Inhalational Versus Propofol-based Intravenous Maintenance of Anesthesia for Emergence Delirium in Adults: A Meta-analysis and Trial Sequential Analysis

Abstract

Emergence delirium (ED) is a severe postoperative complication that increases the risk for injury, self-extubation, and hemorrhage. Inhalational maintenance of anesthesia is a risk factor for ED in pediatric patients, but its impact in adults is undefined. This meta-analysis compares the incidence of ED between inhalational and propofol-based intravenous maintenance of anesthesia. Following a search of the PubMed, Medline, Embase, and Cochrane Library databases, 12 high-quality randomized controlled trials including 1440 patients, were identified for inclusion in the meta-analysis. Compared with propofol-based intravenous maintenance of anesthesia, inhalational maintenance increased the incidence of ED in adults (risk ratio [RR], 2.02; 95% confidence interval [CI]: 1.30-3.14; P=0.002). This was confirmed by sensitivity analysis, trial sequential analysis, and subgroup analyses of studies that assessed ED via Aono’s four-point scale (RR, 3.72; 95% CI: 1.48-9.31; P=0.005) and the Ricker Sedation Agitation Scale (RR, 3.48; 95% CI: 1.66-7.32; P=0.001), studies that included sevoflurane for maintenance of anesthesia (RR, 1.87; 95% CI: 1.13-3.09; P=0.02), studies that reported ED as the primary outcome (RR, 2.73; 95% CI: 1.53-4.86; P=0.0007), and studies that investigated ocular (RR, 2.98; 95% CI: 1.10-8.10; P=0.03), nasal (RR; 95% CI: 1.27-6.50; P=0.01), and abdominal (RR, 3.25; 95% CI: 1.12-9.40; P=0.03) surgeries, but not intracranial surgery (RR, 0.72; 95% CI: 0.34-1.54; P=0.40). In summary, inhalational maintenance of sevoflurane was a risk factor for ED compared with propofol-based intravenous maintenance in adults who underwent ocular, nasal, and abdominal surgeries but not intracranial surgery.

Emergence from anesthesia, which is defined as the transition from unconsciousness to full wakefulness, is normally smooth and uneventful.¹ Emergence delirium (ED) is characterized by agitation, restlessness, and hyperactivity; therefore, it is also called emergence agitation or emergence excitement and represents a type of inadequate emergence.² ED is an adverse postoperative complication that can occur in patients of any age. It most often occurs in children, with a reported incidence of 10% to 80%;³ in the adult population, the incidence of ED has been reported to lie between 4.7% and 22.2%.⁴ ED increases the risk for injury, self-extubation, hemorrhage, and prolonged hospitalization.⁵

Inhalational anesthesia and propofol-based intravenous anesthesia are 2 common strategies for the maintenance of general anesthesia. The comparative safety of these 2 strategies has been investigated by anesthesiologists for many years. Many studies and the European Society of Anesthesiology evidence-based and consensus-based guidelines on postoperative delirium 2017 have identified inhalational maintenance of anesthesia as a risk factor for pediatric ED.^3,6–9 One meta-analysis published in the Cochrane Library in 2018¹⁰ found no difference between propofol-based intravenous and inhalational maintenance of anesthesia in the incidence of postoperative delirium, which is typically considered as delirium that occurs over 24 hours after general anesthesia among elderly people undergoing noncardiac surgery. Whether inhalational maintenance of anesthesia is a risk factor for adult ED is unclear. Therefore, the aim of this meta-analysis was to compare the incidence of ED between inhalational and propofol-based intravenous maintenance of anesthesia in adults undergoing surgery. Secondarily, we compared the incidence of ED between anesthetic approaches in various subgroups, that is, different ED screening scales, different inhalational anesthetics, and whether ED was assessed as a primary or secondary outcome.

METHODS

This meta-analysis was conducted following recommendations of the Cochrane Handbook for Systematic Reviews of Interventions¹¹ and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA).¹² The protocol was registered with the International Platform of Registered Systematic Review and Meta-Analysis Protocols (INPLASY202070089, https://inplasy.com/inplasy-2020-7-0089/).

Search Strategy and Inclusion Criteria

The PubMed, Medline, Embase, and Cochrane Library databases were comprehensively searched from inception to June 25, 2021, by 2 reviewers (Y.Y. and L.F.) independently, according to the search strategy outlined in the Supplementary Material (Supplemental Digital Content 1: Search strategy, https://links.lww.com/JNA/A477) and without restrictions on language or publication date. Any potentially relevant trials were manually searched based on the references of the identified trials and systematic reviews. The search strategy results are shown in Figure 1.

FIGURE 1: Search results and selection procedure.

Trials were assessed independently by 2 reviewers (Y.Y. and L.F.) and included in the meta-analysis if they met the following criteria: (1) Population—adult surgical patients aged 18 years and above for whom extubation and emergence after propofol-based intravenous or inhalational maintenance of anesthesia was intended in the operation room, postanesthesia care unit, or intensive care unit; (2) Intervention—inhalational maintenance of anesthesia; (3) Comparator—propofol-based maintenance of anesthesia; (4) Outcome—directly reported number of patients who experienced ED during the period from unconsciousness to full wakefulness that could be used to calculate the incidence of ED; (5) Study design—randomized controlled trial (RCT). In cases of disagreement, a consensus was reached through discussion with a third reviewer (B.C.).

Exclusion criteria were: (1) patients with preoperative cognitive disorder, mental disorder, or neurological disease potentially associated with symptoms of delirium; (2) unarousable state during the first 24 hours after surgery; (3) patients who underwent postoperative sedation.

Data Extraction

A standard data extraction form was used to retrieve relevant data independently by 2 reviewers (Y.Y. and L.F.) who were blinded to the authors and institution of the studies. Discrepancies were identified and resolved through discussion with a third reviewer (B.C.). The extracted data included first author, country, study design, sample size, publication date, age, sex, American Society of Anesthesiology physical status score, type of anesthesia and surgery, inclusion and exclusion criteria for patients, screening scale and diagnostic criteria, and incidence of ED (the primary outcome of this study).

Assessment of Methodological Quality

Two reviewers (Y.Y. and L.F.) assessed the quality of RCTs independently based on the guidelines in the Cochrane Handbook for Systematic Reviews of Interventions to create a “risk of bias” table that included the following contents: details of methods of random sequence generation, allocation concealment, blinding, incomplete outcome data, selective outcome reporting, and other biases (commercial sponsorship). The overall quality of each study was evaluated as “low,” “high,” or “unclear” risk of bias. The quality of RCTs was further assessed quantitatively by a modified Jadad 7-point scale; Jadad score ≥4 was considered to be of high quality.¹³ When the number of included studies was ≥10, publication bias was assessed by funnel plot.¹⁴ Furthermore, publication bias was evaluated by Egger test regression.

Statistical Analysis

The meta-analysis was conducted using Review Manager software 5.4 (Cochrane, UK) and Stata 16 (StataCorp LLC, College Station, TX). Risk ratios (RRs) with 95% confidence intervals (CIs) were used to assess the incidence of ED. Statistical heterogeneity was assessed by P and I². All of the meta-analyses were performed by a random-effects model. The Mantel-Haenszel method was used to combine separate statistics. A P-value <0.05 was considered statistically significant. Subgroup analysis was conducted according to the existing features in studies: (1) quality of studies; (2) different types of surgery; (3) different screening scales for assessing ED; (4) different inhalational anesthetics (sevoflurane, enflurane, isoflurane, desflurane, halothane), and; (5) studies reporting ED as primary or secondary outcomes. Sensitivity analysis was conducted by omitting one study in turn.

Trial sequential analysis software 0.9.5.10 (Copenhagen Trial Unit, Denmark) was used to examine the reliability and conclusiveness of the available evidence according to a previous meta-analysis.^15,16 When the cumulative Z curve crossed the trial sequential analysis boundary, a sufficient level of evidence for the anticipated intervention effect was reached, and no further studies were needed. Otherwise, when the Z curve failed to cross the trial sequential analysis boundary, and the required information size was not reached, the evidence to reach a conclusion was insufficient. Two-sided tests with a type I error of 5%, a power of 80%, a low bias-based relative risk reduction, and an ED incidence of 5% in control were used to calculate the required information size.⁴

RESULTS

A total of 1333 studies were identified; among them, 1312 studies were excluded following a review of the title and abstract. After a full-text review of the remaining 21 articles, 9 were excluded for the reasons shown in Figure 1. Two discrepancies occurred between the 2 initial reviewers (Y.Y. and L.F.). First, according to our inclusion criteria, all surgery types should be included, but one reviewer recommended that 3 studies^17–19 investigating neurosurgery should be excluded because between patients undergoing neurosurgery are different from the general surgical population in ways which could affect the incidence of ED. This disagreement was resolved by the third reviewer (B.C.) who recommended including these 3 studies and using sensitivity analysis and subgroup analysis to assess their effects. Second, one reviewer recommended that 3 studies^17,19,20 should be excluded because they did not mention the exclusion of “patients with preoperative cognitive disorder, mental disorder or neurological disease potentially associated with symptoms of delirium.” However, after a full-text review, no study reported patients meeting our exclusion criteria. Thus, the third reviewer (B.C.) again recommended including these studies and using sensitivity analysis to assess their effects. Finally, 12 RCTs including 1440 patients were included in the meta-analysis.^17–28 The characteristics and interventions of the included studies are summarized in Table 1.

TABLE 1 - Characteristics of Included Studies

		Inhalational Group/Intravenous Group
References	Methods	Cases (Male %)	Age (Mean±SD)	Patients	Surgery	Intravenous Maintenance of Anesthesia	Inhalational Maintenance of Anesthesia	ED Assessment	ED Was the Primary Outcome
Bastola et al17	RCT, parallel design, single-center	50/25 (70/72)	41±13/38±13	Inclusion criteria: 20-60 y old, ASA I-II, Glasgow Coma Score of 15 Exclusion criteria: with ischemic and/or congestive heart disease, hypertension, diabetes mellitus, chronic obstructive pulmonary disease, hepatic and renal dysfunction	Elective craniotomy for supratentorial tumors	Propofol 5-10 mg/kg/h+60% N2O	Sevoflurane 1%-2% (or desflurane 2%-4%)+60% N2O	Assessed during the emergence period, defined as require verbal, pharmacological or physical restraint	No
Citerio et al18	RCT, parallel design, multicenter	411 (51)	55±13	Inclusion criteria: 18-75 y old, Glasgow Coma Score of 15, no clinical signs of intracranial hypertension Exclusion criteria: severe cardiovascular, renal or liver disease, pregnancy, allergies to any anesthetic agent, body weight >120 kg, drug abuse or psychiatric conditions and documented disturbance, surgery of the hypothalamic region	Elective supratentorial intracranial surgery	Propofol infused 6-10 mg/kg/h	0.75-1.25 MAC sevoflurane	Assessed during the emergence period, the definition of ED was not mentioned	No
Jellish et al21	RCT, parallel design, multicenter	93/93 (40/43)	44±13/42±14	Inclusion criteria: >18 y old, ASA I-II Exclusion criteria: a history of neurological disease, malignant hyperthermia, preexisting renal insufficiency, general anesthesia within the last 2 wk, a positive pregnancy test or were breast feeding	Elective surgical procedures up to 3 h	Propofol along with 67% N2O at 3 L/min	Sevoflurane with 67% N2O at 3 L/min	Assessed on admission to the PACU and repeated every 10 min, scored between 0 and 2	No
Jo et al22	RCT, parallel design, single-center	40/40 (60/77)	40±17/43±19	Inclusion criteria: 20-79 y old, ASA I-III Exclusion criteria: a diagnosis of mental illness, taken a neuroleptic or benzodiazepine for >2 wk within 1 mo of surgery, allergic to opioids or nonsteroidal anti-inflammatory drugs, diagnosed with Asthma or nasal polyps	Open rhinoplasty, septoplasty, turbinoplasty, endoscopic sinus surgery	TCI of propofol at 2.0-2.5 μg/mL, BIS maintained 40-60	Sevoflurane and 50% N2O in O2	Assessed immediately after extubation, defined as RSAS≥5 or RASS≥1	Yes
Kim et al23	RCT, parallel design, single-center	20/20 (80/75)	33±3/35±2	Inclusion criteria: 20-60 y old, ASA I-II Exclusion criteria: with signs or symptoms of infection in the upper respiratory tract, history of sleep apnea, administered medication for psychiatric diseases	Elective for closed reduction of nasal bone fracture	Propofol 200 μg/kg/min, BIS maintained 40-60	4% sevoflurane with 50% N2O at 2 L/min, BIS maintained 40-60	Assessed after the patients had a response to language stimulation in PACU, defined as AFPS≥3	Yes
Liu et al24	RCT, parallel design, single-center	20/20 (20/25)	50±2/51±2	Inclusion criteria: 25-65 y old, ASA I-II Exclusion criteria: with signs or symptoms of heart diseases, open fracture of wrist, local infection of wrist, administered medication for psychiatric diseases	Elective for closed reduction of distal radius fracture	Propofol 200 μg/kg/min, BIS maintained 40-60	0.5%-1.5% isoflurane with 50% N2O, BIS maintained 40-60	Assessed when the patients were able to describe their feelings in PACU, defined as AFPS≥3	Yes
Na et al20	RCT, parallel design, single-center	42/41 (45/42)	60±5/59±3	Inclusion criteria: 20-80 y old, ASA I-III Exclusion criteria: allergy to the anesthetic agents, anticipated difficult airway, BMI >30, chronic obstructive pulmonary disease, heart failure, unstable angina	Elective vitrectomy	TCI of propofol, BIS maintained 40-60	Desflurane inhalation, BIS maintained 40-60	Assessed in the PACU, defined as RSAS ≥5	No
Necib et al19	RCT, parallel design, single-center	35/31 (46/55)	50±11/52±11	Inclusion criteria: 18-75 y old, ASA I-II Exclusion criteria: frontal tumors, blindness and motor compromise of the upper limbs, absence of French speaking	Elective craniotomy for supratentorial tumors	TCI of propofol at 3 μg/mL and remifentanil at 4 ng/mL, BIS maintained 45-55	Sevoflurane and boluses of 0.1 μg/kg sufentanil to maintain BIS 45-55	Assessed during the emergence period, defined as AFPS≥3	No
Nishikawa et al25	RCT, parallel design, single-center	25/25 (48/52)	71±7/71±8	Inclusion criteria: >18 y old, ASA I-II Exclusion criteria: with anticoagulation, symptomatic coronary artery disease, cardiac valvular regurgitation or stenosis, central nervous system or neuromuscular disorders, major or minor tranquilizer medication, psychotic symptoms or cognitive impairment as judged by a psychiatrist	Elective laparoscope-assisted surgical procedures >3 h	TCI of propofol at 4 μg/mL	5% sevoflurane and 100% oxygen at 3 L/min	Assessed at 30 s intervals during the emergence period, excitement was recorded	No
Talih et al26	RCT, parallel design, single-center	45/45 (44/60)	29±8/29±8	Inclusion criteria: >18 y of age, ASA I-II Exclusion criteria: bleeding diathesis, using anticoagulant, hypnotic-antipsychotic-antidepressant drugs or substance addicts, consumed alcohol, had rhinoplasty surgery	Rhinoplasty	TCI of propofol, PSI maintained 25-50	1-1.1 MAC sevoflurane at 1 L/min	Assessed immediately after extubation, defined as RASS≥1	Yes
Tawfik et al27	RCT, parallel design, multicenter	128/124 (46/41)	43±17/46±15	Inclusion criteria: 18-75 y old, ASA I-II Exclusion criteria: pregnancy, dementia, known psychiatric disorders, hepatic or pancreatic insufficiency, a history of habitual drug or alcohol abuse, underwent any surgical procedure under local anesthesia in the past 3 y, a known allergy, sensitivity to volatile anesthetics or to propofol, patients undergoing any bilateral procedures	Elective oculoplastic procedures	Propofol 0.5 mg/kg premixed with lidocaine for sedation	Sevoflurane 8% mixed with oxygen at 6 L/min for sedation	Assessed during the emergence period by a simplified sedation score	No
Zhang et al28	RCT, parallel design, single-center	40/40 (75/75)	69±7/70±6	Inclusion criteria: 65-78 y old, ASA I-II Exclusion criteria: mental disorder, severe dysfunction of heart, lung, liver kidney, spinal deformity, contraindications of epidural anesthesia, a history of severe trauma or surgical treatment	Selective radical surgery for gastric cancer	TCI of propofol at 1.5-3.0 μg/mL	1.5%-3.5% sevoflurane	Assessed during the emergence period, defined as RSAS≥5	No

AFPS indicates Aono’s four-point scale; ASA, American Society of Anesthesiologists; BIS, bispectral index; BMI, body mass index; EA, emergence agitation; MAC, minimal alveolar concentration; PACU, postanesthetic care unit; PSI, patient state index; RASS, Richmond Agitation Sedation Scale; RCT, randomized controlled trial; RSAS, Ricker Sedation Agitation Scale; TCI, target-controlled infusion.

Risk of Bias Assessment

The quality of the studies was assessed by a modified Jadad score and the risk of bias (Table 2); RCTs with a modified Jadad score ≥4 were considered high quality. All included RCTs reported use of a computer or random number generator for randomization, and 5^{17–19,21,25} reported allocation concealment via a sealed envelope. All RCTs presented complete data, with no selective reporting or commercial sponsorship. Due to the characteristic of the intervention, blinding of the anesthesiologist was impossible, but blinding of outcomes was reported in 9 of the RCTs.^18–25,27 Thus, these 9 studies^18–25,27 were regarded as having a low risk of bias, and the remaining 3 studies^17,26,28 were considered to have a moderate risk of bias. The nearly symmetrical funnel plot indicated that a low risk of publication bias existed among the included studies (Fig. 2). Moreover, the Egger test indicated that there was no significant publication bias in the meta-analysis (P=0.303).

TABLE 2 - Quality Assessment of Randomized Controlled Trials Based on the Guidelines in the Cochrane Handbook for Systematic Reviews of Interventions and a Modified Jadad 7-point Scale

References	Sequence Generation	Allocation Concealment	Blinding of Participants and Personnel	Blinding of Outcome Assessment	Incomplete Outcome Data	No Selective Outcome Reporting	Commercial Sponsorship	Jadad Score	Quality
Bastola et al17	√	√	–	?	√	√	√	6	Good/moderate risk of bias
Citerio et al18	√	√	–	√	√	√	√	6	Good/low risk of bias
Jellish et al21	√	√	–	√	√	√	√	6	Good/low risk of bias
Jo et al22	√	?	–	√	√	√	√	5	Good/low risk of bias
Kim et al23	√	?	–	√	√	√	√	5	Good/low risk of bias
Liu et al24	√	?	–	√	√	√	√	5	Good/low risk of bias
Na et al20	√	?	–	√	√	√	√	5	Good/low risk of bias
Necib et al19	√	√	–	√	√	√	√	6	Good/low risk of bias
Nishikawa et al25	√	√	–	√	√	√	√	6	Good/low risk of bias
Talih et al26	√	?	–	?	√	√	√	4	Good/moderate risk of bias
Tawfik et al27	√	–	–	√	√	√	√	4	Good/low risk of bias
Zhang et al28	√	?	–	?	√	√	√	4	Good/moderate risk of bias

√ indicates low risk of bias; ?, unclear risk of bias; –, high risk of bias.

FIGURE 2: Funnel plot of the meta-analysis of the incidence of emergence delirium. RR indicates risk ratio.

Meta-analysis

Twelve high-quality studies with 1440 patients were included in the meta-analysis (Fig. 3). The incidence of ED after inhalational maintenance of anesthesia was significantly higher than that after propofol-based intravenous maintenance (RR, 2.02; 95% CI: 1.30-3.14; P=0.002); no significant heterogeneity was observed (I²=25%). This significant difference remained when omitting one study in turn for sensitivity analysis (data not shown). Furthermore, subgroup analysis revealed that the incidence of ED after inhalational maintenance of anesthesia was significantly higher than that after propofol-based intravenous maintenance regardless of whether the enrolled studies had a low risk of bias (RR, 2.17; 95% CI: 1.16-4.06; P=0.02) or moderate risk of bias (RR, 2.01; 95% CI: 1.18-3.41; P=0.01) (Supplemental Digital Content 2: Forest plots for subgroup analysis of the incidence of ED based on study quality, https://links.lww.com/JNA/A478). Moreover, trial sequential analysis of the incidence of ED among all included studies showed that the Z curve crossed the conventional boundary, trial sequential analysis boundary, and required information size (Fig. 4).

FIGURE 3: Forest plots showing the meta-analysis of the incidence of emergence delirium following inhalational and intravenous maintenance of anesthesia. CI indicates confidence interval; M-H, Mantel-Haenszel.

FIGURE 4: Trial sequential analysis of the incidence of emergence delirium. RIS indicates required information size.

Three studies investigated elective intracranial surgery,^17–19 2 studies investigated elective ocular surgery,^20,27 3 studies investigated nasal surgery,^22,23,26 2 RCTs assessed abdominal surgery,^25,28 1 RCT assessed closed reduction of distal radius fracture,²⁴ and 1 RCT did not mention the site of operation.²¹ Except for intracranial surgery (RR, 0.72; 95% CI: 0.34-1.54; P=0.40), the meta-analysis found that the incidence of ED after inhalational maintenance of anesthesia was significantly higher than that after propofol-based intravenous maintenance in ocular surgery (RR, 2.98; 95% CI: 1.10-8.10; P=0.03), nasal surgery (RR, 2.87; 95% CI: 1.27-6.50; P=0.01), and abdominal surgery (RR, 3.25; 95% CI: 1.12-9.40; P=0.03); moreover, no significant heterogeneity was detected for subgroup analysis of intracranial (I²=0%), ocular (I²=0%), nasal (I²=31%), or abdominal (I²=0%) surgeries (Supplemental Digital Content 3: Forest plots for subgroup analysis of the incidence of ED based on types of surgery, https://links.lww.com/JNA/A479). These results indicate that inhalational maintenance of anesthesia is not a risk factor for ED after intracranial surgery but is a risk factor for ED after ocular, nasal, and abdominal surgery.

ED was assessed using Aono’s four-point scale in 3 studies,^19,23,24 using the Ricker Sedation Agitation Scale in 3 studies,^20,22,28 and using the Richmond Agitation Sedation Scale in 2 studies;^22,26 the remaining studies used different methods to assess ED. The meta-analysis found that the incidence of ED after inhalational maintenance of anesthesia was significantly higher than that after propofol-based intravenous maintenance when assessed by Aono’s four-point scale (RR, 3.72; 95% CI: 1.48-9.31; P=0.005) and the Ricker Sedation Agitation Scale (RR, 3.48; 95% CI: 1.66-7.32; P=0.001) but not when assessed by the Richmond Agitation Sedation Scale (RR, 2.86; 95% CI: 0.76-10.73; P=0.12) (Supplemental Digital Content 4: Forest plots for subgroup analysis of the incidence of ED based on ED screening scales, https://links.lww.com/JNA/A480). No heterogeneity existed among the studies that assessed ED with the Aono’s four-point scale (I²=0%) or the Ricker Sedation Agitation Scale (I²=0%), but there was heterogeneity among the studies that assessed ED with the Richmond Agitation Sedation Scale (I²=48%) (Supplemental Digital Content 4, https://links.lww.com/JNA/A480).

Ten studies^{17–19,21–23,25–28} compared sevoflurane-based maintenance with propofol-based intravenous maintenance of anesthesia, 2 studies^17,20 compared desflurane-based maintenance with propofol-based intravenous maintenance, and one study²⁴ compared isoflurane-based maintenance with propofol-based intravenous maintenance. The meta-analysis found that sevoflurane-based maintenance of anesthesia was associated with a higher incidence of ED than propofol-based intravenous maintenance (RR, 1.87; 95% CI: 1.13-3.09; P=0.02) with no significant heterogeneity (I²=27%); however, there was no difference in the incidence of ED between desflurane-based maintenance of anesthesia and propofol-based intravenous maintenance (RR, 1.47; 95% CI: 0.21-10.30; P=0.70) (Supplemental Digital Content 5: Forest plots for subgroup analysis of the incidence of ED based on inhalational anesthetics, https://links.lww.com/JNA/A481).

Four studies^22–24,26 reported ED as the primary outcome, and 8 studies^{17–21,25,27,28} reported ED as the secondary outcome. The meta-analysis found that inhalational maintenance of anesthesia was associated with a higher incidence of ED than propofol-based intravenous maintenance (RR, 2.73; 95% CI: 1.53-4.86; P=0.0007), with no significant heterogeneity (I²=11%) when ED was reported as the primary outcome; however, no difference was detected when ED was reported as the secondary outcome (RR, 1.54; 95% CI: 0.86-2.76; P=0.15) (Supplemental Digital Content 6: Forest plots for subgroup analysis of the incidence of ED based on the primary or secondary outcome, https://links.lww.com/JNA/A482). When the 3 RCTs^17–19 that assessed intracranial surgery were omitted, inhalational maintenance of anesthesia was associated with a higher incidence of ED than propofol-based intravenous anesthesia maintenance (RR, 2.30; 95% CI: 1.23-4.28; P=0.009), with no heterogeneity (I²=0%) when ED was reported as the secondary outcome (Supplemental Digital Content 7: Forest plots for subgroup analysis of the incidence of ED based on the primary or secondary outcome when the studies investigating intracranial surgery were omitted, https://links.lww.com/JNA/A483).

DISCUSSION

This study aimed to investigate whether inhalational maintenance of anesthesia increases the incidence of ED in adults as compared with propofol-based intravenous maintenance. Pooled data from our meta-analysis involving 12 high-quality RCTs of 1440 adult patients found that inhalational maintenance of anesthesia was associated with a significantly higher incidence of ED than propofol-based intravenous maintenance. This finding was further confirmed by sensitivity analysis and subgroup analyses of studies with a low and moderate risk of bias, studies that assessed ED by Aono’s four-point scale or the Ricker Sedation Agitation Scale, studies that used sevoflurane for the maintenance of anesthesia, studies that reported ED as the primary outcome, and studies that investigated ocular, nasal, and abdominal surgeries but not intracranial surgery. Surprisingly, when studies that investigated intracranial surgery were excluded, inhalational maintenance of anesthesia had a significantly higher incidence of ED than propofol-based intravenous maintenance when ED was reported as a secondary outcome. Moreover, the trial sequential analysis indicated that the current evidence is sufficient to support the higher incidence of ED in adults after inhalational maintenance of anesthesia than that after propofol-based intravenous anesthesia maintenance during ocular, nasal, abdominal, and intracranial surgeries, and that no further studies are needed. Taken together, our data indicate that sevoflurane-based inhalational maintenance of anesthesia is a risk factor for ED in adults after ocular, nasal, and abdominal surgeries.

The European Society of Anesthesiology evidence-based and consensus-based guidelines on postoperative delirium 2017 suggest that abdominal and cardiothoracic surgeries are risk factors for postoperative delirium.⁹ Similarly, this study found that abdominal surgery was a risk factor for ED, but the effect of cardiothoracic surgery is unclear due to the lack of RCTs. Moreover, although this study found that intracranial surgery was not a risk factor for ED, only 3 studies were included, and there are many different types of neurosurgery and aspects of anesthesia management that can affect the occurrence of ED. Therefore, more high-quality RCTs are needed to investigate the effects of cardiothoracic and neurosurgical surgery on ED.

This meta-analysis found no difference between desflurane-based anesthesia maintenance and propofol-based intravenous anesthesia maintenance. This finding was consistent with those of a prospective randomized double-blind study of 144 adults that reported that desflurane anesthesia was associated with less ED than sevoflurane anesthesia undergoing orthognathic surgery²⁹ but was inconsistent with a meta-analysis of 1196 children that reported a comparable incidence of ED between desflurane and sevoflurane anesthesia.³⁰ There are 2 possible explanations for this discrepancy. First, there were only 2 studies of 294 patients receiving desflurane maintenance in our meta-analysis, and, second, the effect of desflurane anesthesia is different between adults and children. Taken together, this indicates that more high-quality RCTs are required to confirm this finding.

Currently, no standard scale is able to diagnose delirium accurately in any setting. This meta-analysis included at least 7 scales to assess ED, of which Aono’s four-point scale, the Ricker Sedation Agitation Scale, and the Richmond Agitation Sedation Scale were most frequently used. However, the difference in the incidence of ED between inhalational maintenance and propofol-based intravenous maintenance of anesthesia was only detected when ED was assessed by Aono’s four-point scale or the Ricker Sedation Agitation Scale, and not when it was assessed by the Richmond Agitation Sedation Scale. The reason that the Richmond Agitation Sedation Scale identified no difference may be that only 2 such studies were included, one of which investigated intracranial surgery. Thus, this finding needs further investigation in high-quality studies.

The implications of this meta-analysis for clinical practice are that inhalational anesthesia maintenance with sevoflurane can increase the incidence of ED in adults, similar to that in children. Propofol-based intravenous anesthesia could be suitable for patients with risk factors for ED. A prophylactic dose of propofol at the end of inhalational anesthetic regimens,³¹ or treatment with dexmedetomidine or fentanyl,⁸ which has been verified to decrease the incidence of ED in children, should be investigated in adults.

There are several limitations to this meta-analysis. First, various types of surgeries were included in the meta-analysis which increased the heterogeneity of the study population. Second, due to the limited sample size, the incidence of ED in the propofol-based intravenous maintenance and desflurane-based maintenance, and subgroup analysis of the Richmond Agitation Sedation Scale, are inconclusive. Third, the effects of sex and gender on the incidence of ED could not be determined. Finally, a broad range of ages (18 y and above) were included in the meta-analysis, but the associated subgroup analysis could not be performed because of insufficient data; this may increase confounding bias for the meta-analysis.

CONCLUSIONS

Based on this meta-analysis of 12 high-quality RCTs including 1440 patients, compared with propofol-based intravenous maintenance of anesthesia, sevoflurane-based inhalational maintenance is a risk factor for ED in adults undergoing ocular, nasal, and abdominal surgeries. Anesthesiologists should consider the risk of ED when using sevoflurane for general anesthesia and consider alternative measures to prevent ED.