Does a Combined Intravenous-volatile Anesthesia Offer Advantages Compared to an Intravenous or Volatile Anesthesia Alone

A Systematic Review and Meta-analysis


BMC Anesthesiol. 2021;21(52) 

In This Article


The combination of two different hypnotics, namely volatile anesthetics and propofol to maintain anesthesia, is rather unusual. Nevertheless, from a pharmacological and practical point of view the combination of these two agents might be useful.

Even in subhypnotic doses propofol is known to have antiemetic properties.[15] Further there is a degree of exposure dependent effect of volatiles on PONV occurrence.[16] However, this may explain the significantly lower PONV rates in patients with CIVA compared to balanced anesthesia in the RR. An analysis comparing CIVA with BAL within 24 h shows only a non-significant result. Interestingly, CIVA is comparable to TIVA in regards of risk for PONV. These results have to be interpreted cautiously, as a significant heterogeneity was evident. Apart from statistical heterogeneity there are variable factors which may influence the occurrence of PONV due to non-standardized anesthetic practice. The choice of opioid for anesthesia induction, maintenance and postoperative pain therapy might have influenced the occurrence of PONV.[17,18] Some of the included studies applied different opioids for the different treatment groups like fentanyl and remifentanil for anesthesia induction and maintenance.[5] The postoperative pain therapy strategy varied in the choice of substances and was inconsistently reported. Another factor to consider was the induction agent used for anesthesia. Two studies used barbiturates in the BAL group[7,11] with a greater likelihood of PONV[19,20] and a possible overestimation of the PONV reducing effect of CIVA compared to BAL. However, the risk for postoperative vomiting due to volatiles is restricted to the early postoperative hours[16] suggesting a PONV preventive effect by adding propofol to volatiles in the early postoperative period. This PONV preventing effect diminishes within 24 h. It remains unclear if this is due to the propofol clearance under a certain plasma level threshold and prolonged effect of volatiles on the area postrema, or if there are other factors influencing the occurrence of PONV. Not all included studies reported on established risk-factors of PONV like e.g., smoking or proportion of female patients, so that there might be a disbalance between the groups especially in those studies with small group sizes although a randomization has been performed.

Propofol and volatile anesthetics such as sevoflurane act additively,[21,22] and the primary organ of elimination for propofol is the liver, whereas volatiles are eliminated through the lungs. Theoretically the use of lower doses of two additive hypnotics with different elimination pathways should result in a shorter postanesthetic recovery time, which is reflected by the time to extubation. In this meta-analysis we found no difference between the combination of intravenous and volatiles anesthetics. However, some studies (Liang and Hensel) indicate a positive effect for CIVA. The overall effect might be diminished due to the fact that all included studies used a processed intraoperative electroencephalogram to measure the depth of hypnosis. In addition, Propofol TCI was frequently used, resulting in a very precise control of hypnosis depth and regain of consciousness in the TIVA group. This could be one reason why there is only a small positive effect evident for the CIVA regime. Thus, the advantage could be greater compared to anesthesia without measuring the depth of hypnosis and with a conventional propofol infusion pump. It is striking, however, that the study with greatest benefit for the CIVA group focused on patients undergoing major abdominal operations, namely intestinal and gastric surgery, which result in a longer duration of anesthesia. The average duration of surgery was approximately 60–70 min longer than in the study by Chi and Hensel et al. and 90 min longer than in the study of Lai et al. Thus, if only one hypnotic drug is used, a prolonged surgery or anesthesia can lead to a higher accumulation in the body, resulting in a longer elimination time. Here an advantage for combined intravenous volatile anesthesia could therefore arise. To prove this, we performed a meta-regression, which showed a strong correlation between duration of surgery and time to extubation with a R2 = 0.89 (p = 0.000) (see Supplementary Figure 1). The combination of two hypnotics could have a positive effect on postanesthetic recovery time and time to extubation depending on the duration of anesthesia. However, there are inconsistent conditions about termination of the administration of anesthetics among the included studies. Chi et al. did not state the conditions of termination.[5] Hensel et al. defined the time to extubation as the time point from which the anesthetic administration was completely terminated.[6] We assume that the administered amount of sevoflurane and/or propofol may have been reduced when the end of surgery has been anticipated. Liang et al. defined the starting timepoint as the turn-off of anesthetics administration after surgery was complete.[9] Extubation was performed with a BIS value above 70 and spontaneous breathing. Lai et al. stopped anesthetic administration at the end of procedure and extubated after consciousness was regained.[8,13] These unequal conditions restrict the findings in the regression analysis.

When comparing CIVA to TIVA, we found less movement during surgery in the CIVA group. Volatile anesthetics act inter alia on the spinal cord and suppress movement.[23] This effect is significantly more pronounced for volatiles than for propofol,[24,25] which may lead to a more favorable outcome when sevoflurane or isoflurane is added to propofol. Movement during surgery might further depend on muscle relaxation and intraoperative pain control. Only two studies delivered data on movement during surgery. The study by Hensel and colleagues included 270 patients per group in various surgical procedures. They only used 0.3–0.5 mg of rocuronium once with anesthesia induction and they reported no significant difference for the intraoperative remifentanil consumption between the CIVA and TIVA group. But they observed movements during surgery in 3% vs. 14% (CIVA vs. TIVA) of the patients. The study by Lai and colleagues investigated CIVA vs. TIVA in non-intubated video-assisted thoracoscopic surgery (VATS).[8] They used laryngeal mask airway while muscle relaxants were not used. For pain management all patients received a thoracic epidural anesthesia with additional surgical intercostal blocks. The TIVA group showed a significantly higher intraoperative fentanyl consumption than the CIVA group (145 vs. 128 μg). The patients with TIVA had significantly higher rates of movement compared to the CIVA group (17 vs. 5). This limited data suggests a possible benefit for adding volatiles to suppress movements during surgery. However, more high-quality studies are needed to draw further conclusions.

Postoperative pain differed neither between CIVA and BAL nor between CIVA and TIVA. However, some studies, which investigated postoperative pain comparing TIVA to BAL, showed a beneficial effect on postoperative pain and opioid intake in TIVA. A meta-analysis by Peng and colleagues addressed this topic and found a statistically significant benefit for propofol with questionable clinical relevance. This result was accompanied by a significant heterogeneity.[26] A recent study investigating the effect of propofol on post-sternotomy pain found no effect on acute or chronic pain.[27] Postoperative pain perception is more likely to be influenced by the use of a multi-modal pain management. Dexamethasone has a strong anti-inflammatory potential and is a useful co-analgetic.[28] The purpose for using dexamethasone in the included studies was PONV prevention. Only two studies provided data on postoperative pain[6,7] of which only one used dexamethasone risk stratified in according to the PONV risk.[6] The use of dexamethason might have influenced postoperative pain perception, but as there was no significant difference in PONV risk score between the groups, the effect of dexamethasone should be equally adjusted. Barbiturates have been associated with hyperalgesia.[29] The study by Kawano et al. used the barbiturate thiamylal for anesthesia induction only for the BAL group. Recent research could not find evidence supporting the association between barbiturates and hyperalgesia.[30] So, an influence of barbiturates on measured pain is rather unlikely.

The strength of this study is to be the first meta-analysis to address this topic. We included 10 studies with 1960 patients. However, the studies included are of moderate to low quality with significant heterogeneity, which limits the significance of our results. Apart from statistical heterogeneity there is also a relevant heterogeneity from a clinical point of view. Besides the different surgical interventions, there is a huge variability in anesthetic management between the studies. Solely a small number of studies used premedication, PONV risk was inconsequently reported and PONV prophylaxis was carried out by some, while others prescribed dexamethasone to all patients. All TIVA and CIVA patients received propofol for induction of anesthesia. The BAL patients received among propofol also barbiturates. The intraoperative analgesia concepts contained lidocaine, fentanyl, remifentanil, nalbuphine and regional anesthesia and the procedure at the end of surgery with regard to turning off the anesthetic agents differed between the studies. The lack of standardization limits the comparability and explanatory power of the CIVA concept.

However, an anesthetic regimen with a comparable PONV incidence to TIVA, with less intraoperative movements and with a shorter time to extubation would be desirable from a patient, surgical and economic view. Therefore, we suggest a thoroughly planned multi-center randomized controlled trial to compare the different concepts. This study should include three treatment arms: CIVA, TIVA and BAL for standardized surgical procedures. Also a standardized anesthetic concept (including standardized risk adapted PONV prophylaxis and standardized pain control) using a processed electroencephalogram with a predefined anesthesia depth and remifentanil as sole opioid should be implemented. The CIVA concept uses two different anesthetics to reduce the overall dose and dose dependent side-effect of single substances use. Nevertheless, side effects could be relevant and should be monitored as well as its cost effectiveness.