Use of Biomarkers to Optimize Fluid Dosing,CRRT Initiation and Discontinuation in Pediatric ICU Patients With AKI

Overview

Acute Kidney Injury (AKI) is a common clinical problem defined by an abrupt (\< 48 hour) increase in serum creatinine (SCr) resulting from an injury or insult that causes a functional or structural change in the kidney. Despite significant advancements in the care of the critically ill child, mortality rates observed in critically ill children who develop AKI have not improved. The investigators have shown even "small" increases in SCr, which is the standard kidney function marker, are associated with increased child mortality, even when outcome was controlled for significant patient co-morbidity. Furthermore, the investigators have also shown that the amount of fluid accumulation observed in critically ill children with AKI is independently associated with mortality suggesting that earlier dialysis may improve survival. However, the investigators also do not want to dialyze patients who don't ultimately need dialysis, as it is an invasive procedure. The data cited above highlight the need not only to detect AKI early, but also predict it severity in order to optimize clinical decision making with respect to fluid administration and dialysis initiation. While substantial research has been expended to validate NGAL as an early marker of AKI, it has not been studied in the context of clinical decision support to guide a therapeutic intervention. The investigators hypothesize that NGAL levels can be used to determine predict which critically ill children will develop severe and prolonged AKI with substantial volume overload, thereby providing the clinician with a diagnostic tool to guide CRRT initiation.

The specific aims of this proposal are: 1. Determine if NGAL (POC plasma and confirmatory urine) concentrations can predict which critically ill children will ultimately develop significant (\>10%) positive ICU fluid accumulation Hypothesis to be tested: Elevated plasma NGAL concentrations (initial plasma threshold \> 250 ng/ml) will predict which critically ill children will develop a positive ICU net fluid accumulation of \> 10% of ICU admission weight 2. Determine if NGAL (POC plasma and confirmatory urine) concentrations can predict which critically ill children who develop \>10-20% ICU fluid overload will recover urine output and kidney function rapidly Hypothesis to be tested: Elevated plasma NGAL concentrations (initial urinary threshold \>1 ng/mg Cr ) will predict which critically ill children who develop \>10-20% FO will not have an improvement in AKI as determined by an improvement of at least one pRIFLE strata within 24-48 hours of developing pRIFLE-I or pRIFLE-F 3. Determine if NGAL (POC plasma and confirmatory urine) concentrations can predict kidney function recovery in critically ill children develop \>10-20% ICU fluid overload who receive continuous renal replacement therapy Hypothesis to be tested: Decreasing NGAL concentrations will be associated with improvement in urine output and initial resolution of AKI in \< 72 hours This pilot study will be novel in that the investigators will evaluate NGAL levels in near real-time, twice daily to guide clinical decision support in terms of fluid administration effect assessment and CRRT provision in this critically ill pediatric population. Specifically, the investigators will use the NGAL data daily to 1) drive initiation of CRRT in children with elevated NGAL and \> 10-20% fluid overload and 2) drive CRRT discontinuation in patients with decreasing NGAL concentrations. In addition, the investigators will employ an adaptive study design to readjust the threshold NGAL during the time course of the study if the data suggest adjustment will enrich the data pool.