Abstract
BACKGROUND: Chronic kidney disease (CKD) increases morbidity and mortality in most vascular procedures. However, a binary classification of estimated glomerular filtration rate (eGFR) <60 mL/min/1.73 m2, which is often used in both modeling and clinical trials, may not be optimal for predicting clinical outcomes.
OBJECTIVE: Determine the optimal eGFR cutoff for use in risk stratification and prediction models.
METHODS: Vascular Quality Initiative (VQI) data for non-emergent, first-time OAR, EVAR, TEVAR, CEA, CAS, PVI, Supra- and infra-inguinal bypass were analyzed from to 2013-2023 and divided into cohorts based on eGFR (≥60, 45-59, 30-44, <30, and preoperative dialysis). χ2 and logistic regression were used to evaluate perioperative outcomes.
RESULTS: Compared to patients with eGFR ≥60, those with eGFR 45-59 had similar odds of mortality following all procedures, except TEVAR. Driven by this group, the combined cohort showed a slight increase in the odds of mortality for eGFR 45-59 (0.6% vs. 0.7%, aOR 1.16, P=0.002). Those in the 30-44 group demonstrated increased odds of mortality both overall and in the individual procedure groups (0.6% vs. 1.2%, aOR 1.78, P<0.001). The odds of mortality continued to increase with worsening eGFR. The overall rate of new permanent dialysis was low for all eGFR cohorts, with a 0.02% difference between those with eGFR >60 and those in the 45-59 cohort (0.04% vs. 0.06%; a OR 1.65, P<0.001). The odds of permanent dialysis likewise continued to increase with decreasing eGFR.
CONCLUSIONS: Rather than a binary eGFR cutoff of ≥60 and <60 to stratify patient risk, better risk stratification may be achieved by using five groups of ≥60, 45-59, 30-44, <30, and preoperative dialysis.