Accuracy of cumulative volumes of fluid challenge to assess fluid responsiveness in critically ill patients with acute circulatory failure: a pharmacodynamic approach

Background: The relationship between the dose (volume of fluid) and the effect (increase of stroke volume [SV]) has been poorly described. We hypothesised that the analysis of the dynamic response of SV during fluid challenge (FC) helps to determine the optimal volume of FC, along with its diagnostic accuracy parameters for fluid responsiveness. Methods: A prospective observational study was conducted in critically ill patients with circulatory failure. Patients monitored with oesophageal Doppler and assigned to an FC of 500 ml of crystalloid were included. The areas under the curve (AUC) and 95% confidence intervals (CI95) of the receiver operating characteristic curves for cumulative volumes from 50 to 450 ml were determined for fluid responsiveness (SV increase ≥15% from baseline) along with other parameters of diagnostic accuracy. In the pharmacodynamic analysis, dose-effect and dose-response models were constructed, with determination of median and 90% effective dose (ED50 and ED90). Results: Forty-five patients were included. The AUC increased with cumulative volumes of FC up to 250 ml (AUC250 0.93 [CI95: 0.85-1.00]), followed by a plateau above 0.95 of AUC. The optimal volume was 250 ml, associated with a specificity of 0.89 [CI95: 0.78-1.00], a sensitivity of 0.92 [CI95: 0.69-1.00], and a threshold of 9.6% increase in SV. The ED50 was 156 [CI95: 136-177] ml and the ED90 was 312 [CI95: 269-352] ml. Conclusions: A volume of FC of 250 ml with a threshold of 9.6% increase in SV showed the highest accuracy in detecting fluid responsiveness in critically ill patients with shock.