Skip to content


  • Poster presentation
  • Open Access

Validation of the Cockroft-Gault formula for estimation of creatinine clearance in the critically ill population

  • 1,
  • 1,
  • 1 and
  • 1
Critical Care200610 (Suppl 1) :P277

  • Published:


  • Glomerular Filtration Rate
  • Cimetidine
  • Peritoneal Dialysis
  • Creatinine Clearance
  • Ranitidine


Estimation of creatinine clearance by the Cockroft-Gault (CG) formula is used in daily practice to calculate the glomerular filtration rate (GFR) in critically ill patients. This formula was derived from a cohort of noncritically ill patients and has not been validated against a standard measurement of the GFR such as 24-hour urine creatinine clearance in the critically ill patient.


To quantify the discrepancy in GFR estimated by the CG formula and 24-hour urine creatinine clearance.


A prospective cohort study was conducted in 50 adult patients in a mixed medical-surgical ICU. Inclusion criteria were ITU stay >48 hours and indwelling urinary catheter. Exclusion criteria were age <18 years, pregnancy, hemodialysis or peritoneal dialysis, urine output <400 ml/day and patients receiving cimetidine, ranitidine, cefoxitin, trimethoprim or diuretics. We estimated the GFR by the CG formula and measured the GFR by the 24-hour urine creatinine clearance. A Bland and Altman plot was used to find the percentage difference between the paired observations. The association between the two methods was measured by the product moment correlation coefficient.


The mean GFR estimated by the CG formula was 95.19 (CI) and that by 24-hour urine creatinine clearance was 90.69 (CI). The bias as measured by the mean percentage of error (MPE) and the precision as measured by the mean absolute percentage of error (MAPE) between these two methods were -8.07 and 103.18, respectively. The correlation coefficient of the CG formula as a measure of the GFR was 0.65 (P < 0.0001).

Figure 1


We conclude that CG formula has a strong correlation with the measured GFR but is not a reliable measure and overestimates the GFR in the critically ill population.

Authors’ Affiliations

AMRI Hospitals, Kolkata, India


© BioMed Central Ltd 2006