Abstract [1306]
What is the best site for measuring the pleth variability index (PVI) during a surgery procedure?
Pavlakovitch I., Desebbe O., Cannesson M., Bastien O., Lehot J.-J.
Hospices Civils de Lyon Hopital L Pradel, Department of Anaesthesiology and Intensive Care, Bron-Lyon, France
Background: The PVI (Masimo corp) is a non invasive and continuous monitoring of fluid responsiveness in mechanically ventilated patients. The PVI is derived from the respiratory variations in the perfusion index (PI) measured from the photoplethysmograpich waveform. The PVI has been evaluated only in steady state conditions and from the finger site. However, the cephalic region seems to allow better detection of the respiratory signal within the plethysmographic waveform than the finger and seems to be less sensitive to sympathetic tone.
The goal of our study was to evaluate three different sites for PVI measurement (ear, forehead, and finger) and to test their ability to detect a pulse pressure variation (PPV) > 10% during ongoing surgery.
Materials and Methods: Sixteen patients were studied during abdominal aortic aneurysm. Hemodynamic data, PPV (automatically and continuously displayed on Philips Monitors) and PVI at three sites (forhead, ear and finger) were recorded at 5 different steps (after induction, after incision, after aortic clamping, before and after aortic declamping). Pulse oximeters (LNOPAdt, Masimo Corp) attached to the forehead, ear and finger were connected to a monitor (Radical 7, Masimo Corp). PVI calculates the respiratory variations in the plethysmographic waveform amplitude as: PVI= (PImax-PImin)/PImax where PImax and PImin are the maximum and the minimum Perfusion Index values over a given period of time; PVI is averaged over 2 minutes.
Results and Discussion: The PVI was correlated with the PPV (r=0.43 for PVIfinger with p< 0.001; r=0.60 for PVIforehead with p < 0.001; r=0.72 for PVIear with p< 0.001). A PVIforehead > 14% and a PVIear > 15% predicted a PPV > 10% with a sensitivity of 73% and a specificity of 72%
 areapcutoffsensibility (%)specificity (%)
[Area under the curve for PVI to predict a PPV>10%]

Conclusion: Ear and forehaed seems to provide the best accuracy for PVI determination compared to the finger.This observation may be related to a smaller vasomotor tone at these sites than at the finger.