Passive leg raising-induced changes in pulse pressure variation to assess fluid responsiveness in mechanically ventilated patients: a multicentre prospective observational study.

Abstract

Passive leg raising-induced changes in cardiac index can be used to predict fluid responsiveness. We investigated whether passive leg raising-induced changes in pulse pressure variation (ΔPPV_PLR) can also predict fluid responsiveness in mechanically ventilated patients. In this multicentre prospective observational study, we included 270 critically ill patients on mechanical ventilation in whom volume expansion was indicated because of acute circulatory failure. We did not include patients with cardiac arrythmias. Cardiac index and PPV were measured before/during a passive leg raising test and before/after volume expansion. A volume expansion-induced increase in cardiac index of >15% defined fluid responsiveness. To investigate whether ΔPPV_PLR can predict fluid responsiveness, we determined areas under the receiver operating characteristic curves (AUROCs) and grey zones for relative and absolute ΔPPV_PLR. Of the 270 patients, 238 (88%) were on controlled mechanical ventilation with no spontaneous breathing activity and 32 (12%) were on pressure support ventilation. The median tidal volume was 7.1 (inter-quartile range [IQR], 6.6-7.6) ml kg^-1 ideal body weight. One hundred sixty-four patients (61%) were fluid responders. Relative and absolute ΔPPV_PLR predicted fluid responsiveness with an AUROC of 0.92 (95% confidence interval [95% CI], 0.88-0.95; P<0.001) each. The grey zone for relative and absolute ΔPPV_PLR included 4.8% and 22.6% of patients, respectively. These results were not affected by ventilatory mode and baseline characteristics (type of shock, centre, vasoactive treatment). Passive leg raising-induced changes in pulse pressure variation accurately predict fluid responsiveness with a small grey zone in critically ill patients on mechanical ventilation. NCT03225378.