A fixed vaporizer setting does not necessarily produce the same anesthetic depth throughout cardiopulmonary bypass (CPB). A randomized study of 74 adults compared entropy-guided sevoflurane titration with a fixed 1.5% setting during elective on-pump cardiac surgery. The guided approach produced lower entropy values and higher end-oxygenator sevoflurane concentrations, but also greater vasoactive-inotropic requirements. Ventilation duration and ICU stay were not significantly different.
Why this question matters during bypass
During CPB, anesthetic delivery moves from the lungs to the oxygenator. Hemodilution, temperature changes, circuit characteristics and the patient’s physiology can alter the relationship between a vaporizer dial setting and anesthetic effect. Measuring agent concentration at the oxygenator exhaust provides information about gas delivery; processed EEG monitoring provides a different view of the patient’s cerebral response.
Kumar and colleagues examined these two measurements together. Their study asks whether titrating sevoflurane to an entropy target changes the concentration measured at the end of the oxygenator, and whether those changes translate into different hemodynamic or early recovery outcomes.
Study design and anesthetic strategy
This prospective, single-center randomized trial enrolled 74 adults undergoing elective on-pump cardiac surgery, with 37 assigned to each strategy. The intervention titrated sevoflurane to an entropy range of 40–60; the comparator used a fixed vaporizer setting of 1.5% v/v. Both groups received fentanyl infusion. The administering anesthesiologist knew the allocation, while postoperative outcome assessors were blinded.
The bypass protocol used mild hypothermia at 32–34°C and hollow-fiber membrane oxygenators. Patients requiring deep hypothermic circulatory arrest were excluded. Investigators recorded state entropy, response entropy, vaporizer settings, end-oxygenator anesthetic concentration, mean arterial pressure, pump flow, sweep gas and arterial blood gas measurements across operative phases.
State entropy principally reflects processed EEG activity, while response entropy also incorporates higher-frequency activity, including electromyographic contributions. These indices are measurements used to guide titration; reaching a target does not by itself demonstrate improved clinical outcomes.
Higher concentrations and lower entropy values
The entropy-guided strategy produced lower state and response entropy values, with reported P values of 0.01 and 0.03, respectively. Vaporizer dial settings and end-oxygenator sevoflurane concentrations were higher in the guided group, with a reported P value below 0.001. The authors describe more consistent maintenance of the entropy target.
The important distinction is between a vaporizer input, the concentration measured in oxygenator exhaust and the patient’s response. These are related observations, but they are not interchangeable. An exhaust reading is not a direct blood sevoflurane measurement, and a 1.5% dial setting should not be interpreted as an established universal CPB dosing target.
Hemodynamic support increased without a demonstrated recovery benefit
Mean arterial pressure was lower with entropy-guided administration, although the authors report acceptable ranges in both groups. The highest vasoactive-inotropic score (VIS) during the first 12 hours after ICU arrival was higher in the guided group: 28.9 ± 1.4 versus 21.7 ± 2.3, with P = 0.0001 as reported in Table 2. This is a clinically relevant tradeoff when interpreting the deeper anesthetic state achieved by the intervention.
Table 2 reports mechanical ventilation durations of 2.13 ± 0.34 hours versus 2.8 ± 0.42 hours (P = 0.15), and ICU stays of 4.24 ± 1.21 days versus 4.6 ± 1.32 days (P = 0.42), for guided and fixed-setting groups, respectively. Neither comparison reached statistical significance. Arterial blood gas measurements also showed no significant between-group differences.
No participant reported intraoperative awareness during the postoperative modified Brice interview. With only 74 patients, however, the trial cannot establish that either strategy prevents this uncommon event. Likewise, nonsignificant differences in recovery measures do not establish equivalence or prove the absence of harm.
Reporting limitations deserve attention
The paper inconsistently assigns the Group A and Group B labels across sections. This review therefore uses the strategy names and the explicitly labeled table columns. Several table summaries and test statistics also appear internally inconsistent, including overall versus group-level values. The numerical results above are reproduced as reported and should be interpreted cautiously rather than treated as independently validated estimates.
Enrollment is described as occurring from July 2023 through March 2025, while the reported trial-registration identifier dates to November 2023. That chronology merits clarification when assessing prospective registration. The study was powered for an anesthetic concentration comparison, not rare awareness events or definitive clinical recovery outcomes.
Other limits include the small sample, single-center setting, open-label titration and restriction to mild hypothermia. The authors did not use an anesthetic gas scavenging system for the measurement setup. Consequently, the results do not establish that the same exhaust concentrations would be obtained with a different sampling or scavenging arrangement, and the setup should not be read as a recommendation to omit scavenging.
Relevance for perfusion teams
The study illustrates the value of considering anesthetic delivery measurements alongside cerebral monitoring and hemodynamics. It also shows why achieving a tighter processed EEG target must be weighed against the associated cardiovascular support requirements. The findings do not establish a superior universal sevoflurane regimen or justify increasing vaporizer settings in every CPB patient.
For perfusion practice, the research question centers on coordination: how oxygenator-delivered anesthesia, temperature transitions, exhaust measurements and patient response are interpreted by the cardiac anesthesia and perfusion team. Larger studies with clearer reporting and clinical endpoints are needed to determine whether a guided strategy improves outcomes beyond its effects on monitoring values.
Source: Kumar V, Kumar A, Gourav KP, Devarakonda B. Relationship between entropy-guided sevoflurane administration and end oxygenator sevoflurane concentration in adult patients undergoing cardiopulmonary bypass surgery. Annals of Cardiac Anaesthesia. 2026;29(3):378–385. doi:10.4103/aca.aca_21_26.
Funding and disclosures: The authors report no financial support or sponsorship and no relevant financial conflicts.
