Acute kidney injury is independently associated with mortality in very low birthweight infants: a matched case-control analysis

Abstract

The independent impact of acute kidney injury (AKI) on survival in very low birthweight (VLBW; < or =1,500 g) critically ill infants has not been studied. Cases (non-survivors n = 68) were matched to, at most, two controls (survivors n = 127) by incidence density sampling with replacement, birthweight (+/- 50 g), gestational age (+/- 1 week), and availability of serum creatinine (SCr) levels before the index patient's time of death. Maternal/infant demographic characteristics, co-morbidities, complications and interventions were explored. No difference existed between patients and controls in mean gestational age and birthweight (the matching variables), race, or gender. Compared with the controls, cases had younger mothers, less placental separation, fewer occurrences of hyponatremia, more intra-ventricular hemorrhage, and received chest compressions and cardiac drugs. A 1 mg/dl increase in SCr was associated with almost two-times higher odds of death [odds ratio (OR) = 1.94, 95% confidence interval (95% CI) 1.13-3.32]. OR increased when confounding variables were adjusted (adjusted OR 3.44, 95% CI 1.23-9.61). Similarly, a 100% increase in SCr from trough level was associated with an increased OR = 1.53 (95% CI 1.14-2.04) and became stronger, after adjustment of variables (adjusted OR = 1.90, 95% CI 1.10-3.27). After confounding variables had been controlled for, AKI was independently associated with mortality in VLBW infants. Further prospective multi-center studies are needed to determine whether this association exists.