Karim Lakhal Stephan Ehrmann Isabelle Runge Dalila Benzekri-Lefe `vre Annick Legras Pierre Franc ¸ois Dequin Emmanuelle Mercier Michel Wolff Bernard Re ´gnier Thierry Boulain Central venous pressure measurements improve the accuracy of leg raising-induced change in pulse pressure to predict fluid responsiveness Received: 9 August 2009 Accepted: 29 December 2009 Published online: 29 January 2010 Ó Copyright jointly hold by Springer and ESICM 2010 This study was presented in part at the annual meeting of the Socie ´te ´ de Re ´animation de Langue Franc ¸aise, Paris, January 2007 (Re´animation 2007; 16:S046) and at the annual meeting of the American Thoracic Society, San Francisco, May 2007 (Am J Respir Crit Care Med 2007; 175:A33). Electronic supplementary material The online version of this article (doi:10.1007/s00134-010-1755-2) contains supplementary material, which is available to authorized users. K. Lakhal  M. Wolff  B. Re ´gnier Service de Re ´animation Me ´dicale et Maladies Infectieuses, Ho ˆpital Bichat- Claude Bernard, Assistance Publique des Ho ˆpitaux de Paris, Paris, France e-mail: lakhal_karim@yahoo.fr S. Ehrmann  A. Legras  P. F. Dequin  E. Mercier Service de Re ´animation Me ´dicale Polyvalente, CHRU de Tours, 37044 Tours, France I. Runge  D. Benzekri-Lefe `vre  T. Boulain ( ) ) Service de Re ´animation Me ´dicale, Ho ˆpital La Source, Centre Hospitalier Re ´gional, avenue de l’Ho ˆpital, 45067 Orle ´ans Cedex 1, France e-mail: thierry.boulain@chr-orleans.fr Tel.: ?33-2-38514446 Fax: ?33-2-38514142 Abstract Purpose: Passive leg raising (PLR) is a maneuver per- formed to test the cardiac Frank- Starling mechanism. We assessed the influence of PLR-induced changes in preload on the performance of PLR- induced change in pulse pressure (D PLR PP) and cardiac output (D PLR CO) for fluid responsiveness prediction. Methods: Sedated, non- arrhythmic patients with persistent shock were included in this prospec- tive multicenter study. Cardiac output and pulse pressure were measured at baseline (patient supine), during PLR (lower limbs lifted to 45°) and after 500-ml volume expansion. Patients were classified as responders or not. Results: In the whole population (n = 102), the area under the receiver-operating characteristic curve (AUC) was 0.76 for D PLR PP and was higher for D PLR CO (0.89)(p \ 0.05), but likelihood ratios were close to 1. In patients with a PLR-induced increase in central venous pressure (CVP) of at least 2 mmHg (n = 49), D PLR PP and D PLR CO disclosed higher AUCs than in the rest of the population (0.91 vs. 0.66 and 0.98 vs. 0.83; p \ 0.05); positive/negative likelihood ratios were 9.3/0.14 (8% cutoff level) and 30/0.07 (7% cutoff level), respec- tively. Conclusions: A PLR- induced change in CVP C2 mmHg was required to allow clinical use- fulness of PLR-derived indices. In this situation, D PLR PP performed well for predicting fluid responsiveness in deeply sedated patients. Keywords Monitoring  Physiologic [MesH]  Hypovolemia [MesH]  Passive leg raising  Preload  Central venous pressure Introduction In case of persistent circulatory failure after initial resuscitation, intensivists often use additional volume expansion with the aim of increasing cardiac output (CO). However, nearly half of the patients in shock are not responsive to volume expansion [1–3] after the first hours of care, and an unnecessary volume load may be delete- rious [4]. After the initial phase of resuscitation, clinical examination, cardiac volumes and filling pressures [cen- tral venous pressure (CVP); pulmonary artery occlusion pressure (PAOP)] are of little interest to predict the Intensive Care Med (2010) 36:940–948 DOI 10.1007/s00134-010-1755-2 ORIGINAL