Prospective response surface modeling of analgesia nociception index using propofol, fentanyl and sevoflurane synergism for laryngeal mask airway placement

BACKGROUND: Anesthetic agents exhibit synergistic effects that are difficult to quantify, especially when multiple drugs are used. No response surface model (RSM) exists for analgesia based on the Anti-Nociception Index (ANI). Our objective is to develop a three-drug RSM for ANI to explore analgesic interactions and simulate drug administration with propofol, fentanyl, and sevoflurane. METHODS: ANI was prospectively monitored in 52 females undergoing breast surgery during laryngeal mask airway (LMA) insertion. A change in ANI (delta-ANI) of less than 20% after insertion was defined as a non-reactive response, with sensitivity analyses using <10% and <30% thresholds. Model performance was assessed using accuracy (<50% difference between predicted probability and observation), area under the receiver operating characteristic curve (AUC), Spearman's ρ, and Brier score, with 5-fold cross-validation and reduced-model testing. A secondary descriptive analysis was conducted using model-based simulation. RESULTS: = -1.34 suggested additional synergism when all three agents were combined. Among the pairwise interactions, fentanyl/sevoflurane exhibited the strongest synergism. The model accuracy was 81.6%, with an AUC of 0.8 and ρ of 0.5. Simulation showed that both the propofol/fentanyl and sevoflurane/fentanyl combinations could achieve non-reactivity. CONCLUSIONS: The three-drug RSM of propofol, fentanyl, and sevoflurane on the ANI demonstrated synergism during LMA insertion, with fentanyl being the primary contributor. The combination of propofol and fentanyl was optimal for induction, with sevoflurane providing effective maintenance thereafter. TRIAL REGISTRATION: NCT03813875, Jun 14, 2019.