Evaluation of creatinine clearance estimation in an elderly male population

Thirty-one medically stable, elderly males (age 75 +/- 8.3 yrs) participated in a prospective study evaluating the accuracy of 16 methods of estimating creatinine clearance. Serum creatinine values were determined on the mornings of days 1, 4, and 5 to assure stable renal function. On the morning of day 3, a 24-hour urine collection was initiated. A statistically significant correlation existed between the measured and estimated clearance values for all 16 formulas. The correlation (r less than 0.65) was lower than that in previously published studies, however. Five of the formulas (1A, 5A, 5B, 7A, 7B) demonstrated no statistical difference between mean measured and estimated values. In this population, formula 2B was the least biased and formula 9B the most accurate. For all 16 methods, the bias was minimal and the relative accuracy of the estimated methods was comparable. The results support using methods to estimate creatinine clearance only as a rough bedside prediction of renal function in elderly males.