MO648: Evaluation of Fluid Status with Lung Ultrasound and Continuous Blood Volume Monitoring in Chronic Haemodialysis Patients

Abstract

Abstract BACKGROUND AND AIMS Continuous blood volume (BV) monitoring with Haemoscan monitor is a noninvasive photometric method that measures BV% change according to the changes in haematocrit during haemodialysis (HD) session (haemoconcentration depending on HD ultrafiltration rate and interstitial refilling rate). Lung ultrasound (LUS) has recently been shown to be a useful, noninvasive technique for the assessment of extravascular lung water and pulmonary congestion. Interstitial lung water is detected as LUS B-lines and the sum of B lines in 28 measurements sites gives the lung score (LS, 0–280). The aim of the present study was to evaluate the hydration status and its changes during the HD session, both by LUS and by BV monitoring and to identify any correlations between the two methods in chronic HD patients. METHOD All patients transferred to our new dialysis unit were examined clinically and with pre- and post-dialysis LUS, as part of routine assessment. Haemoscan BV monitoring during HD sessions is a feature of the dialysis machine. Changes between pre- and post-dialysis measurements (Delta, Δ) of different clinical and imaging parameters were recorded and evaluated. RESULTS A total of 24 patients, 19 men and 5 women, 65.5 ± 13.7 years old, on dialysis for 73 (7–201) months, with all usual causes of end-stage renal disease (ESRD) and 27 HD sessions of 3–4 h were included. Median pre-dialysis LS was 4 (0–116) and post-dialysis 1 (0–28), Δ LS: 3 (−1 to 102). The final BV change detected by Haemoscan was −8.8 ± 3.6% (−1.7 to −18.4%) and the final fluid loss was 2.43 ± 0.8 L (32.3 ± 8.3 ml/kg BW or 8.8 ± 2.2 mL/kg BW/h). Pre-dialysis systolic blood pressure (SBP)/diastolic blood pressure (DBP) was 159 ± 21/86 ± 17 mmHg and post-dialysis 138 ± 18/78 ± 14 mmHg (Δ SBP 22 ± 24 mmHg and Δ DBP: 8 ± 14 mmHg). Statistically significant correlations were found between BV change and fluid loss (rho = −0.535, P = .004) (the higher the fluid loss during HD, the higher the haemoconcentration) and between Δ LS and Δ SBP (rho = 0.550, P = .003) and Δ DBP (rho = 0.544, P = .003). No correlation was found between BV% change and Δ LS. However, most of the patients had few or no pre-dialysis B lines and a minimal Δ LS, rendering impossible the finding of any correlation. But if we consider the only 7 sessions with ≥10 pre-dialysis B lines (indicating some pulmonary congestion), a strong negative correlation was found between BV% change and Δ LS (rho = −0.786, P = .036), (the higher the hemoconcentration, the higher the pulmonary decongestion). Nevertheless, even in patients without pulmonary congestion (B-lines <5), a significant haemoconcentration during HD session was observed: BV change = −9.8 ± 3.7% (−3.9 to −18.4%) with final fluid loss of 2.5 ± 0.7 L (32 ± 8 mL/kg BW or 8.7 ± 2.3 mL/kg BW/h). CONCLUSION Fluid loss during HD correlates to haemoconcentration (BV % change), but not to Δ LS, even in patients with some pulmonary congestion (pre-dialysis LS ≥10). It seems that inter-dialysis fluid accumulation can be less apparent in the lungs and even patients without clinical signs of fluid overload (dyspnoea, oedema) and no pulmonary congestion (minimal B lines) can lose substantial fluid amounts during HD sessions with significant haemoconcentration. Indeed, in our study, no difference in fluid loss and BV% change was found between patients with high and low pre-dialysis LS. Thus, a high pre-dialysis LS score may indicate fluid overload and pulmonary congestion, but the opposite (low pre-dialysis LS) cannot indicate that the patient is dehydrated. This also may explain why we found a strong correlation of Δ LS and BV% change in those with high pre-dialysis LS, but not in all the patients. Lung ultrasound is a useful tool for fluid overload detection and can be used complementary to BV monitoring: redefining dry weight in those with high pre-dialysis LS and lung congestion and avoiding very high BV% changes which indicate reduced refilling rate and risk of hypotensive consequences.