American College of Clinical Pharmacy
      Search      Cart
         

Mon-98 - Accuracy of predictive equations in estimating measured creatinine clearance for patients with adequate kidney function following traumatic injury

Scientific Poster Session III - Original Research

Original Research
  Monday, November 13, 2023
  01:00 PM–02:30 PM

Abstract

Introduction: Accurate kidney function estimation is essential for dosing drugs with significant kidney elimination. Because critically ill patients with traumatic injuries can experience augmented renal clearance (ARC), it is important to ascertain precision of predictive methods to estimate glomerular filtration rate (GFR).

Research Question or Hypothesis: Does the Cockcroft-Gault (CG) or chronic kidney disease epidemiology collaboration (CKD-EPI) equation accurately predict measured creatinine clearance (mCrCl) in critically injured trauma patients?

Study Design: retrospective analysis

Methods: Accuracy of CG and the race-less CKD-EPI equations (mL/min, not indexed to body surface area) was determined for trauma patients with a 24-hour urine collection via an indwelling catheter. Precision was evaluated via root mean squared error (RMSE) and percent of estimates within 30% of mCrCl (P30). Equations were considered unbiased if the 95% confidence interval (CI) for the mean error included 0. ARC was defined as mCrCl >129 ml/min/1.73m2. Patients with serum creatinine >1.5 mg/dL, kidney replacement therapy, chronic kidney disease, suspected rhabdomyolysis, or incomplete urine collections were excluded. Data are expressed as median [interquartile range].

Results: Two hundred patients were studied. Median age was 44 [28, 60], 76% were male, 45% were obese, 22% had traumatic brain injury, median injury severity score was 26 [20, 35], and 61% experienced ARC. Measured CrCl was 183 [141, 234] ml/min. CG and CKD-EPI equations significantly underpredicted mCrCl (135 [100, 177] and 135 [113, 155] ml/min, respectively, P < 0.001) but were unbiased with a 95% CI of -155 to 44 and -164 to 52 ml/min, respectively. Neither CG nor CKD-EPI demonstrated superior precision with a median error of -46 [-77, -19] versus -45 [-92, -14] (P=0.571), RMSE of 47 [22, 77] versus 46 [17, 92] ml/min (P=0.799), and P30 of 58% versus 57% (P=0.919).

Conclusion: A significant proportion of medications with predominant kidney elimination may be under-dosed in trauma patients when using conventional predictive formulas.

Presenting Author

Julie Farrar Pharm.D., BCCCP

Authors

Delaney Adams Pharm.D.
Regional One Health

Joanna Hudson PharmD
The University of Tennessee

Roland Dickerson Pharm.D., BCNSP, FCCP, FASHP, FCCM, FASPEN
Joseph Swanson Pharm.D., FCCM, FCCP
University of Tennessee Health Science Center