Objective: This pilot study investigated whether serum B-type Natriuretic Peptide (BNP), bioelectrical impedance analysis (BIA), and left ventricular end-diastolic diameter (LVEDD) can be used to predict fluid overload and clinical outcomes in pediatric sepsis. Methods: Pediatric sepsis patients were enrolled. BNP, BIA, and LVEDD were obtained on admission and on Day 3. Diagnostic performances of BNP, BIA, LVEDD and correlation with fluid status were obtained. Results: Twenty-two patients were enrolled. Day 3 BNP was higher in non-survivors (9241 vs. 682.2 pg/mL, p=0.04) and day 3 LVEDD Z-score was lower in non-survivors (-3.51 vs. -0.01, p=0.023). There was no difference in the fluid balance between survivors and non-survivors. Admission BNP >670.34pg/mL predicted vasopressor use with a sensitivity of 85.71% and specificity of 86.67% while ΔBNP>5388.13pg/mL predicted mortality with 100% sensitivity. Day 3 LVEDD <22mm predicted mortality with a sensitivity of 94.74%. Cumulative fluid balance was strongly correlated with BIA and LVEDD (r=0.65, p=0.001; r=0.74, p<0.001 respectively). The median length of stay in hospital days for non-survivors was not significantly different from survivors (4 [1-12] vs. 8 [6-12] days,p=0.21). Conclusion Rise in BNP levels appear to be independent of fluid status and is a good predictor of mortality, vasopressor, and mechanical ventilator use but not of length of hospital stay. LVEDD and BIA are good estimates of cumulative fluid balance but not as predictors of mortality, vasopressor, mechanical ventilator use, and length of hospital stay. Significance of the outcomes of the study was limited due to the small sample size.
Natriuretic Peptide, Brain ; Echocardiography