Volatile versus propofol sedation after cardiac valve surgery: a single-center prospective randomized controlled trial

Intensive care medicine is vital in managing patients after heart valve surgery. Heart valve reconstruction and replacement are complex and invasive procedures requiring extensive monitoring to assure clinical stabilization [1]. During this time of recovery, deep sedation is frequently employed [2, 3]. However, prolonged sedation risks negative sequelae that potentially impact the patient's recovery and overall health. Consequently, fast-track strategies promoting extubation within six hours after intensive care unit (ICU) admission have emerged as the normative practice [4‐6]. Notably, prolonged sedation correlates with increased incidence of delirium, contributing to protracted ICU stays, increased mortality rates and postoperative cognitive deficits [7‐9]. Prolonged sedation and ventilation predisposes patients to muscle atrophy, possibly culminating in ICU-acquired weakness (ICUAW) or an elevated risk of ventilator-associated pneumonia [10‐12]. Earlier and predictable awakening with an equally accelerated neurological assessment may also help to detect neurological complications requiring intervention at an early stage. These periprocedural risks underscore the importance of closely monitoring patients to limit the duration of sedation and ventilation [1, 13, 14].

The increase in the number of patients requiring surveillance, alongside a rise in comorbid conditions with the associated surge in nursing workload, also needs to be considered for practicability in the implementation of novel techniques. Different sedation procedures are associated with different personnel and cost expenses, some of which appear to be amenable to bedside evaluation.

The use of sedatives in economically constrained health care systems must be carefully considered and balanced against the potential risks and benefits.

For the aforementioned reasons, shortening sedation should be considered in patients, e.g., by the use of volatile sedatives. While semiclosed circuit ventilators predominate in surgical settings, open circuit ventilators are used exclusively in intensive care. This difficulty prompted the development and deployment of anesthetic conserving devices (ACDs), facilitating the conservation and re-administration of volatile anesthetics after their application into the airway [15, 16]. This method enabled the widespread adoption of volatile sedatives in the ICU. Their application in complex patient cohorts, including those with renal and hepatic insufficiency or obesity, poses minimal challenges, albeit with noted contraindications in patients with pre-existing obstructive pulmonary diseases. The favorable properties of volatile sedatives, characterized by minimal adverse effects, rapid awakening facilitating earlier extubation, and significant bronchodilatory and anti-inflammatory properties, renders them near-ideal agents for sedation, significantly reducing ventilation time [17‐21]. The discourse surrounding the potential diminution of delirium incidence after deep sedation persists, with evidence not consistently indicating delirium rate reductions [22]. Volatile anesthetics offer dosage scalability without the habituation or limitations seen with propofol, beneficial in challenging sedation cases [23, 24]. With respect to the widely used substance propofol, a risk of propofol infusion syndrome (PRIS) has been reported [25]. While the incidence of PRIS is slightly greater than 1%, the use of volatile sedatives carries a genetically predisposed risk of malignant hyperthermia (MH), with an occurrence rate ranging from 1 in 10,000 to 1 in 250,000 anesthetic administrations [26]. Although the critical nature of both conditions, the effective management of MH with dantrolene has reduced mortality rates to 2 and 6%, compared to the higher mortality associated with PRIS [27, 28]. The impact of necessary equipment on personnel resources remains uncertain, with studies indicating low substance consumption over extended periods, yet lacking clear data for short term usage [29‐32]. As an approach to clarify the alleged additional effort, we conducted a study to evaluate the advantages of corresponding sedation procedures with an assessment of the workload in a cohort of patients undergoing valve surgery. The primary study outcome focused on the time to extubation, while secondary outcomes encompasses the workload required for sedation management, course of the surgical procedure, neurocognitive recovery time, setup duration, incidence of acute renal failure, and catecholamine usage, among others.

Among 1604 patients screened, 175 met eligibility criteria. Written informed consent was obtained from 100 patients. Within the intended study period, six patients were excluded, with specifics detailed in Fig. 1.

Accordingly, 47 patients were treated with propofol in the postoperative phase, while 47 patients were sedated with sevoflurane. Patient characteristics are presented in Table 1.

Our analysis revealed no significant differences between the study groups in terms of surgical procedures, demographics, pre-existing medical conditions, or the EuroSCORE II risk stratification [37]. The laboratory assessments conducted six hours after admission indicated a significant difference in creatinine levels, with a corresponding increase within the propofol group, albeit without a corresponding increase in acute renal failure incidents. Furthermore, a significant decrease in the interleukin 6 level was observed six hours after ICU admission. A differential analysis of the evaluated clinical and laboratory treatment parameters is delineated in Table 2.

The median preparation time for sevoflurane sedation was 26:31 min (IQR 23:38; 31:54) compared to 18:35 min (IQR 13:23; 23:19) for propofol-based sedation, indicating a significant additional median preparation time of 9:56 min (± 4:16 min) for additional (p < 0.001). During the sedation phase, sufficient sedation (RASS -3/-4) was achieved at a MAC₅₀ of 0.7 (± 0.1), with no significant need for additional sedatives to attain sufficient sedation depth. The corresponding average consumption of propofol was 841 mg ± 1498 mg, while for sevoflurane, it was 14 ± 4 ml.

Sevoflurane treatment was associated with significantly reduced times until eye opening (86 ± 28 min versus 151 ± 71 min; p < 0.001), command compliance (hand grip) (93 ± 38 min versus 164 ± 75 min; p < 0.001) and extubation (91 ± 39 min versus 167 ± 77 min; p < 0.001) following cessation of postsurgical sedation, as detailed in Fig. 2.

No significant difference in the incidence of delirium on the first postoperative day was observed between the two groups (propofol, n = 3, 6.4% versus sevoflurane, n = 4, 8.5%, p = 0.665) or on the third postoperative day (n = 4, 8.5% vs. n = 8, 17.0%), p = 0.416.

In terms of catecholamine usage for circulatory stabilization during deep sedation, no significant differences were observed in norepinephrine dosages between treatments within the first 60 min in the ICU: 485 ± 408 µg for sevoflurane treatment and 735 ± 1876 µg for propofol treatment. The numbers of patients requiring the antidiuretic hormone analogue vasopressin (n = 7 versus 1, p = 0.007), as well as epinephrine (n = 11 versus 4, p = 0.033), were significantly higher under propofol sedation.

Throughout the monitoring of potentially adverse effects, oxygenation levels (p = 0.710) or ventilation parameters exhibited no significant variances, and there was no severe lactate accumulation. According to international diagnostic criteria, nine patients in the propofol group experienced acute renal failure compared with four in the volatile sedation group (p = 0.149); all resolving without intervention or renal replacement therapy. Postoperative nausea and vomiting were noted once in each treatment group. We did not observe any liver failure in the study cohort, and there were no in-house fatalities reported across both groups.

In our cohort of open cardiac valve surgery patients who underwent cardiopulmonary bypass intraoperatively, a significant time advantage with regard to the ability to extubate under short-term sedation was observed, corroborating the findings of prior research [17, 19]. After 60 min of surveillance, the time to extubation was notably shorter for sevoflurane (31 min), compared to propofol-induced sedation (107 min). This aligns with several comprehensive analyses of published studies on the utilization of volatile sedatives in ICU environments, which have highlighted the effectiveness of agents such as isoflurane and sevoflurane across varied patient populations [19, 38, 39]. A salient point of convergence among these studies is the observed significant reduction in time to extubation, a result that aligns closely with our own observations. Furthermore, our study allowed us to quantify the time required to set up and dismantle a corresponding ACD system for bedside application in the ICU.

We were able to show that a minimal technical effort of less than ten minutes enabled a significant acceleration of the awakening, exceeding one hour, thus permitting an earlier assessment of the patient's neurological condition. The increased technical effort seamlessly integrated into clinical workflows, proving its non-disruptive and delegable nature, and was well accepted by the ICU staff, including physicians, nurses, and surgeons.

The consumption rates of propofol and sevoflurane observed, were surprisingly higher than those previously reported, particularly for sevoflurane, where we identified an average usage of 14 milliliters, significantly surpassing the approximately 8 ml described in extended ICU sedation scenarios [15, 30, 40‐43]. This discrepancy might be attributed to the initial filling of the ACD and the saturation process of the patient's physiological compartments.

Regarding the safety profile of sevoflurane, no complications were observed during the study. The exclusions were primarily attributed to intraoperative complications or postoperative surgical bleeding. This underscores the feasibility and safety of the short-term sevoflurane sedation in cardiothoracic patients, in line accordance with previously published research [17]. Concerns regarding renal damage from elevated serum fluoride levels with the use of sevoflurane were not substantiated by our laboratory findings [44, 45]. Additionally, diverging from other studies, an increased need for catecholamines was not detected when using sevoflurane in this susceptible patient group, aligning with the cardiovascular stability observed in patients subjected to prolonged volatile sedation [44]. Conversely, propofol sedation significantly increased the necessity for vasopressin and epinephrine [46, 47], a phenomenon likely attributable to propofol’s established vasodilatory effects.

Consequently, vasodilatory effects might have been absent under deep sedation induced by midazolam; yet, the application of benzodiazepines could have notably prolonged the awakening period [48, 49].

Although we did not observe any intraprocedural cerebrovascular incidents within our cohort, such events remain a relevant risk, manageable by early detection and timely intervention. The accelerated ability to extubate patients and the consequent swift neurological assessment could diminish the delay to conduct diagnostic imaging for neuropathological identification and facilitating appropriate treatment. This principle extends to the management of unexpected coronary perfusion malfunctions. In addition to these specific valve surgery aspects, the broader impact of the link between delirium and both the length and depth of postoperative sedation warrants attention.

Economic aspects, especially the substantial investment costs for the ACD, repeatedly deter the adoption of volatile sedation in the ICU. Nonetheless, the well-documented clinical benefits of inhaled sedation are likely to justify the initial outlay and minimal additional personnel effort outweighing these concerns in the long term.

A significant decrease in interleukin 6 levels, identified within in the volatile sedation group, corresponds with the previously described anti-inflammatory effects of these sedatives. The direct benefits of mitigating proinflammatory responses, have not been investigated in the context of postoperative rehabilitation, highlighting the imperative for further exploration of this class of agents for long-term intensive care sedation.

We believe that our results are likely to be transferable to other patient groups and may demonstrate comparable benefits in terms of controllability, a minimal safety profile, and rapid awakening.

This study has several limitations. Conducted at a single center without the possibility of practitioner blinding, the research design could introduce selective sedative use, potentially skewing outcomes. Our research center had pre-existing expertise administering volatile sedation using both AnaConDa and MIRUS ACD systems, thus requiring minimal additional training for staff. Conversely, institutions adopting to ACD-based sedation techniques a new may encounter prolonged times for setup and dismantling periods for bedside functionality. Conducting a multicenter trial would be advantageous to comprehensively assess these variations. Moreover, limiting participation to individuals scheduled for elective heart valve surgery, consequently reduced the study’s generalizability to a wider spectrum of cardiac surgery patients.

Due to the short duration of sedation and the consecutive short duration of ventilation and ICU stay, we were unable to determine whether pneumonia occurred during the course of the treatment. Furthermore, as a pilot study with a limited study population, certain findings might have been more definitive with a larger study cohort.

The short-term application of volatile sedatives for intensive care sedation is associated with additional work and technical effort for the ICU staff while also significantly accelerating the wake-up response and neurological assessability following cardiothoracic valve surgery.