Prediction of worsening postoperative renal function in older candidates to elective cardiac surgery: Choosing the best eGFR formula may not be enough.

BACKGROUND

Though renal impairment is highly prevalent in older patients and influence post-operative outcomes in cardiac surgery; its prognostic relevance is debated and not fully assessed by surgical risk scores.

OBJECTIVE

We investigated the predictive role of estimated glomerular filtration rate formulas for in-hospital worsening renal function (WRF) after cardiac surgery.

METHODS

We prospectively enrolled in single-center cohort study, patients aged ≥ 75 years candidate to elective cardiac surgery. Four creatinine-based equations were used to calculate estimated glomerular filtration rate (eGFR) formulas: Cockroft-Gault, Modification of Diet in Renal Disease, Chronic Kidney Disease Epidemiology, and Berlin Initiative Study 1 formulas. Each patient underwent geriatric and clinical evaluation before surgery with calculation of the Society of Thoracic Surgeons scores. In-hospital WRF was defined as a composite of an increase in SCr ≥0.5 mg/dl or the occurrence of grade III KDIGO acute kidney injury. The association between each eGFR equation, alone and in models including clinical variables, and WRF was analyzed using logistic regressions and ROC analysis.

RESULTS

WRF occurred in 69 patients (19.8%), and the predictors of WRF were previous acute myocardial infarction, hypertension, 4-mt gait speed performance, and preoperative eGFR, irrespective of the equation used. With all equations, inclusion of these additional variables in the logistic regression models improved the prediction of WRF (AUCs 0.798-0.810).

CONCLUSIONS

An accurate assessment of renal function and of physical performance should be incorporated into cardiac surgery risk scores to improve prediction of in-hospital WRF and, hence, risk stratification in older adults undergoing elective cardiac surgery.