Background and Goal of Study: The Surgical Pleth Index (SPI) is proposed to assess anti-nociception during general anaesthesia(1). We compared its accuracy with that of haemodynamic parameters and identified factors that may confound its interpretation.
Materials and Methods: After IRB approval, 30 ASA I-II consenting patients scheduled for intracranial surgery received general anaesthesia based on propofol (ppf) - remifentanil (remi) TCI's. The propofol target was 3 μg/ml for all of them. They were randomly and blindly assigned to three groups, according to the target concentration of remi (2, 4, or 6 ng/ml). Both targets were achieved before head holder insertion (HHI). Volaemic status was assessed using the Delta Down(2) immediately before HHI. SPI, heart rate (HR), and mean arterial pressure (MAP) were continuously recorded. Stepwise ordinal regressions were performed to assess the effect of volaemia, and medical history of treated high blood pressure on the ability of SPI, MAP, and HR to appropriately predict the remi target.
Results and Discussion: HHI was associated to an increase in SPI, MAP and HR, more marked at low remi concentrations. The accurate prediction rate of the remi target, based on the highest (max) SPI, MAP or HR observed during the first five min after HHI was 0.5, which is better than chance in a 3 options design. It rose up to 0.6 when each model took additionally into account volemia and chronic therapy for high blood pressure. When max SPI, MAP and HR models combining those factors were concordant, the accurate prediction rate was 0.8.
Conclusion(s): The capacity of SPI to predict patient analgesic regimen during anaesthesia is equivalent to that of MAP and HR, and is improved by taking into account volaemia and chronic anti-hypertensive treatment. Combining SPI with MAP and HR offers the best accurate prediction.
1 Huiku, BJA 2007, 98 (4): 447-55 (2) Deflandre, BJA 2008, 100 (2): 245-50.