Abstract [1284]
 
 
Plethysmography variability index: Comparison with pulse pressure variation in vascular surgery
 
Guinet P., Daviet L., Wodey E., Laviolle B., Ecoffey C.
Pontchaillou Hospital, University of Rennes 1, Department of Anaesthesiology and Intensive Care, Rennes, France
 
Background and Goal of Study: Fluid optimization is a daily issue in the operating room. Pulse Pressure Variation (PPV) is a reliable predictive index of cardiac output response after volume expansion (1), but requires an arterial catheter. Respiratory variations of plethysmographic curve show similarities with PPV curve although non invasive (2). We sought to test the reliability of Plethysmography Variability Index (PVI) in comparison with PPV.
Materials and Methods: Observational study in vascular surgery patients (carotid and aortic surgery). PPV (Philips Intellivue MP 60) and PVI (Masimo Corp) were recorded 7 times in each patient: T0, after induction of general anesthesia; T1: table position at +20; T2: Trendelenbourg, -20; T3: table at 0, after fluid expansion (500ml Ringer's lactate); T4: beginning of carotid or aortic clamping period; T5: end of clamping; T6: end of surgery. In a subgroup of patients, arterial pressure and plethysmographic raw curves were recorded in order to allow manual assessment of PPV (PPVm) and PVI (PVIm). Agreement between PPV, PVI, PPVm and PVIm was evaluated using a Bland-Altman plot and Intraclass Correlation Coefficient (ICC, strong correlation if >0,6) (3).
Results and Discussion: 41 patients (7 women, 34 men, mean age 69) were included. Correlation betwwen PPV and PVI over the 287 measurements was not strong (0,55). Correlation for each time point was 0,27 at T0, 0,60 at T1, 0, 69 at T2, 0,30 at T3, 0,58 at T4, 0,50 at T5 and 0, 49 at T6. Bland-Altman plot showed a broad dispersion for high values of PPV and PVI. In the subgroup of 15 patients (8 carotid and 7 aortic surgeries) with manual computing of PPVm and PVIm, ICC was 0,87 between PPV and PPVm, 0,73 between PPVm and PVIm (strong correlations), and 0,51 between PVI and PVIm.
Conclusion(s): Correlation between PPV and PVI is not strong in our vascular surgery patients. PVI may not be as reliable as PPV for guiding volume expansion. Nevertheless the strong correlation between PPV and PVIm underlines the potential interest of the plethysmographic curve in providing a predictive index for fluid opimization.
References:
(1) Michard F, Boussat S, Chemla D,et al. Am J Respir Crit Care Med. 2000;162(1):134-8.
(2) Cannesson M, Besnard C, Durand PG et al. Crit Care. 2005 5;9(5):R562-8.
(3) Falissard B. Comprendre et utiliser les statistiques dans les sciences de la vie. Ed Masson. 1996