Estimation of Creatinine Clearance in Morbidly Obese Patients

Jasmina A. Demirovic; Amy Barton Pai; Manjunath P. Pai


Am J Health Syst Pharm. 2009;66(7):642-648. 

In This Article


We conducted a single-center, cross-sectional study at the University of New Mexico Hospital (UNMH). This study was approved by the human research review committee and the scientific review panel of the UNMH's General Clinical Research Center. Written informed consent was obtained from participants before the study began. Morbidly obese patients (BMI of ≥40.0 kg/m2) age 18-75 years admitted to UNMH between August 1, 2005, and May 31, 2007, were screened for study inclusion. Patients with any of the following criteria were excluded from the study: > 20% change in SCr (from hospital admission to day of study initiation), admission to an intensive care unit, inability to walk unassisted, diagnosis of congestive heart failure, treatment with any known nephrotoxic agents (e.g., amphotericin B, cyclosporine, tenofovir), or treatment with cimetidine, trimethoprim-sulfamethoxazole, corticosteroids, or probenecid. In addition, pregnant or lactating women and patients with liver disease (Child-Pugh category B or C) were excluded.

Measurement of CLcr

To measure CLcr, we performed a timed 24-hour urine collection. Patients were instructed to void before the urine collection was started. Urine specimens were stored on ice until the analysis was performed. Analysis was performed within 4 hours after the end of collection to avoid additional conversion of creatine to creatinine. SCr values were obtained from the UNMH clinical database or by performing venous blood sampling at some point during the day of the urine collection if not performed as a routine standard of care. Urinary and serum creatinine concentrations were measured by an automated method using a Vitros 5,1 FS analyzer (Ortho Clinical Diagnostic, Rochester, NY) using an enzymatic method (Tricore Reference Laboratories, Albuquerque, NM).[18] The interday and intraday coefficients of variation for this assay were < 2%. Measured CLcr was calculated using the standard formula as outlined in the appendix. Adequate 24-hour urine collection was verified by comparing the measured 24-hour urinary creatinine output (urinary creatinine concentration x urine volume) with the expected urinary creatinine output of approximately 15 mg/kg of IBW per day for women and 20 mg/kg of IBW per day for men. In addition, CLcr was estimated in milliliters per minute for each patient with the Cockcroft-Gault, Salazar-Corcoran, and MDRD4 equations. In addition, we substituted six body-size descriptors (TBW, IBW, ABW0.3, ABW0.4, FFW, and LBW) into the Cockcroft-Gault equation to reflect various clinical approaches.

Determination of FFW Using BIA

We determined FFW and total body water using BIA according to the National Institutes of Health Technology Assessment Conference Statement.[19] All patients were required to fast for at least four hours before BIA measurement. A portable impedance Quantum II analyzer (RJL Systems, Detroit, MI) using the standard tetrapolar method was utilized. Briefly, two distal current-introducing electrodes were placed on the dorsal surfaces of the hand and foot. In addition, two voltage-sensing electrodes were applied at the wrist and ankle. Measurements were made with the patients in supine position for at least 15 minutes and application of a 50-kHz alternating current. Patients' arms and legs were abducted at a 30-45° angle from the trunk. The resistance measurements from the impedance device were used along with height, weight, and age in a regression equation derived by Gray and colleagues[20] to calculate FFW.

Statistical Analyses

One-way analysis of variance was used to compare measured and calculated CLcr and GFR values; post hoc comparisons were performed with Tukey's test. Ordinary least-squares linear regression was used to compare slopes, intercepts, and correlation coefficients for equation-estimated CLcr or GFR with 24-hour measured CLcr. The influence of leverage points was assessed with natural logarithmic transformation and by Cook's D. Bias was defined as the mean difference between the estimated CLcr for each equation and the measured CLcr, while precision was defined as the root mean-squared error. The 95% confidence intervals were constructed around the slopes, y-intercepts, and bias. Bland-Altman plots were generated to describe bias relative to the mean of the measured and estimated CLcr values. Equation accuracy was defined as the percentage of patients with estimated CLcr values that were ≤ 30% and ≤ 50% of the measured CLcr.[21] All analyses were performed using STATA IC, version 10 (Stata Corp., College Station, TX).


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.