Información de la revista
Vol. 47. Núm. 5.
Páginas 267-279 (Mayo 2023)
Descargar PDF
Más opciones de artículo
Vol. 47. Núm. 5.
Páginas 267-279 (Mayo 2023)
Original article
Acceso a texto completo
Halogenated anesthetics vs intravenous hypnotics for short and long term sedation in the intensive care unit: A meta-analysis
Anestésicos halogenados vs. hipnóticos intravenosos para sedación a corto y largo plazo en la unidad de cuidados intensivos: un metaanálisis
V. Likhvantseva,
Autor para correspondencia

Corresponding author.
, G. Landonib,c, N. Ermokhinaa, M. Yadgarova, L. Berikashvilia, K. Kadantsevaa,d, O. Grebenchikova, L. Okhinkoe, A. Kuzovleva
a V. Negovsky Reanimatology Research Institute, Moscow, Russia
b Department of Anesthesia and Intensive Care, IRCCS San Raffaele Scientific Institute, Milan, Italy
c Vita-Salute San Raffaele University, Milan, Italy
d A. Loginov Moscow Clinical Scientific Center, Moscow, Russia
e V. Demikhov Municipal Hospital №. 68, Moscow, Russia
Este artículo ha recibido
Información del artículo
Texto completo
Descargar PDF
Figuras (3)
Mostrar másMostrar menos
Material adicional (1)

To comprehensively assess peer-reviewed studies using volatile (VA) or intravenous (i/v) anesthetics for sedation in intensive care units (ICUs), with the hypothesis that the type of sedation may have an impact on survival and other clinically relevant outcomes.


Systematic review and meta-analysis of randomized and non-randomized trials.




Critically ill and postoperative patients.



Measurements and main results

Studies comparing VA versus i/v anesthetics used in the ICU settings were independently systematically searched. Finally, 15 studies (1520 patients of predominantly surgical profile needed VA sedation for less than 96h) were included. VA had no impact on all-cause mortality (very low quality of evidence, Odds Ratio=0.82 [0.60–1.12], p=0.20). However, VA were associated with a reduction in duration of mechanical ventilation (p=0.03) and increase in ventilator-free days (p<0.001). VA also reduced postoperative levels of cardiac troponin (24h), time to extubation (p<0.001) and awakening (p=0.04).


In this meta-analysis, volatile sedation vs propofol caused the increase in ventilator-free days, the reduction in the duration of mechanical ventilation, time to extubation and the troponin release in medical or surgical ICU patients, while in surgical ICU patients the time to awakening was shortened.

Volatile agents
Intravenous agents
Intensive care
Mechanical ventilation
Troponin level
Critically ill

Evaluar exhaustivamente los estudios revisados por pares que utilizan anestésicos volátiles (AV) o intravenosos (iv) para sedación en unidades de cuidados intensivos (UCI), con la hipótesis de que el tipo de sedación puede tener un impacto en la supervivencia y otros resultados clínicamente relevantes.


Revisión sistemática y metaanálisis de ensayos aleatorizados y no aleatorizados.




Se incluyeron críticamente enfermos y postoperatorios.



Mediciones y resultados principales

Los estudios que comparaban los AV vs. los anestésicos iv utilizados en la UCI se buscaron de forma independiente y sistemática. Finalmente, se incluyeron 15 estudios (1.520 pacientes de perfil predominantemente quirúrgico necesitaron sedación de AV durante menos de 96h). El AV no tuvo impacto en la mortalidad por cualquier causa (calidad de los datos probatorios muy baja, Odds Ratio=0,82 [0,60-1,12], p=0,20). Sin embargo, el AV se asoció con una reducción de la duración de la ventilación mecánica (p=0,03) y aumento de los días sin ventilación mecánica (p<0,001). La AV también redujo los niveles postoperatorios de troponina cardíaca (24 horas), el tiempo hasta la extubación (p<0,001) y el despertar (p=0,04).


En este metaanálisis, la sedación volátil vs. propofol causó el aumento de los días sin ventilación, la reducción de la duración de la ventilación mecánica, el tiempo hasta la extubación y la liberación de troponina en pacientes de la UCI médica o quirúrgica, mientras que en pacientes de la UCI quirúrgica el tiempo hasta el despertar se acortó.

Palabras clave:
Agentes volátiles
Agentes intravenosos
Cuidados intensivos
Ventilación mecánica
Nivel de troponina
Enfermedad crítica
Texto completo

Sedation, ‘the act of calming patients by the administration of sedative medications’,1 is frequently used in intensive care units (ICUs) to prevent arouse and delirium associated harm, relieve anxiety, and reduce the stress of being mechanically ventilated.2 Since agitation and anxiety occurs in about 30%–80% of patients being treated in ICU settings,1 sedation is a highly sought strategy for ICU patients.

Volatile anesthetics (VA) sedation uses isoflurane or sevoflurane to achieve the desired level of sedation.3 It is a more recent and a less frequently used strategy than the traditional intravenous strategy in the ICU setting.4 However, with the adoption of user friendly devices for VA delivery – AnaConDa (Sedana Medical, Danderyd Sweden) and MIRUS (Pall Medical, Dreieich, Germany) – VA have become increasingly popular among ICU practitioners.5,6 There are potential lungs protective properties7,8 and anti-inflammatory activity9,10 coupled with the intended endothelium-saving effect,11,12 that contribute to an increased use of VA in the ICU. Malignant hyperthermia and possible environmental pollution are the most common drawbacks of using VA.13,14

Early studies confirmed safety and feasibility of VA in the ICU settings15,16 and a meta-analysis of randomized trials found a significant reduction in time to extubation17 with findings confirmed in more recent systematic reviews.18

The aim of this systematic review and meta-analysis was to comprehensively assess published randomized and non-randomized peer-reviewed studies which compared VA and i/v anesthetics for ICU sedation, with the hypothesis that the type of sedation may have an impact on mortality and other clinically relevant outcomes.


This study was carried out in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.19 Meta-analysis is registered in PROSPERO (ID: CRD42021277313).

Search strategy

A systematic search of studies published over the past 10 years (2011–2021) was carried out in PubMed, Medline, Cochrane Central Register of Controlled Trials (CENTRAL) in the Cochrane Library, Google Scholar and the Russian Science Citation Index (RSCI) by four independent researchers (NE, LB, MY and KK). The search was carried out in the form of queries with details available in the supplemental material (Supplemental appendix 1). Additionally, the authors used the backward snowballing method (analysis of references of included articles and retrieved reviews) for further studies. We did not restrict search by language. Medical Subject Headings (MeSH) terms were applied.

Study selection

The links obtained from the database were first independently examined at the title/annotation level by two researchers (NE and LB). Randomized controlled trials, prospective and retrospective cohort studies comparing inhaled (volatile) versus i/v sedation in the intensive care unit were considered. After removing duplicates, the appropriate publications were selected. The final decision on inclusion in this study was made based on the analysis of full-text articles. Divergences were resolved by consensus.

For this study, the following inclusion criteria were used: adult patients (≥18 years) who underwent inhalation or i/v sedation in the intensive care unit (no restrictions on dose or time of administration). The exclusion criteria were mixed groups (both inhaled and i/v sedation were used), small sample size (less than 25 patients in two groups), duplicate publications, animal studies, clinical guidelines, articles without a comparison group.

Outcome measures and data extraction

Basic information about the study (design, sample size, intervention plan, inclusion criteria), information about the study objects (age, gender, underlying disease and surgery), types of anesthetics and doses, target sedation level, anesthetic-conserving device, ICU treatment outcomes (mortality, duration of mechanical ventilation, troponin level at 1 postoperative day, ventilator-free days, time to extubation, awakening time, length of stay (LOS) in ICU, hospital LOS, catecholamine requirements) was collected. We have not considered the type of anesthetic used in the surgery room. The data were independently extracted by two researchers (NE and MY) and subsequently compared with each other for verification. The primary outcome of this study was all-cause mortality (30-day mortality). Secondary endpoints were: length of stay in ICU and length of hospitalization (days), duration of mechanical ventilation (days) and ventilator-free days, catecholamine requirements, cardiac troponin levels on the first postoperative day, time to extubation (min), and awakening time (min).

Internal validity and risk of bias assessment

The internal validity and risk of bias of the included studies were assessed by two peer reviewers (MY and LB) and peer-reviewed by a third (VL) according to the latest version of the “ROB 2” (the Cochrane tool for assessing risk of bias in randomized trials) and “ROBINS-I” (Risk Of Bias In Non-randomized Studies of Interventions) tools.20,21 Discrepancies in assessment were resolved by consensus. Publication bias was assessed using the Egger's test (MedCalc Statistical Software, version 19.5.6),22,23 and also by visual examination of the “funnel plot” charts. We also used a GRADE systematic approach to rate the certainty of evidence.24 Two review authors (MY and LB) worked independently to assess the quality of evidence, disagreements were resolved by consensus.

Data analysis and synthesis

To calculate and visualize the results of the meta-analysis in forest plots, the Cochrane tool “RevMan, version 5.3” was used. The heterogeneity of studies was assessed using the Cochran's Q test, and the degree of statistical agreement was measured using the coefficient of heterogeneity I2. The quantitative results of individual studies were brought to the form “mean±standard deviation” and the standardized difference of mean values (SMD) and its 95% confidence interval were calculated. Since many quantitative parameters in this meta-analysis a priori have a deeply skewed distribution, we used the log transformation for summary data of ventilator-free days, time to extubation, time to awakening and ICU/hospital length of stay.25 We used Cochrane handbook recommendations to re-express SMDs using rules of thumb for effect sizes (<0.40=small effect, 0.40–0.70=moderate effect, >0.70=large effect).20 Binary research results were used to calculate the odds ratio with the corresponding 95% confidence interval (CI) using the inverse variance method (Mantel–Haenszel method). For a pooled estimate of the magnitude of the standardized mean difference, two models were used: a fixed-effects model (in the case of low statistical inconsistency, I2<50%) and a random-effects model (I250% and/or p<0.05). Statistical significance was set at 0.05 for hypothesis testing.

Sensitivity analysis

Sensitivity analyses were performed by considering additional subgroups – analyzing only studies with a low or moderate risk of bias (Hi-QOL studies), only RCT/non-randomized studies and by sequentially removing each study and reanalyzing the remaining data set (producing a new analysis for each study removed). Additionally, a subgroup of studies which used a prolonged sedation (>12h) regimen was evaluated.

ResultsStudy characteristics

During the initial search, 427 articles were found, of which 31 were eligible. Upon careful reading of the full-text articles, 16 studies were excluded (Supplemental Table 1). Ultimately, 15 full-text articles published between January 2011 and July 2021were included in the meta-analysis. A flowchart illustrating the study selection process is presented in Fig. 1.

Figure 1.

Flowchart used to select the final 15 manuscripts.


The characteristics of the included studies are summarized in Table 1.

Table 1.

Characteristics and description of the 15 trials included in the meta-analysis.

Study  Design  N, VA  N, i/v  Volatile agent (dose)  Location volatile started  Anesthetic-conserving device  Comparator (dose)  Journal  Population 
Bellgardt M et al. (2016)26  Retrospective cohort  72  128  Isofluran MAC 0.3–0.8  Operating room  AnaConDa  Propofol 2–4mg/kg/h; midazolam 0.05–0.2 mg/kg/h  Eur J Anaesthesiol  Postoperative surgical patients 
Guerrero Orriach JL et al. (2013)27  RCT  20  40  Sevoflurane MAC 0.5–0.7  Operating room  AnaConDa  Propofol TCI 1–1.5mg/ml  J Crit Care  Postoperative cardiac surgery patients 
Guinot PG et al. (2020)28  RCT  42  39  Sevoflurane  Operating room  MIRUS  Propofol  Medicine  Postoperative cardiac surgery patients 
Hellström J et al. (2012)29  RCT  50  50  Sevoflurane ET 0.5–1.0%  ICU  AnaConDa  Propofol started at 2mg/kg/h  Scand Cardiovasc J  Postoperative cardiac surgery patients 
Jabaudon M et al. (2017)30  RCT  25  25  Sevoflurane started at 6ml/h  ICU  AnaConDa  Midazolam started at 0.1mg/kg/h  Am J Respir Crit Care Med  Adult patients with ARDS 
Jerath A et al. (2015)30  RCT  67  74  Sevoflurane or isofluran MAC 0.1–0.3  Operating room  AnaConDa  Propofol 0.6–1.5mg/kg/h  Crit. Care Med  Postoperative cardiac surgery patients 
Jung S et al. (2020)32  Retrospective observational  25  24  Sevoflurane ET 0.5%  ICU  AnaConDa  Propofol 3.66±1.30μg/kg/h  Acute Crit Care  Postoperative surgical patients 
Krannich А et al. (2017)33  Retrospective observational propensity-matched  110  110  Isofluran ET 0.5–1.5%  ICU  AnaConDa  Midazolam 0.03–0.2mg/kg/h  Crit Care Med  Adult patients with non-traumatic cardiac arrest 
Marcos-Vidal JM et al. (2014)34  Prospective cohort  67  62  Sevoflurane ET 0.5–1.0%  ICU  AnaConDa  Propofol 1–4mg/kg/h  Heart Lung Vessel  Postoperative cardiac surgery patients 
Meiser A et al. (2017)35  Retrospective case match  19  19  Isofluran 3–10ml/h  ICU  AnaConDa  Propofol. midazolam  Respir Care  Adult patients with ARDS 
Mesnil M et al. (2011)36  RCT  19  28  Sevoflurane ET 0.5%  ICU  AnaConDa  Propofol started at 2mg/kg/h  Intensive Care Med  Adult patients, more than 24h of sedation for mechanical ventilation 
Plotnikov GP et al. (2014)37  RCT  20  58  Sevoflurane 5.0±2.3ml/h  ICU  AnaConDa  Propofol 4.35±1.75mg/kg/h. dexmedetomidin 0.35±0.2mg/kg/h  Byulleten’ NTSSSKH im. A.N. Bakuleva RAMN  Postoperative cardiac surgery patients 
Resepov NА et al. (2017)38  RCT  20  20  Sevoflurane started at 2ml/h (max 5ml/h)  ICU  AnaConDa  Propofol started at 0.5mg/kg/h  Messenger of Anesthesiology and Resuscitation  Adult patients with sepsis-associated delirium 
Soro M et al. (2012)39  RCT  36  37  Sevoflurane ET 0.5–1.0%  Operating room  AnaConDa  Propofol 1–4mg/kg/h  Eur J Anaesthesiol  Postoperative cardiac surgery patients 
Staudacher DL et al. (2018)40  Retrospective observational propensity-matched  36  178  Isofluran ET 0.5–1.0%  ICU  AnaConDa  Propofol  J Crit Care  Adult patients after cardiopulmonary resuscitation 

Abbreviations: VA: volatile anesthetics, i/v: intravenous anesthetics, ICU: intensive care unit, MAC: minimum alveolar concentration, RCT: randomized controlled trial, ET: end-tidal, TCI: target-controlled infusion.

The meta-analysis included 1520 patients: VA (628 patients); i/v sedation (892 patients). Among the 15 included studies, 9 were RCTs. The following settings were represented: seven studies in sedation in postoperative cardiac surgery patients,27–29,31,34,37,39 two articles in postoperative surgical patients,26,32 one study in patients who were assigned to more than 24h of mechanical ventilation sedation,36 five studies in the intensive care unit critically ill patients (non-traumatic cardiac arrest and successful CPR,33,40 sepsis-associated delirium38 and moderate to severe ARDS30,35).

Fourteen included studies used the AnaConDa device in the ICU for inhaled sedation, and only one study used MIRUS.28 Among the 15 included studies, 10 compared sevoflurane with propofol, 4 compared isoflurane with propofol and midazolam, and one study had multiple comparisons.

Four studies30,32,36,38 used long-term sedation (12h or more). Five of 9 RCTs and two of six non-randomized studies had medium or high quality (Supplemental Figures 1 and 2).

Quantitative data synthesis

Fourteen studies (8 RCTs) including 1453 patients reported mortality data which were not different between groups 112/609 (18.4%) in the VA group versus 202/844 (23.9%) in the i/v group (Odds Ratio (OR)=0.82 [0.60–1.12]; p-value for effect is 0.20; p-value for heterogeneity=0.15; I2=34%, Fig. 2; Table 2).

Figure 2.

Forest plot for hospital mortality representing the odd's ratio for volatile vs. intravenous sedation effects on all-cause mortality for the included studies. The plot displays the study, sample size, odds ratio (OR), confidence interval (CI), and p-value. The size of the squares indicates the weight of the studies (taking into account sample size and standard deviations); the diamond represents the pooled OR with CI. Hi-QOL – studies with low–moderate risk of bias.

Table 2.

Outcomes and sensitivity analysis.

Outcome  Trials  N, VA  N, i/v  SMD/OR  95% CI  p-value for overall effect  p-value for heterogeneity  I2, %  p-value for publication bias 
All studies  14  609  844  0.82  0.60–1.12  0.20  0.15  34  p=0.19 
RCTs  280  323  1.08  0.50–2.32  0.85  0.61  p=0.02 
Non-randomized trials  329  521  0.87  0.72–1.05  0.15  0.05  62  p=0.86 
Studies with low–moderate bias  289  429  1.12  0.77–1.64  0.55  0.75  p=0.60 
Prolonged sedation  70  69  0.73  0.30–1.79  0.50  0.70  p=0.001 
Duration of mechanical ventilation
All studies  371  616  −0.46  −0.88–(−0.04)  0.03*  <0.001  88  p=0.49 
RCTs  134  181  −0.60  −1.31–0.10  0.09  <0.001  88  p=0.85 
Non-randomized trials  237  435  −0.31  −0.87–0.26  0.29  <0.001  90  p=0.86 
Studies with low–moderate bias  191  333  −0.52  −0.72–(−0.33)  <0.001*  0.09  54  p=0.44 
Prolonged sedation  64  73  −0.11  −0.45–0.22  0.51  0.70  p=0.31 
Troponin level at 1 postoperative day
All studies  179  171  −0.52  −0.84–(−0.20)  0.001*  0.10  52  p=0.49 
RCTs  112  109  −0.50  −0.98–(−0.03)  0.04*  0.06  65  p=0.35 
Studies with low–moderate bias  62  59  −0.61  −1.56–0.34  0.21  0.02  82  p<0.001 
Ventilator-free days
All studies  202  400  0.46  0.28–0.64  <0.001*  0.79  p=0.40 
RCTs  75  75  0.49  0.17–0.82  0.003*  0.29  10  p<0.001 
Non-randomized trials  127  325  0.44  0.23–0.66  <0.001*  0.78  p=0.59 
Studies with low–moderate bias  61  203  0.44  0.14–0.75  0.004*  0.42  p<0.001 
Time to extubation
All studies  177  191  −1.59  −2.26–(−0.91)  <0.001*  <0.001  87  p=0.06 
Awakening time
All studies  122  280  −1.30  −2.54–(−0.06)  0.04*  <0.001  95  p=0.41 
RCTs  86  102  −1.99  −3.92–(−0.06)  0.04*  <0.001  94  p<0.001 
All studies  13  588  794  0.01  −0.17–0.18  0.93  0.008  55  p=0.88 
Hospital LOS
All studies  367  589  −0.06  −0.35–0.23  0.69  <0.001  76  p=0.03 
Catecholamine requirements
All studies  362  550  1.13  0.79–1.61  0.50  0.41  p=0.43 

Abbreviations: VA: volatile anesthetics, i/v: intravenous anesthetics, CI: confidence interval; ICU: intensive care unit; LOS: length of stay; RCT: randomized controlled trial; OR: odds ratio; SMD: standardized mean difference.


Significant overall effect.

Analysis of subgroups of RCTs, non-randomized, Hi-QOL studies (studies with low–moderate risk of bias) and studies with long-term sedation did not reveal an association of sedation with mortality. Funnel plot for mortality is presented in Supplement (Supplemental Figure 3).

A pooled analysis of data from five studies (606 patients, four studies with short-term sedation, two RCTs) showed that patients on VA had an increase in ventilator-free days (moderate effect size: SMD=0.46 [0.28–0.64]; p-value for the effect <0.001; p-value for heterogeneity=0.79; I2=0%, Fig. 3; Table 2). The results were confirmed in all sub-analyses: RCTs (p=0.003), non-randomized (p<0.001) and Hi-QOL studies (2 trials, p=0.004).

Figure 3.

Forest plot for ventilator-free days. The plot displays the study, sample size, log-transformed standardized mean difference (SMD), confidence interval (CI), and p-value. The size of the squares indicates the weight of the studies (taking into account sample size and standard deviations); the diamond represents the pooled SMD with CI. Hi-QOL – studies with low–moderate risk of bias.


A pooled analysis of 987 patients (9 studies, 5 RCTs) showed that sedation with VA was associated with a reduction in duration of mechanical ventilation (SMD=−0.46 [−0.88 to −0.04] – moderate effect size; p-value for effect=0.03; p-value for heterogeneity <0.001; I2=88%, Supplemental Figure 4, Table 2). The results were confirmed in Hi-QOL studies (5 trials, p<0.001), but not when considering only RCTs (p=0.09) or non-randomized studies (p=0.29) (Table 2). When considering only studies using prolonged sedation there was also no significant effect (p=0.51).

Four studies with short-term sedation using sevoflurane (3 RCTs, 350 patients) reported cardiac troponin levels (1 day after surgery) with postoperative cardiac surgery patients on VA having a statistically significantly lower troponin levels (moderate effect size: SMD=−0.52 [from −0.84 to −0.20]; p-value for effect 0.001; p-value for heterogeneity=0.10; I2=52%, Supplemental Figure 5; Table 2). The result was robust when considering RCTs (p=0.04) but not Hi-QOL studies (2 trials, p=0.21).

Data from 4 studies (all RCTs, 3 with short-term sedation) in 368 patients showed that sedation with VA was associated with a decrease in time to extubation in both medical and surgical patients (large effect size: SMD=−1.59 [−2.26 to −0.91]; p-value for effect <0.001; p-value for heterogeneity <0.001; I2=87%, Supplemental Figure 6; Table 2). Results on awakening time (large effect size: SMD=−1.30 [−2.54 to −0.06]; p-value for effect 0.04; p-value for heterogeneity <0.001; I2=95%) were confirmed when non-randomized studies were excluded (p=0.04, Supplemental Figure 7 and Table 2). The difference in the awakening time was established only for surgical patients, not for medical ones. However, only one study was of high quality. Moreover, the time to extubation and the awakening time were less in VA group than in i\v group, regardless of the duration of sedation (p<0.05). Despite the depth of sedation across the majority of the studies were approximately similar (from RASS=−1 to RASS=−3), these results were difficult to compare them to the ones in the other studies where the depth of sedation were measured according to other scales.

No differences were observed in length of ICU stay (p=0.93), length of hospital stay (p=0.69), and need for catecholamines (p=0.5, Table 2 and Supplemental Figures 8–10).

Certainty of evidence for all studied outcomes was qualified using GRADE approach for RCTs. Overall, very low quality of evidence shows that volatile sedation has no impact on hospital mortality. Certainty of evidence for other outcomes ranged from very low to high. The reasons for the decrease in the quality of evidence are summarized and presented in Table 3.

Table 3.

Certainty of evidence from RCTs for studied outcomes (Grade approach).

Outcome  No. of participants and RCTs  Risk of bias  Inconsistency  Indirectness  Imprecision  Publication bias  Upgrades  Overall quality of evidence 
Mortality statement
No effect on ICU patients  525, 7 RCTs  Not serious (0)  Serious (−1)  Not serious (0)  Very serious (−2)  Serious (−1)  None  ⊕OOOVery low 
Duration of mechanical ventilation (DMV) statement
Volatile anesthetics do not reduce DMV in ICU patients  315, 5 RCTs  Serious (−1)  Very serious (−2)  Not serious (0)  Very serious (−2)  Not serious (0)  None  ⊕OOOVery low 
Troponin level at 1 p/o day statement
Volatile anesthetics reduce troponin level after cardiac surgery  221, 3 RCTs  Not serious (0)  Serious (−1)  Not serious (0)  Serious (−1)  Not serious (0)  None  ⊕⊕OOLow 
Ventilator-free days (VFD) statement
Volatile anesthetics increase VFD in ICU patients  150, 2 RCTs  Serious (−1)  Very serious (−2)  Not serious (0)  Not serious (0)  Serious (−1)  None  ⊕OOOVery low 
Time to extubation (TE) Statement
Volatile anesthetics reduce TE in ICU patients  368, 4 RCTs  Serious (1)  Not serious (0)  Not serious (0)  Not serious (0)  Not serious (0)  None  ⊕⊕⊕OModerate 
Awakening time (AT) statement
Volatile anesthetics reduce AT in ICU patients  188, 2 RCTs  Serious (−1)  Serious (−1)  Not serious (0)  Serious (−1)  Not serious (0)  None  ⊕OOOVery low 
LOS in ICU statement
No effect on ICU patients  532, 7 RCTs  Not serious (0)  Serious(−1)  Not serious (0)  Not serious (0)  Not serious (0)  None  ⊕⊕⊕OModerate 
Hospital LOS statement
No effect after cardiac surgery  455, 5 RCTs  Not serious (0)  Serious (−1)  Not serious (0)  Serious (−1)  Not serious (0)  None  ⊕⊕OOLow 
Catecholamine requirements statement
No effect on ICU patients  520, 6 RCTs  Serious (−1)  Not serious (0)  Serious (−1)  Serious (−1)  Not serious (0)  None  ⊕OOOVery low 

Abbreviations: RCT, randomized controlled trial; p/o, postoperative; ICU, intensive care unit; LOS, length of stay.

Key: 0, no evidence downgrade; −1, serious limitation; −2, very serious limitation; +1, evidence upgrade. Baseline evidence level for RCTs: high.


Overall analysis showed that the use of VA for sedation in patients in the ICU does not affect hospital mortality (OR=0.82 [0.60–1.12]; p=0.20). This finding is consistent with previous meta-analyses.17,41 The level of evidence obtained in current study for mortality outcome was downgraded to very low due to clinical inconsistency, presence of publication bias and serious imprecision (wide confidence interval for overall effect). It should be noted that the presence of publication bias was found, and statistical heterogeneity was not high (I2=34%).

In this study, inhalation sedation in the ICU was associated with an increase in ventilator-free days, but the final level of evidence was very low. To the best of the authors’ knowledge, this was the first meta-analysis which has shown impact of VA on patients’ ventilator-free days.

This meta-analysis performed mostly from studies including surgical patients suggests that volatile sedation is associated with a reduction in duration of mechanical ventilation. In a meta-analysis by Jerath et al. (2017)41 (523 patients, 8 studies) the duration of mechanical ventilation was lower in VA group (p=0.03).

According to our data, there is moderate quality evidence that VA reduce time to extubation in ICU patients and very low quality evidence that VA reduce time to awakening (when including only RCTs). In the meta-analysis of Landoni et al. (2016) the use of halogenated agents was associated with a significant reduction in time to extubation (p<0.00001).17 Meta-analysis of Kim et al. (13 studies, 1027 patients) showed that volatile sedation delivered through a special AnaConDa device in the ICU shortens the awakening (p=0.004) and extubation time (p<0.001) compared to i/v anesthetics sedation.4 A similar result was obtained in a meta-analysis by Jerath et al. (2017),41 which also showed a decrease in extubation time when using inhaled sedation compared with i/v sedation (p<0.00001).

Halogenated anesthetics are known to have cardioprotective effects,26,42 which, according to our study, was expressed in a decrease in the content of specific markers – cardiac troponins on the first postoperative day in postoperative cardiac surgery patients. This fact has also been confirmed by a number of other studies.43,44 Results from a meta-analysis by Kim et al. (2017) also indicate a decrease in troponin levels in patients with VA, and the effect size was largest between 12 and 24hours after ICU admission (p=0.003).4 Patients in VA group had 0.71ngml−1 (95% CI: 0.23 to 1.2) lower troponin levels according to a meta-analysis by Spence et al. (2017).45

According to the results of our research there was no impact of VA on catecholamine requirements, which was confirmed in the sensitivity analysis and in the review of studies with prolonged sedation, however the evidence was of very low quality. The impact of VA and i/v anesthetics on this outcome has not been assessed in other meta-analyses to date.

No relationship was found between the type of sedation and length of ICU stay and in hospital LOS in current meta-analysis. This fact has also been confirmed in other studies. Thus, in a meta-analysis of Landoni et al. (2016) no relationship was found between VA and ICU LOS (p=0.13) and hospital LOS (p=0.08).17 Similar results were obtained in meta-analyses of Jerath et al. (hospital LOS: p=0.74), Kim et al. (ICU LOS: p=0.513; hospital LOS: p=0.059) and Spence et al. (ICU LOS: p=0.65; hospital LOS: p=0.11).4,41,45

We were the first to use the GRADE approach to assess the quality of the evidence regarding the effect of sedation on mortality. Increase of ventilator free days due to volatile sedation firstly depicted in current meta-analyses and it seems to be a new and important piece of information that have been brought to our knowledge. It looks like that the next strength of our meta-analysis is that we performed very thorough sensitivity analyses, looking at both RCT alone and low-risk of bias studies.

Authors acknowledge that this study also has some limitations. All RCTs included in the meta-analysis were single center studies and therefore external validity is limited. The meta-analysis was performed mostly from studies included surgical patients so the cohort of medical critically ill patients is poorly represented. The sample sizes of the included studies were small and only four studies used long-term sedation (12h or longer). Clinical heterogeneity (inconsistency) of studies (mixed surgery, varying depth of sedation, different time points for measuring outcomes) and risks of bias also reduced the level of evidence. We failed to spot any difference in effect on the investigated parameters between isoflurane and sevoflurane been used for inhaled sedation. Thus, some of the results may have been insufficient. We also have not considered the type of anesthetic used in the surgery room. Finally, we were unable to analyze other factors such as laboratory parameters, cognitive status and major morbidity as most of the included studies did not report these data.


Volatile anesthetics had no effect on hospital mortality (very low evidence), however they reduced duration of mechanical ventilation, troponin level, time to extubation, and the time to awakening in medical and surgical ICU patients. Thus, it seems like, despite of new data, brought by current meta-analysis, further large high-quality randomized controlled studies are to be performed to provide more knowledge towards this still unclear challenge due to the low power of published studies.

Author contributions

Conceived and designed the análisis: VL, GL, NE, MY, LB, KK, OG, AK

Collected the data: NE, MY, LB, KK

Contributed data or analysis tools: GL

Performed the analysis: MY, KK

Wrote the paper: VL, GL, NE, MY, LB, KK, OG, LO, AK

All authors have approved the final article.



Conflict of interest


Appendix A
Supplementary data

The following are the supplementary data to this article:

D. Neme, Z. Aweke, H. Micho, S. Mola, B. Jemal, T. Regasa, et al.
Evidence-based guideline for adult sedation, pain assessment, and analgesia in a low resource setting intensive care unit: review article.
Int J Gen Med, 13 (2020), pp. 1445-1452
M.C. Balas, G.L. Weinhouse, L. Denehy, G. Chanques, B. Rochwerg, C.J. Misak, et al.
Interpreting and implementing the 2018 pain, agitation/sedation, delirium, immobility, and sleep disruption clinical practice guideline.
Crit Care Med, 46 (2018), pp. 1464-1470
S. Misra, T.A. Koshy.
Review of the practice of sedation with inhalational anaesthetics in the intensive care unit with the AnaConDa(®) device.
Indian J Anaesth, 56 (2012), pp. 518-523
H.Y. Kim, J.E. Lee, H.Y. Kim, J. Kim.
Volatile sedation in the intensive care unit: a systematic review and meta-analysis.
Medicine (Baltimore), 96 (2017), pp. e8976
H. Bomberg, H.V. Groesdonk, M. Bellgardt, T. Volk, A. Meiser.
AnaConDaTM and MirusTM for intensive care sedation, 24h desflurane versus isoflurane in one patient.
Springerplus, 5 (2016), pp. 420
R. Blondonnet, A. Quinson, C. Lambert, J. Audard, T. Godet, R. Zhai.
Use of volatile agents for sedation in the intensive care unit: a national survey in France.
PLoS One, 16 (2021), pp. e0249889
B. O’Gara, D. Talmor.
Lung protective properties of the volatile anesthetics.
Intensive Care Med, 42 (2016), pp. 1487-1489
A. Jerath, N.D. Ferguson, B. Cuthbertson.
Inhalational volatile-based sedation for COVID-19 pneumonia and ARDS.
Intensive Care Med, 46 (2020), pp. 1563-1566
X. Guo, J. Deng, B. Zheng, H. Liu, Y. Zhang, Y. Ying.
HDAC1 and HDAC2 regulate anti-inflammatory effects of anesthetic isoflurane in human monocytes.
Immunol Cell Biol, 98 (2020), pp. 318-331
A. Suleiman, A.B. Qaswal, M. Alnouti, M. Yousef, B. Suleiman, M.E. Jarbeh.
Sedating mechanically ventilated COVID-19 patients with volatile anesthetics: insights on the last-minute potential weapons.
Sci Pharmaceut, 89 (2021), pp. 6
J. Dong, M. Xu, W. Zhang, X. Che.
Effects of sevoflurane pretreatment on myocardial ischemia-reperfusion injury through the AKT/hypoxia-inducible factor 1-alpha (HIF-1α)/vascular endothelial growth factor (VEGF) signaling pathway.
Med Sci Monit, 25 (2019), pp. 3100-3107
M. Kim, A. Ham, K.Y. Kim, K.M. Brown, H.T. Lee.
The volatile anesthetic isoflurane increases endothelial adenosine generation via microparticle ecto-5′-nucleotidase (CD73) release.
T. Migita, K. Mukaida, M. Kobayashi, H. Hamada, M. Kawamoto.
Sevoflurane-induced malignant hyperthermia.
Acta Anaesthesiol Scan, 56 (2012), pp. 351-356
J.S. Yasny, J. White.
Environmental implications of anesthetic gases.
Anesth Prog, 59 (2012), pp. 154-158
K.D. Röhm, M.W. Wolf, T. Schöllhorn, A. Schellhaass, J. Boldt, S.N. Piper.
Short-term sevoflurane sedation using the Anaesthetic Conserving Device after cardiothoracic surgery.
Intensive Care Med, 34 (2008), pp. 1683-1689
E. L’her, D. Lenaïg, R. Pili, G. Prat, J.M. Tonnelier, M. Lefevre, et al.
Feasibility and potential cost/benefit of routine isoflurane sedation using an anesthetic-conserving device: a prospective observational study.
Resp Care, 53 (2008), pp. 1295-1303
G. Landoni, L. Pasin, L. Cabrini, A.M. Scandroglio, M.B. Redaelli, C.D. Votta.
Volatile agents in medical and surgical intensive care units: a meta-analysis of randomized clinical trials.
J Cardiothorac Vasc Anesth, 30 (2016), pp. 1005-1014
H. Wang, C. Wang, Y. Wang, H. Tong, Y. Feng, M. Li, et al.
Sedative drugs used for mechanically ventilated patients in Intensive Care Units: a systematic review and network meta-analysis.
Curr Med Res Opin, 35 (2019), pp. 435-446
A. Liberati, D.G. Altman, J. Tetzlaff, C. Mulrow, P.C. Gøtzsche, J.P. Ioannidis, et al.
The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration.
J Clin Epidemiol, 62 (2009), pp. e1-e34
Cochrane handbook for systematic reviews of interventions.
J.A. Sterne, M.A. Hernán, B.C. Reeves, J. Savović, N.D. Berkman, M. Viswanathan, et al.
ROBINS-I: a tool for assessing risk of bias in non-randomised studies of interventions.
BMJ, 355 (2016), pp. i4919
C.B. Begg, M. Mazumdar.
Operating characteristics of a rank correlation test for publication bias.
Biometrics, 50 (1994), pp. 1088-1101
M. Egger, G.D. Smith, M. Schneider, C. Minder.
Bias in meta-analysis detected by a simple, graphical test.
G.H. Guyatt, A.D. Oxman, G. Vist, R. Kunz, Y. Falck-Ytter, P. Alonso-Coello, GRADE Working Group: Rating quality of evidence and strength of recommendations GRADE: an emerging consensus on rating quality of evidence and strength of recommendations, et al.
J.P. Higgins, I.R. White, J. Anzures-Cabrera.
Meta-analysis of skewed data: combining results reported on log-transformed or raw scales.
Stat Med, 27 (2008), pp. 6072-6092
M. Bellgardt, H. Bomberg, J. Herzog-Niescery, B. Dasch, H. Vogelsang, T.P. Weber, et al.
Survival after long-term isoflurane sedation as opposed to intravenous sedation in critically ill surgical patients: retrospective analysis.
Eur J Anaesthesiol, 33 (2016), pp. 6-13
J.L. Guerrero Orriach, M.G. Ortega, M.R. Aliaga, P. Iglesias, M.R. Navarro, J.C. Manas.
Prolonged sevoflurane administration in the offpump coronary artery bypass graft surgery: beneficial effects.
J Crit Care, 28 (2013), pp. 879
P. Guinot, O. Ellouze, S. Grosjean, V. Berthoud, T. Constandache, M. Radhouani, et al.
Anaesthesia and ICU sedation with sevoflurane do not reduce myocardial injury in patients undergoing cardiac surgery. A randomized prospective study.
J. Hellström, A. Öwall, P. Sackey.
Wake-up times following sedation with sevoflurane versus propofol after cardiac surgery.
Scand Cardiovasc J, 46 (2012), pp. 262-268
M. Jabaudon, P. Boucher, E. Imhoff, R. Chabanne, J.S. Faure, L. Roszyk, et al.
Sevoflurane for sedation in acute respiratory distress syndrome. A randomized controlled pilot study.
Am J Respir Crit Care Med, 195 (2017), pp. 792-800
A. Jerath, S. Beattie, T. Chandy, J. Karski, G. Djaiani, V. Rao, et al.
Volatile-based short-term sedation in cardiac surgical patients: a prospective randomized controlled trial.
Crit Care Med, 43 (2015), pp. 1062-1069
S. Jung, S. Na, H.B. Kim, H.J. Joo, J. Kim.
Inhalation sedation for postoperative patients in the intensive care unit: initial sevoflurane concentration and comparison of opioid use with propofol sedation.
Acute Crit Care, 35 (2020), pp. 197-204
A. Krannich, C. Leithner, M. Engels, J. Nee, V. Petzinka, T. Schröder, et al.
Isoflurane sedation on the icu in cardiac arrest patients treated with targeted temperature management: an observational propensity-matched study.
Crit Care Med, 45 (2017), pp. 384-390
J.M. Marcos-Vidal, R. Gonzalez, C. Garcia, C. Soria, M. Galiana, B. De Prada.
Sedation with sevoflurane in postoperative cardiac surgery: influence on troponin T and creatinine values.
Heart Lung Vessel, 6 (2014), pp. 33-42
A. Meiser, H. Groesdonk, S. Bonnekessel, T. Volk, H. Bomberg.
Inhalation sedation in subjects with ARDS undergoing continuous lateral rotational therapy.
Respir Care, 63 (2018), pp. 441-447
M. Mesnil, X. Capdevila, S. Bringuier, P.O. Trine, Y. Falquet, J. Charbit, et al.
Long-term sedation in intensive care unit: a randomized comparison between inhaled sevoflurane and intravenous propofol or midazolam.
Intensive Care Med, 37 (2011), pp. 933-941
G.P. Plotnikov, E.V. Grigor’ev, D.L. Shukevich.
Options for sedation in mechanically ventilated patients.
Byull NTSSSKH AN Bakul RAMN, 15 (2014), pp. 45-52
Rezepov NА, O.N. Ulitkina, Skripkin YuV, T.S. Zabelina, V.V. Likhvantsev.
Inhalation sedation in those with sepsis-associated delirium.
Messenger Anesthesiol Resus, 14 (2017), pp. 48-54
M. Soro, L. Gallego, V. Silva, M.T. Ballester, J. Lloréns, A. Alvariño, et al.
Cardioprotective effect of sevoflurane and propofol during anaesthesia and the postoperative period in coronary bypass graft surgery: a double-blind randomised study.
Eur J Anaesthesiol, 29 (2012), pp. 561-569
D. Staudacher, S. Hamilton, D. Duerschmied, P.M. Biever, M. Zehender, C. Bode, et al.
Isoflurane or propofol sedation in patients with targeted temperature management after cardiopulmonary resuscitation: a single center study.
J Crit Care, 45 (2018), pp. 40-44
A. Jerath, J. Panckhurst, M. Parotto, N. Lightfoot, M. Wasowicz, N.D. Ferguson, et al.
Safety and efficacy of volatile anesthetic agents compared with standard intravenous midazolam/propofol sedation in ventilated critical care patients: a meta-analysis and systematic review of prospective trials.
Anesth Analg, 124 (2017), pp. 1190-1199
S.K. Dharmalingam, G.J. Amirtharaj, A. Ramachandran, M. Korula.
Volatile anesthetic preconditioning modulates oxidative stress and nitric oxide in patients undergoing coronary artery bypass grafting.
Ann Card Anaesth, 24 (2021), pp. 319-326
S.G. De Hert, P.W. ten Broecke, E. Mertens, E.W. Van Sommeren, I.G. De Blier, B.A. Stockman, et al.
Sevoflurane but not propofol preserves myocardial function in coronary surgery patients.
Anesthesiology, 97 (2002), pp. 42-49
G. Landoni, G.G. Biondi-Zoccai, A. Zangrillo.
Desflurane and sevoflurane in cardiac surgery: a meta-analysis of randomized clinical trials.
J Cardiothorac Vasc Anesth, 21 (2007), pp. 502-511
J. Spence, E. Belley-Côté, H.K. Ma, S. Donald, J. Centofanti, S. Hussain, et al.
Efficacy and safety of inhaled anaesthetic for postoperative sedation during mechanical ventilation in adult cardiac surgery patients: a systematic review and meta-analysis.
BJA, 118 (2017), pp. 658-669
Copyright © 2022. Elsevier España, S.L.U. y SEMICYUC
Medicina Intensiva
Opciones de artículo
Material suplementario
es en

¿Es usted profesional sanitario apto para prescribir o dispensar medicamentos?

Are you a health professional able to prescribe or dispense drugs?