Fluid overload and mortality in children receiving continuous renal replacement therapy: the prospective pediatric continuous renal replacement therapy registry

Background: Critically ill children with hemodynamic instability and acute kidney injury often develop fluid overload. Continuous renal replacement therapy (CRRT) has emerged as a favored modality in the management of such children. This study investigated the association between fluid overload and mortality in children receiving CRRT.

Study design: Prospective observational study.

Setting & participants: 297 children from 13 centers across the United States participating in the Prospective Pediatric CRRT Registry.

Predictor: Fluid overload from intensive care unit (ICU) admission to CRRT initiation, defined as a percentage equal to (fluid in [L] - fluid out [L])/(ICU admit weight [kg]) x 100%.

Outcome & measurements: The primary outcome was survival to pediatric ICU discharge. Data were collected regarding demographics, CRRT parameters, underlying disease process, and severity of illness.

Results: 153 patients (51.5%) developed < 10% fluid overload, 51 patients (17.2%) developed 10%-20% fluid overload, and 93 patients (31.3%) developed > or = 20% fluid overload. Patients who developed > or = 20% fluid overload at CRRT initiation had significantly higher mortality (61/93; 65.6%) than those who had 10%-20% fluid overload (22/51; 43.1%) and those with < 10% fluid overload (45/153; 29.4%). The association between degree of fluid overload and mortality remained after adjusting for intergroup differences and severity of illness. The adjusted mortality OR was 1.03 (95% CI, 1.01-1.05), suggesting a 3% increase in mortality for each 1% increase in severity of fluid overload. When fluid overload was dichotomized to > or = 20% and < 20%, patients with > or = 20% fluid overload had an adjusted mortality OR of 8.5 (95% CI, 2.8-25.7).

Limitations: This was an observational study; interventions were not standardized. The relationship between fluid overload and mortality remains an association without definitive evidence of causality.

Conclusions: Critically ill children who develop greater fluid overload before initiation of CRRT experience higher mortality than those with less fluid overload. Further goal-directed research is required to accurately define optimal fluid overload thresholds for initiation of CRRT.