Multiple logistic regression analysis of potential confounding.

Adjustment for potential confounding of the effect of infused volume and of baseline diabetes and renal disease on induction of CI-AKI by ioxitalamate in the intent-to-treat (n = 500) and high risk (n = 250) samples. sCr (serum creatinine mg/dL); sCys C (serum cystatin C; mg/dL). Definition of CI-AKI (contrast-induced acute kidney injury): sCr≥(baseline+0.3 mg/dL) and/or both sCr and sCys C≥(baseline · (1+10%)). aRegression models adjusted or unadjusted for potential confounding by infused volume and by baseline diabetes and renal disease. bPredictor (all treatments included saline): NAC (N-acetylcysteine), NaHCO3 (sodium bicarbonate), saline alone; infused volume (total volume of oxitalamate plus treatment administered intravenously). cbolded 2-tailed P-values were considered statistically significant and refer to the null hypothesis that the odds ratio = 1. dDifference in odds ratios (%) = (100 · ((Unadjusted odds ratio) – (Adjusted odds ratio))/(Unadjusted odds ratio)); a difference of at least 10% (bolded value) was considered evidence of confounding. ePatients at high risk of developing CI-AKI due to affliction with diabetes mellitus or renal dysfunction at baseline or due to receiving >140 mL of ioxitalamate. Multiple logistic regression analysis of potential confounding.