The response of the composite variability index to a standardized noxious stimulus during propofol-remifentanil anesthesia

Abstract

Background: Recently the Composite Variability Index (CVI) was developed to quantify nociception. This index is derived from the standard deviations (s) of the Bispectral Index (sBIS) and the electromyogram (sEMG). The primary aim of our study was to compare CVI before and after a noxious stimulus. As secondary end points, we investigated the influence of remifentanil on the CVI and tested the ability of the CVI to indicate patient movement after a noxious stimulus under changing remifentanil concentrations. Furthermore, we measured the increase in CVI after a noxious stimulus in comparison to other clinical variables (BIS, sBIS, sEMG, heart rate [HR], and systolic blood pressure [BP(sys)]).

Methods: Twenty-four patients without a history of cardiac disease were investigated. Anesthesia was induced with propofol administered by target-controlled infusion. A standardized noxious electrical stimulus was applied (50 Hz, 70 mA, 30 seconds) to the ulnar nerve at increasing or decreasing remifentanil effect-compartment concentrations (Ce(remi)). Changes in baseline and poststimulus CVI, BIS, sBIS, sEMG, HR, and BP(sys) were investigated. Parameters' ability to indicate movement after a noxious stimulus was evaluated with the prediction probability (P(K)).

Results: All investigated parameters (except BP(sys)) increased significantly after a noxious stimulus at 0, 1, 2, or 3 ng·mL(-1) Ce(remi). The association between poststimulus maximal parameters and movement were P(K) = 0.81 for HR, P(K) = 0.78 for sEMG, and P(K) = 0.72 for CVI (pairwise difference to CVI statistically nonsignificant). The association between ΔsEMG or ΔCVI (poststimulus value minus baseline value) and movement was significantly higher (P(K) = 0.76 and 0.75, respectively) compared with ΔHR (P(K) = 0.53) (P = 0.008 and P = 0.01, respectively). Receiver operating characteristic analysis revealed a threshold value for movement for ΔCVI of >0.39 (sensitivity of 0.71, specificity of 0.74) and for ΔsEMG of >0.31 (sensitivity of 0.68, specificity of 0.78).

Conclusion: In paralyzed patients, ΔsEMG and ΔCVI might help identify inadequately low levels of analgesia with an acceptable sensitivity and specificity. The impact of profound neuromuscular block on the CVI should be investigated in further studies.