SIGH and end-expiratory occlusion test for assessing fluid responsiveness in critically ill patients undergoing pressure support ventilation.

Background: Assessing fluid responsiveness is problematic for critically ill patients with spontaneous breathing activity, such as during Pressure Support Ventilation (PSV), since spontaneous breathing activity physiologically affects heart-lung interplay. We compared the reliability of two hemodynamic tests in predicting fluid responsiveness in this clinical setting: SIGH, based on a ventilator-generated sigh applied at 35 cmHO for 4 s and the end-expiratory occlusion test (EEOT).

Methods: Prospective study conducted in a general intensive care unit (ICU) and enrolling patients in PSV showing different inspiratory effort [assessed by airway occlusion pressure (P0.1)] and requiring volume expansion (VE). Hemodynamic variables were recorded by means of the MOSTCARE system, patient received a VE using 4 ml/kg of crystalloids over 10 min and were considered responders if a cardiac output (CO) ≥ 10% was observed. The reliability of SIGH and EEOT in discriminating fluid responsiveness was assessed using receiver operating characteristic (ROC) curve approach and the area (AUC) under ROC curves was compared. For the EEOT, we considered the percent changes of CO between baseline the end of the test, while for the SIGH, the percent changes of pulse pressure (PP) between baseline and the lowest value recorded after SIGH application.

Results: Sixty ICU patients were enrolled, and 56 patients analysed. The AUC of PP changes after SIGH was 0.93 (0.84-0.99) [sensitivity of 93.1% (78.0-98.7%); specificity of 91.6 (73.0-98.9%)]; best threshold - 25% PP from baseline (grey zone - 15%/35%)]; and greater than the AUC of CO changes after EEOT [0.67 (0.52-0.81); sensitivity of 72.4% (54.3-85.3%) specificity of 70.3% (73.0-98.9%)]; best threshold 4% of CO increase from baseline (grey zone - 1%/10%)]. In the subgroup having a P0.1 < 1.5 cmHO, the AUC of SIGH [0.98 (0.94-0.99)] and of EEOT [0.89 (0.72-0.99] were comparable (p = 0.26).

Conclusions: In a selected ICU population undergoing PSV, SGH reliably predicted fluid responsiveness and performed better than the EEOT, which is, however, still reliable in the subgroup of ICU patients having a small extent of inspiratory efforts.