In his counterpoint editorial in a recent issue of CHEST (September 2010), Schmidt2 maintained that the development of alternative goal-directed therapy/resuscitation protocols that are simpler and noninvasive may facilitate implementation. At our institution, we experienced difficulties similar to those described by Mikkelsen et al3 since, even after extensive educational interventions in the year 2007, <50% of eligible patients were treated in accordance to the EGDT protocol, in most cases because a CVC was not placed in the patient in the ED. These patients received an inadequate amount of fluids within the first 6 h, probably because when central venous pressure was not measured, only clinical signs were used as a guide to fluid replacement. We are now enrolling patients to a noninvasive protocol that uses inferior vena cava collapsibility >30% and the absence of B lines at lung ultrasound as a guide for fluid challenge.4 CVC placement is reserved for patients who need noradrenaline infusion. A lactate clearance >10% rather than adequate central venous oxygenation values is used as a sign of improving tissue perfusion.5 To date, we have enrolled 27 patients with sepsis whose eligibility was defined as a serum lactate level of ≥4 mmol/L (seven patients) or systolic BP <90 mm Hg after volume resuscitation (20 patients). Average fluid infusion was 4,120 ±1,473 mL. A CVC needed to be inserted for the administration of vasoactive drugs in only 44% of cases. Mortality was 44.4% in patients with hypotension and 14.3% in normotensive patients with high lactate levels. Both values are comparable with those published by Rivers6 in his original study. In conclusion, we believe that given the increasing diffusion of clinical ultrasound in the ED, a noninvasive approach such as the one we are using could prove helpful in overcoming many difficulties inherent in the classic EGDT protocol and could favor its well-deserved diffusion.