Can passive leg raise predict the response to fluid resuscitation in ED?

Abstract

Objective

Passive leg raise (PLR) can be used as a reversible preload challenge to stratify patients according to preload response. We aim to evaluate the accuracy of PLR, monitored by a non-invasive cardiac output monitor in predicting to response to fluid resuscitation in emergency department (ED).

Methods

We recruited adult patients planned to receive a resuscitation fluid bolus. Patients were monitored using a thoracic electrical bioimpedance (TEB) cardiac output monitor (Niccomo, Medis, Germany). A 3-min PLR was carried out before and after fluid infusion. Stroke volume changes (ΔSV) were calculated and a positive response was defined as ≥ 15% increase.

Results

We recruited 39 patients, of which 37 were included into the analysis. The median age was 63 (50–77) years and 19 patients were females. 17 patients (46%) were fluid responders compared to 11 (30%) with positive response to PLR1. ΔSV with PLR1 and fluid bolus showed moderate correlation (r = 0.47, 95% confidence interval, CI 0.17–0.69) and 62% concordance rate. For the prediction of the response to a fluid bolus the PLR test had a sensitivity of 41% (95% CI 22–64) and specificity of 80% (95% CI 58–92) with an area under the curve of 0.59 (95% CI 0.41–0.78). None of the standard parameters showed a better predictive ability compared to PLR.

Conclusion

Using TEB, ΔSV with PLR showed a moderate correlation with fluid bolus, with a limited accuracy to predict fluid responsiveness. The PLR test was a better predictor of fluid responsiveness than the parameters commonly used in emergency care (such as heart rate and blood pressure). These data suggest the potential for a clinical trial in sepsis comparing TEB monitored, PLR directed fluid management with standard care.