429 Fluid Balances Show Poor Agreement with Daily Body Weight Changes in Sick Neonates

Abstract

Background: Fluid balance monitoring is a commonly used complex, time-consuming and therefore expensive method to assess patient fluid volume status. In clinical practice fluid balances commonly appear to be inaccurate. Studies on the reliability of fluid balance are lacking however, particularly in newborns. Aim: To determine the accuracy of the fluid balance in sick newborns. Methods: Fluid balance totals (ml/day) were compared to daily body weight changes (g/day) in sick newborns admitted to neonatal high care. Correlation coefficients and Bland-Altman plots were used to assess the agreement between fluid balance and daily weight changes. We chose 50 ml as the clinically relevant limit of agreement. Results: In 170 patients, 482 fluid balances and weight differences over the same period were recorded. The correlation between fluid balance and weight change was moderate r=0,65 (p < 0,01). The difference between the weight change and fluid balance was more than 50 ml (the limit of agreement) in 40% of the cases, with 95% of the differences ranging from -125 to +104 ml. Conclusion: Agreement between fluid balance totals and daily weight changes is poor. Fluid balances both under- and overestimated weight changes in unpredictable patterns. Because of the inaccuracy of this time-consuming and complex method, recording a fluid balance is not recommended in neonatal practice.