Different Renal Function Equations and Dosing of Direct Oral Anticoagulants in Atrial Fibrillation

Figures

• 

  Figure 1. Mean eGFRs and the Differences Among the 3 Equations in Different sCr Strata Compared to the CG formula, the MDRD equation overestimated the eGFRs across a wide range of sCr levels, whereas the CKD-EPI formula overestimated the eGFRs only among strata with an sCr of <2.6 mg/dL. In the sCr strata of 1.20-1.29 mg/dL and 1.80-1.89 mg/dL among which CG-eGFRs were already below 50 mL/min and 30 mL/min, respectively, the eGFRs calculated by the MRDR or CKD-EPI formulas were still higher than 50 mL/min and 30 mL/min. When CG-eGFRs were lower than 15 mL/min at the sCr 3.50-3.59 mg/dL stratum, the MDRD-eGFRs and CKD-EPI–eGFRs were 15.9 and 14.1 mL/min, respectively. ∗CG-eGFRs of <50, 30, or 15 mL/min. CG = Cockcroft-Gault; CKD-EPI = Chronic Kidney Disease Epidemiology Collaboration; eGFR = estimated glomerular filtration rate; MDRD = Modified Diet in Renal Disease; sCr = serum creatinine.

• 

  Figure 2. Agreements of the MDRD, CKD-EPI, and CG Formulas for the Classifications of eGFR (<15, 15-50, and ≥50 mL/min) At the cutoff values of eGFR of <15, 15-50, and ≥50 mL/min, the agreements were 78% between the MDRD and CG formulas and 81% between the CKD-EPI and CG formulas. The disagreements of different equations were largely due to overestimations (21% for MDRD and 17% for CKD-EPI). Abbreviations as in Figure 1.

• 

  Figure 3. Mean eGFRs and the Differences Among the 3 Equations in Patients Aged <75 and ≥75 years Compared to the CG equation, the MDRD or CKD-EPI formulas obviously overestimated eGFRs for (B) patients older than 75 years but not (A) those younger than 75 years. ∗CG-eGFRs of <50, 30, or 15 mL/min. Abbreviations as in Figure 1.

• 

  Figure 4. Mean eGFRs and the Differences Among the 3 Equations in Patients With BWs of <50 and ≥50 kg Compared to the CG equation, the MDRD or CKD-EPI formulas obviously overestimated eGFRs for (A) patients having a BW lower than 50 kg but not (B) those with a BW higher than 50 kg. ∗CG-eGFRs of <50 mL/min, 30 mL/min, or 15 mL/min. BW = body weight; other abbreviations as in Figure 1.

• 

  Figure 5. Percentages of Concordance and Discordance of Dosages of Each DOAC With the Different Equations Overall, there were 7.6% and 6.5% of patients treated with DOACs whose dosages were discordant between the MDRD/CG and CKD-EPI/CG formulas, respectively. (A) For the MDRD versus CG formulas, the discordance rates were lowest for edoxaban (1%) and highest for rivaroxaban (13%). (B) For the CKD-EPI versus CG formulas, the discordance rates were 1% for edoxaban, 3% for apixaban, 5% for dabigatran, and 11% for rivaroxaban. DOAC = direct oral anticoagulant; other abbreviations as in Figure 1.

• 

  Figure 6. Risk of Clinical Outcomes for Patients Treated With DOACs Compared to Warfarin Among patients receiving DOACs whose dosages were defined as on label based on the MDRD or CKD-EPI formulas, only those whose dosages were truly on label (concordance) based on the CG formula were associated with a lower risk of major bleeding compared to warfarin (interaction P < 0.001). ∗Adjusted for differences between the warfarin and DOAC groups, including age, sex, BW, CHA2DS2-VASc score, HAS-BLED score, chronic obstructive pulmonary disease, and sCr. aHR = adjusted HR; IS/SE = ischemic stroke/systemic embolism; other abbreviations as in Figures 1, 4, and 5.

• 

  Central Illustration. Overestimations of Estimated Glomerular Filtration Rate, Discordance Rate of Direct Oral Anticoagulant Dosing, and Risk of Major Bleeding Compared to the CG equation, the MDRD and CKD-EPI formulas would overestimate eGFRs in a considerable percentages of patients with atrial fibrillation (21% for MDRD and 17% for CKD-EPI), especially for patients aged ≥75 years and with a BW of <50 kg (58.8% for MDRD and 50.9% for CKD-EPI). For patients treated with DOACs, approximately 7.6% (MDRD) and 6.5% (CKD-EPI) of them had dosages of DOACs that were discordant compared to the CG equation (5% for dabigatran [Dabi], 11%-13% for rivaroxaban [Riva], 3% for apixaban [Api], and 1% for edoxaban [Edo]). Only patients whose dosages of DOACs were truly on label based on the CG formula were associated with a lower risk of major bleeding compared to warfarin. aHR = adjusted hazard ratio; BW = body weight; CG = Cockcroft-Gault; CKD-EPI = Chronic Kidney Disease Epidemiology Collaboration; DOAC = direct oral anticoagulant; eGFR = estimated glomerular filtration rate; MDRD = Modified Diet in Renal Disease.