A prospective randomized comparison of a pulse-contour analysis monitor versus a non-invasive bioreactance monitor in a stroke-volume based goal-directed fluid resuscitation protocol in brain-dead organ donors

Background: Brain-dead (BD) organ donors are frequently hypotensive and hypovolemic requiring fluid resuscitation. We previously published our four-hour stoke volume (SV)-based fluid resuscitation protocol that expeditiously corrected the fluid deficit and significantly decreased time on vasopressors. The SV was measured by pulse-contour analysis (PCA). Objective: To determine if the measurement of stroke volume by bioreactance (BR) was equivalent to PCA in a goal-directed fluid resuscitation protocol in BD donors. Methods: We performed a prospective randomized trial of fluid resuscitation in BD donors comparing the two monitors. Results: In 169 deceased donors there were 1481 comparative measurements of SV. The correlation coefficient was 0.32. A Bland Altman analysis revealed a minimal bias (-1.8 ml) but large limits of agreement (±62.9 ml) and a percentage error of 80.5 %. A 10 % SV increase with a 500 ml fluid bolus had conflicting results between the monitors in 31.3 % of 1309 measurements. As a reference standard, the Fick method of measuring SV was used 49 times with simultaneous measurements of PCA-SV and BR-SV. The mean Fick SV (96.0 ± 30.6 ml) was significantly greater than BR (79.4 ± 18.1 ml; p<.01) and the PCA (77.2 ± 22.5 ml; p<.01). The mean SV differences for Fick-BR (16.6 ± 27.2, 95 % CI 8.78–24.42) and Fick-PCA (18.7 ± 29.0, 95 % CI 10.42–27.08) were not significantly different (p=.57). Conclusion: Although this study demonstrated inaccuracies, imprecision, and disagreement between BR and PCA SV measurements, there was equipoise in clinical outcomes when used with our fluid resuscitation protocol, as previously published.