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.