unit