Official FDA recommendation: do not use any calculated clearance greater than 125 ml/min. This is pretty logical. Physiologically a kidney can’t filter any faster than that.
Do not use the reported EGFR that lab offers for carboplatin dosing (or for any other drug dosing adjustments for that matter). While EGFR uses more information and may be more “accurate”, no drug dose adjustments have been validated based on this number. Drug dosing adjustments are uniformly based on Cockcroft-Gault calculated clearance.
Consider that at the fringes, Cockcroft-Gault is not a great equation. A measured creatinine of less than 0.8 mg/dl may not be reflective of fantastic renal function. Using values of 0.1-0.7 mg/dl may result in very optimistic estimates of clearance. Many providers choose not to use a value of less than 0.8 or 1 mg/dl for this reason. Early carboplatin studies used actual creatinine values but in modern times it is usual to temper this.
Use of carboplatin in dialysis is often hotly debated, refer to recent literature to form your own conclusion. Two very good options: estimate clearance based on current SCr and administer PRIOR TO dialysis, or estimate clearance at 0 ml/min and administer just after dialysis.
AUC x = (CG clearance + 25)x with x being the target AUC achieved and the 25 representing nonrenal clearance.