Patient State Index at Laryngoscopy and Tracheal Intubation Conditions

Background. Inadequate hypnotic depth at laryngoscopy may provoke laryngeal reflexes, hemodynamic responses and airway trauma. The 2023 Good Clinical Research Practice (GCRP) guideline recommends that depth of anesthesia at intubation be measured objectively. Previous studies used the bispectral index; the Patient State Index (PSI) has not been studied, and the two indices are not interchangeable.

Design. Single-center, prospective, observational cohort study of consecutive eligible patients in one operating room, where all intubations are performed by the same attending anesthesiologist. Processed-EEG (SedLine) and quantitative neuromuscular monitoring are part of routine care; no additional monitoring is applied.

Routine practice (not mandated by the protocol). Lidocaine 1 mg/kg and fentanyl 1 µg/kg, then propofol target-controlled infusion (Eleveld model, effect-site target 3 µg/mL). When the target is reached, the infusion is stopped and rocuronium 0.6 mg/kg is given; the trachea is intubated by direct laryngoscopy when the train-of-four count reaches zero. The protocol prescribes no drug, dose or timing; the drugs actually given and any deviation are recorded.

Blinding and safety. From induction until intubation conditions have been graded, the SedLine display is turned away from the intubating anesthesiologist. This does not alter care, because the timing of intubation is routinely based on the effect-site target and the train-of-four count, both of which remain visible. An independent observer records the EEG variables, watches the display, and immediately informs the anesthesiologist if PSI exceeds 60 for more than 30 seconds after the neuromuscular blocking agent; all clinical decisions remain with the anesthesiologist. Awareness is assessed on the first postoperative day with a modified Brice questionnaire.

Measurements. T0 = baseline; T1 = neuromuscular blocking agent; T2 = laryngoscope insertion; T3 and T4 = 1 and 3 minutes after intubation. Because the SedLine displays PSI with a delay of about 25 seconds, the PSI at each time point is the median of trend values from 5 seconds before to 25 seconds after the event, which reflects the 30 seconds of EEG preceding it. Participants with a median artifact percentage above 10% in the T2 window are flagged. The intubating anesthesiologist grades the three GCRP/Copenhagen components (ease of laryngoscopy, vocal cord position, reaction to intubation; excellent = 2, good = 1, poor = 0), which are summed into a composite score of 0 to 6; the standard categorical grade is also reported. The independent observer grades the reaction component a second time to assess inter-observer agreement.

Analysis. Proportional odds ordinal logistic regression of the composite score on the PSI at laryngoscopy, adjusted for three pre-specified confounders: predicted propofol effect-site concentration at laryngoscopy, interval from neuromuscular blocking agent to laryngoscopy, and age. The effect measure is the odds ratio per 10-unit increase in PSI with its 95% confidence interval. The question is whether the PSI adds information beyond what the anesthesiologist already knows; the model is associational, not causal. Pre-specified sensitivity analyses exclude, in turn, participants with Cormack-Lehane grade 3-4, a failed first attempt, a safety alert, an additional dose of neuromuscular blocking agent, a train-of-four count above zero, or artifact above 10%. Secondary outcomes are exploratory. Model checks and fallback models are specified in the statistical analysis plan.

Sample size. Power was estimated by simulation for the proportional odds model, allowing for the expected ceiling effect in the composite score and for correlation between the PSI and the confounders. To detect an odds ratio of 0.50 per 10 PSI units with 80% power (two-sided alpha 0.05), under the conservative assumption that 75% of participants reach the maximum score, 145 analyzable participants are required; allowing for 10% loss, 162 participants will be enrolled.