P143 INVESTIGATION OF THE VALIDITY AND CLINICAL UTILITY OF BIOIMPEDANCE MEASUREMENTS OF FLUID STATUS IN INTENSIVE CARE David Keane1,2, Stuart Elliot3, Elizabeth Lindley1,2, Andrew Lewington1, Stuart Murdoch3 Departments of 1Renal Medicine, 2Medical Physics and 3Anaesthetics, Leeds Teaching Hospitals NHS Trust INTRODUCTION: Patients admitted tointensive care units (ICU) frequentlyreceive large volumesof intravenous fluid, particularly in the acute stages of their treatment.This fluid is intended to maintain blood volume but will often leak into the tissues. Studies of critically ill patients appear to show a link between body fluid gain and morbidity and mortality (Alsous et al, Chest, 2000). Treatment strategies designed to prevent or reduce excessive fluid overload could improve survival. Fluid management is normally reliantupon fluid balance charts (FBC) which attempt to documenttotal fluid intake and output.However, this method is prone to inaccuraciesand canbe unreliable.Bioimpedance (BI) measurements provide a simple, non-invasive method of assessing body fluid volumes and fluid status and have been shown to improve outcomeswhen used to guide fluid management in patients with renal failure (Onofriescu et al, AJKD 2014). There is a lack of published work around the use of the technology in ICU and this study aimedto determineif BI measurements have the potential to provide an accurate and reliable indication of fluid status in ICU. METHODS: Incident patients on the ICU of a large Teaching Hospital, who were identified within 24 hours of admission and expected to be present on the unit for over 48 hours, were considered for inclusion. Consentwas provided by either relatives or the patient dependent upon capacity.Daily BI measurements were made using the Body Composition Monitor (BCM, Fresenius AG Bad Homburg)fromthe point of identification, based on an estimated weight and a measured height.Clinical observations were also recorded daily. BCM measurements were scored by an experienced user based on the acceptability of the raw data. RESULTS:Consent was obtained from 30 of the 45 patients identified. Four patients died on the ICU during the study. Acceptable measurements without significant artefact were obtained in 91% (246/270) of cases. The difference between BCM and FBC on the first day of measurement ranged from negative 8 litres up to positive 6 litres.The daily change in fluid balance between BCM and FBC agreed in two thirds of cases (139/206). Disagreement was no more likely to occur in septic patients than non-septic patients. An example of the relationship between BCM and FBC measures of fluid status in one patient can be seen in fig. 1, showing good agreement but a significant offset from day 1. Figure 1: FBC and BCM measured hydration for one patient Almost half of patients (10/23) left ICU with more than 3 litres of excess fluid. DISCUSSION: Results suggest that BI measurements can be used reliably in this setting with appropriate training and an understanding of measurement quality. Measurements may be most beneficial on ICU admission where FBC assumes zero balance and afterinterventions, such as surgical procedures, where input/output are not monitored.Objective measurements of fluid overload may allowfor greater confidencewhen removing fluid once patients have stabilised than relying on FBC alone. The results showed that patientsare frequently discharged from ICU with significant excess fluid, often to wards thatlackspecialist support for fluid management. CONCLUSION: BI measurements can provide useful information relating to fluid status that has the potential to assist decision making in the ICU. Further research evaluating how BCM could be incorporated into fluid management of the critically illis warranted.