The laryngeal mask airway (LMA) is a supraglottic airway device that can be inserted after induction of general anesthesia (GA) using either inhalational or intravenous (IV) agents. Propofol, a short-acting lipophilic agent, is the most widely used IV agent for GA induction because it is able to induce relaxation of the masticatory musculature and depression of the upper airway reflexes, and thus allow LMA insertion. 1,2 When used as the sole anesthetic, high-dose propofol can lead to profound hypotension, cardiorespiratory depression, and prolonged apnea. 3,4 To minimize these undesirable effects, other anesthetic agents such as opioids, benzodiazepines, inhaled anesthetics, and neuromuscular blocking agents have been used to reduce the dose of propofol required. Each anesthetic combination confers different benefits and risks that may improve or impair LMA placement. 5 Despite the use of LMAs for over 4 decades, there is no consensus on which anesthetic combination facilitates the greatest LMA insertion success while minimizing adverse events.
The aim of this study was to examine differences in adverse outcomes between anesthetic agents used during LMA insertion following GA induction. We performed a systematic review and network meta-analysis (NMA) comparing propofol as the sole anesthetic for GA induction with other anesthetic combinations used during LMA insertion. Our primary outcome was prolonged apnea, and our secondary outcomes were airway adverse events, LMA insertion failure, inadequate anesthetic depth, adverse hemodynamic events, unanticipated hospital admission, and patient mortality.
Anesthesia practice among clinicians is diverse in their selection of drug combinations and doses, which is challenging when creating evidence-based statements on airway management. The NMA creates a network of interventions with both direct and indirect comparisons. 6 By combining direct and indirect estimates across a network of interventions in a single analysis, a relative ranking of treatments for a given outcome can be made. This study uniquely employs an NMA to compare the variety of induction agents used during LMA insertion to establish which may reduce risks of prolonged apnea and other adverse outcomes the most.
METHODS
This study was registered at PROSPERO (CRD42020202474). The Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension statement for NMA was used to report our methodology and results. 6 Eligibility criteria included the following:
-
Population: American Society of Anesthesiologists I or II adult patients (≥15 years of age) receiving GA with an LMA for nonemergency medical or dental surgery in hospital or ambulatory clinical setting who are spontaneously breathing during the procedure
-
Interventions: Any anesthetic agent(s) for induction of GA and LMA insertion
-
Comparison: Propofol as sole anesthetic for GA induction and LMA insertion
-
Outcomes:
-
Primary outcome: prolonged apnea, defined as duration (time from induction until first breath) or incidence (≥15 s without breath after LMA insertion) 7,8
-
-
Airway adverse event
-
Incidence of coughing or gagging
-
Incidence of breath holding
-
Incidence of laryngospasm
-
Incidence of hypoxia (oxygen saturation as measured by pulse oximetry <96%) 11
-
Incidence of swallowing
-
-
Insertion failure of LMA
-
Proportion of LMAs unsuccessful placed on first attempt
-
Inadequate ease of insertion (difficult/impossible)
-
-
Inadequate depth of anesthesia
-
Incidence of additional/rescue anesthetic required
-
Inadequate jaw relaxation/mouth opening (Young’s criteria 12 ; full, partial, nil)
-
Incidence of excitatory (head or limb) or nonpurposeful movements
-
-
Adverse hemodynamic event
-
Bradycardia (<60 beats/min) 13
-
Tachycardia (>100 beats/min) 13
-
Hypotension (systolic arterial blood pressure <90 mm Hg 14 or mean arterial pressure decrease >20% of the baseline value) 15
-
Hypertension (systolic arterial blood pressure >160 mm Hg 16 or mean arterial pressure increase >20% of the baseline value) 16
-
-
Unanticipated hospital admission
-
Mortality
-
-
-
Study design: Randomized controlled trials (RCTs)
Data Sources and Search Strategy
Our search strategy was constructed with the help of 2 medical librarians and was applied to the following electronic databases: Cochrane Library (Cochrane Database of Systematic Reviews, Cochrane Central Register of Controlled Trials [CENTRAL], Cochrane Methodology Register), OVID MEDLINE, OVID EMBASE, CINAHL, and SCOPUS (Supplemental Information, Table S1; all supplemental information available online only). The main databases were initially searched on January 9, 2021, and updated on January 19, 2022. A search filter of 1983, the year of the first LMA publication, to present was used for each database. 17 The first 100 hits of Web of Science and ScienceDirect were also searched for additional relevant studies. Completed and ongoing trials were searched through clinicaltrials.gov and World Health Organization International Clinical Trials Registry Platform. Unpublished studies such as dissertations, reports, and conference abstracts were searched for in ProQuest, Google Scholar (the first 100 hits), and the OpenGrey database. Reference lists of previous reviews in the same topic and included studies were searched manually. There were no language or publication date restrictions. We attempted to contact the authors in case of missing information. Search results were imported into EndNote X9 (Clarivate) for duplication screening, then Covidence systematic review software (Veritas Health Innovation) for study screening and selection. 18,19
Study Selection and Data Collection
Two review authors (C.G. and E.G.) independently reviewed and selected trials from screening. C.G. extracted the data from the selected trials and E.G. reviewed and confirmed all extracted data. Disagreements were resolved through discussion and consensus or by consulting a third reviewer (M.W.). Reason for study exclusion and included study characteristics are found in the Supplemental Information (Tables S2 and S3).
Risk of Bias Assessment
The 2 review authors, C.G. and E.G., independently used Cochrane’s risk of bias assessment tool to evaluate the methodological quality of included trials across 6 domains 20 : selection, performance, detection, attrition, reporting, and other biases. Studies were graded as low risk if all the domains were judged to be of low risk, moderate risk if any domain was judged to be of unclear risk, and high risk if any domain was judged to be of high risk or if all domains were judged to be unclear risk. Disagreements were resolved with discussion or with a third reviewer (M.W.).
NMA and Data Synthesis
The included studies’ data were entered into a standardized spreadsheet. If the study contained an intervention with the same anesthetic but different doses, the intervention arms were combined. Anesthetics were grouped together based on related drug class (ie, pharmacodynamic properties) for analysis. Continuous outcome (apnea time) was analyzed with mean and SD. Binary outcomes (presence of an airway adverse event or inadequate depth of anesthesia) were analyzed using the odds ratio (OR). Ordinal outcomes were combined and analyzed as a binary outcome. Missing SDs were calculated from the available CI or standard error and the number of participants. 21 If mean and SD were unavailable, they were estimated using median and range (the formula presented by Hozo and colleagues 22 using median and range) or using range (the formula presented by Walter and Yao 23 ). If numerical data were not reported, authors were contacted.
NMA was conducted using the program R (version 3.5.0; R Project for Statistical Computing) with the Gemtc version 0.8.2 and rjags packages, which interface with Just Another Gibbs Sampler software (version 4.0.0; developed by Martyn Plummer) for Markov chain Monte Carlo modeling. A Bayesian NMA was performed under a hierarchical random effects framework and unified generalized linear model. For continuous outcomes, a normal likelihood and identity link function was used. For binary outcomes, a binomial likelihood with logit link function was used. 24,25 The treatment effects were estimated as ORs or mean differences with associated 95% credible intervals (CrIs) and the surface under the cumulative ranking curve (SUCRA). Heterogeneity was assessed using the I2 statistic. Node-splitting method was used to calculate the inconsistency of the model through direct and indirect evidence with its Bayesian P value. 24 Posterior probabilities were used and scored to rank the interventions for the overall conclusions. Sensitivity analysis was conducted by excluding any high–risk of bias studies. Subgroup analysis was conducted comparing (1) the effect of premedication and (2) the type of LMA.
Quality of Evidence
Two review authors (C.G. and E.G.) used the CINeMA web application (University of Bern) to evaluate the confidence of the main NMA results considering 6 domains: heterogeneity, imprecision, incoherence, indirectness, and within- and across-studies bias. 26 Results were judged as high, moderate, low, or very low confidence. 26 Consensus was reached by consulting the other reviewer or by third reviewer (M.W.) when necessary.
RESULTS
Search Results
In the initial screening stage, 6762 potentially eligible reports were retrieved, 2966 duplicates were removed, and 3540 records were excluded based on title and abstract (Figure 1). A total of 256 full-text trials were assessed for eligibility; 195 studies were excluded, and 61 RCT met the inclusion criteria (Supplemental Information, Tables S2 and S3). The included studies were from India (18), China (8), Turkey (6), Singapore (4), United Kingdom (4), Malaysia (3), South Korea (3), Brazil (2), Iran (2), Taiwan (2), Australia, Canada, Egypt, Ireland, Japan, Lebanon, Poland, Saudi Arabia, and Singapore (Supplemental Information, Table S3).
Citation: Anesthesia Progress 71, 2; 10.2344/22-00033
Study Characteristics
Population.
All studies limited their patients to adults, although the minimal age of inclusion ranged from 15 years to 18 years of age (≥15 y, 3 studies; ≥16 y, 4 studies; ≥17 y, 1 study; and ≥18 y, 44 studies; Supplemental Information, Table S3). Nine studies did not specify their adult age range of inclusion. LMA types used were the first-generation LMA Classic (Teleflex; 14 studies) and second-generation ProSeal LMA (Teleflex; 5 studies) and I-gel (Intersurgical; 1 study). Forty-one studies did not specify LMA type used; however, 16 of these studies were conducted before the introduction of second-generation LMAs 27 in 2000. Twenty-nine studies did not use premedication prior to anesthetic induction. Of the 24 studies that reported premedication use, 12 used an anticholinergic (glycopyrrolate or atropine), 10 used an oral benzodiazepine (midazolam, lorazepam, or temazepam), and 2 used hydroxyzine. Eight studies did not specify if premedication was utilized. The surgical reasons for GA and LMA placement were elective general surgery, orthopedics, and plastic or gynecologic procedures (17 studies); lower abdominal or urologic operations (7 studies); oncological (4 studies); minor, elective, or short procedures (22 studies); and not specified (11 studies).
Interventions.
A total of 48 different anesthetics and anesthetic combinations were identified. Of the 61 studies, 5 had intervention pairings that were disconnected from the comparator network across all outcomes, and were excluded from analysis. An additional 3 trials were excluded from analysis as they became single-arm trials following the anesthetic drug class grouping. In total, 4695 patients and 37 anesthetic treatments were combined into 28 anesthetic drug class groupings for inclusion in the NMA (Table).
Outcome Assessment.
Figure 2 demonstrates the networks of each outcome. As reporting of outcomes differed across studies, analysis was completed for 31 studies for apnea incidence (2548 patients, 20 anesthetic combinations), 29 studies for apnea time (3091 patients, 17 anesthetic combinations), 41 studies for airway adverse events (3679 patients, 26 anesthetic combinations), 41 studies for LMA insertion failure (3853 patients, 26 anesthetic combinations), and 36 studies for inadequate depth of anesthesia (3888 patients, 20 anesthetic combinations). Supplemental Information, Figure S1 summarizes relative treatment effects for all possible comparisons as expressed by ORs with 95% CrIs.
Citation: Anesthesia Progress 71, 2; 10.2344/22-00033
In total, there was a 33% (849 of 2548) incidence of prolonged apnea overall (Table). For prolonged apnea, vapor + midazolam + short opioid demonstrated a high SUCRA score (OR, 0.02; 95% CrI, 4.0 × 10−3 to 0.64; SUCRA = 92%; Figure 3), which indicates a higher likelihood that combination is one of the top ranked for reducing the risk of prolonged apnea. For comparison, propofol alone demonstrated a SUCRA score of 47% in this outcome (Figure 3).
Citation: Anesthesia Progress 71, 2; 10.2344/22-00033
Overall, the mean apnea time was 3.74 ± 3.56 minutes (Table). Propofol + vapor (mean difference = −2.10; 95% CrI, −4.70 to 0.47; SUCRA = 92%) ranked the top combination for reducing the risk of apnea time (Figure 3). For comparison, propofol alone demonstrated a SUCRA score of 71% in this outcome (Figure 3).
In the airway adverse event outcome, coughing and/or gagging was reported the most and had an incidence of 17.8% (582 of 3261; Table). Laryngospasm was examined among 2887 patients and had an incidence of 4.7%. Propofol + midazolam + agonist-antagonist opioid (OR, 5.7 × 10−11; 95% CrI, 2.9 × 10−31 to 0.03; SUCRA = 98%) ranked the top intervention for reducing the risk of airway adverse events (Figure 3). For comparison, propofol alone demonstrated a SUCRA score of 9% in this outcome (Figure 3).
Under the insertion failure outcome, failed LMA insertion on the first attempt occurred in 16.2% of patients (490 of 3028; Table). Furthermore, 16.2% (486 of 2999) were reported to have inadequate ease of LMA insertion. Propofol + midazolam + IV lidocaine (OR, 0.03; 95% CrI, 1.0 × 10−3 to 0.79; SUCRA = 88%) ranked the top intervention to reduce risk of insertion failure (Figure 3). For comparison, propofol alone demonstrated a SUCRA score of 20% in this outcome (Figure 3).
Within the inadequate depth of anesthesia outcome, head and/or limb movement had the highest incidence of occurrence and was reported in 30.4% (811 of 2669) of patients (Table). Propofol + vapor + short opioid (OR, 0.03; 95% CrI, 3.8 × 10−3 to 0.24; SUCRA = 91%) ranked the top intervention to reduce the risk of inadequate depth of anesthesia (Figure 3). For comparison, propofol alone demonstrated a SUCRA score of 9% in this outcome (Figure 3).
Hemodynamic events (bradycardia, tachycardia, and hypotension) were reported in 19 studies, across 1613 patients and 15 anesthetic interventions. The most reported event was hypotension, which occurred in 12.2% (150 of 1232) of patients. However, 75 of 150 hypotension events occurred in the propofol + short opioid intervention. The OR and CrI demonstrated large variability across all interventions following analysis and was excluded in the summed scores of SUCRA ranks (Supplemental Information, Figure S2).
The SUCRA ranks were scored across all outcomes and summed (Supplemental Information, Table S4). Propofol + dexmedetomidine had the highest summed score (Figure 4), as it ranked among the top interventions across the outcomes for apnea incidence (OR, 2.20; 95% CrI, 0.21-39.0; SUCRA = 37%; rank 13 of 20), apnea time (mean difference 0.24; 95% CrI, −2.40 to 2.80; SUCRA = 66%; rank 5 of 17), airway adverse event (OR, 0.07; 95% CrI, 0.02-0.27; SUCRA = 67%; rank 7 of 26), insertion failure (OR, 0.12; 95% CrI, 0.02-0.59; SUCRA = 73%; rank 4 of 26), and inadequate depth of anesthesia (OR, 0.06; 95% CrI, 0.01-0.29; SUCRA = 84%; rank 2 of 20). Eighteen of 28 anesthetic interventions had scores across all outcomes, of which propofol had the lowest summed SUCRA score.
Citation: Anesthesia Progress 71, 2; 10.2344/22-00033
None of the included studies reported hypertension, unanticipated hospital admission, or mortality.
Risk of Bias
The overall risk of bias was graded as low in 19 studies (36%), moderate in 21 studies (40%), and high in 13 studies (24%) (Figure 5). Overall, high risk of bias was found in the domains of selection bias (1 study), performance bias (3 studies), detection bias (4 studies), and attrition bias (7 studies; Figure 5).
Citation: Anesthesia Progress 71, 2; 10.2344/22-00033
Sensitivity and Subgroup Analysis
When the high–risk of bias studies were removed, propofol + dexmedetomidine remained as having the highest overall cumulative SUCRA rank score in reducing adverse outcomes (Supplemental Information, Table S4). The results of the other subgroup analyses, (1) the effect of premedication and (2) the type of LMA (Supplemental Information, Table S4), were similar to those of the primary analysis. The quality of evidence ranged from very low to moderate (Supplemental Information, Table S5). Heterogeneity analysis demonstrated significant inconsistency (I2 = 8.9-99.1%) with low to considerable heterogeneity (Supplemental Information, Tables S6 and S7).
DISCUSSION
This NMA attempted to identify the best anesthetic combination to avoid prolonged apnea and the other secondary outcomes of airway adverse events, inadequate anesthetic depth, LMA placement failure, and adverse hemodynamics after LMA insertion. Anesthesiologists often intend to have brief apnea to achieve sufficient anesthetic depth for LMA placement; however, anesthetic overdose during induction has been frequently identified by bispectral index monitoring. 28,29 Avoiding prolonged apnea and promptly resuming spontaneous breathing are desired; hence, prolonged apnea after LMA insertion was selected as the primary outcome for this study. Return of spontaneous breathing can be evaluated along with other physiology- and medication-related parameters to allow anesthesiologists to titrate their anesthetics by monitoring changes in their patients’ respiration rates and tidal volumes, thereby decreasing risk of anesthetic overdose.
Preventing overdose may also avoid adverse outcomes such as hemodynamic instability, prolonged recovery time, and increased mortality. 29,30 In our study, it was difficult to determine if prolonged apnea was associated with other adverse outcomes such as hypoxia, as there was a lack of reporting in the included studies. Of the 53 studies included for meta-analysis, only 7 reported hypoxia as an outcome. Lack of reporting of hemodynamic events was also seen across the studies. In contrast, by reducing apnea time, the risk of inadequate depth of anesthesia is increased. In this NMA, 36 studies reported an outcome that suggests the risk of inadequate depth of anesthesia (ie, head or limb movement, inadequate jaw relaxation). Likewise, other more serious outcomes, such as laryngospasm, may be seen with insufficient anesthetic depth. Therefore, a balance exists between preventing anesthesia overdose and providing an adequate depth of anesthesia to reduce the risk of adverse events during LMA insertion.
Propofol as the only anesthetic agent for GA induction ranked near lowest for reducing adverse events during LMA insertion; only thiopental demonstrated a poorer ability to prevent adverse outcomes. The addition of other anesthetics with propofol improved the conditions for GA induction and LMA insertion. The combination of propofol + dexmedetomidine ranked the most effective anesthetic combination for reducing adverse events evidenced by SUCRA scores. This finding is supported by the recent meta-analysis completed by Ju and colleagues, 5 which demonstrated the use of dexmedetomidine in comparison with sedative agents, such as fentanyl or midazolam, improved the success rate of LMA placement, lessened respiratory depression, and reduced adverse events, such as coughing and limb movement. In our NMA study, the benefit of dexmedetomidine was compared with all anesthetic combinations used for LMA placement available in literature.
Dexmedetomidine is an α-2 adrenoreceptor agonist that has several beneficial properties for anesthesia induction that may have helped rank it as the top anesthetic when combined with propofol for LMA insertion. 31 Unlike other anesthetics that provide sedation (eg, opioids), dexmedetomidine has minimal negative effect on respiratory drive. 31 The respiratory-sparing effect makes dexmedetomidine ideally suited for the management of a difficult airway, and it has been used for awake fiberoptic intubation. 32,33 Dexmedetomidine has analgesic, antitussive, and anxiolytic properties that have been demonstrated to reduce patient reactivity during airway-stimulating events (eg, awake intubation), sparing the need for opioids or benzodiazepines. 32,33 Moreover, dexmedetomidine also has antisialagogue properties, potentially reducing the risk for cough and laryngospasm. 32,34
In our study, the combination of propofol + dexmedetomidine ranked 13 of 20 in apnea incidence and 5 of 17 in apnea time. It is possible that the respiratory-sparing properties of dexmedetomidine were negated by the combined use of propofol. Despite the middle ranking of propofol + dexmedetomidine for apnea, the other desirable properties of dexmedetomidine may have helped it rank higher in the other outcomes of insertion failure, inadequate depth of anesthesia, and airway adverse events. Undesirable effects of dexmedetomidine include the potential for bradycardia and hypotension. 31 Of the 4 studies examining propofol + dexmedetomidine included in our meta-analysis, the incidences of bradycardia and hypotension were 6.8% and 1.3%, respectively. Overall, the quality of evidence of the included studies was low to moderate for propofol + dexmedetomidine, and further, better-quality studies are required to establish the full benefit of that anesthetic combination for LMA insertion.
The implication of this study is that propofol + dexmedetomidine could be seen as more favorable and have increased use among anesthesiologists because of its effectiveness for reducing adverse events during LMA insertion. The use of dexmedetomidine with propofol has clinical and economic considerations that go beyond reducing adverse events and increasing patient safety. First, the use of dexmedetomidine may increase the cost of GA compared with other commonly used anesthetics. Dexmedetomidine may cost twice as much as using propofol alone and as much as 4 times when compared with midazolam. 35 However, considering the total cost of managing some more serious adverse events, dexmedetomidine may prove more favorable. When used in the intensive care unit (ICU) for intubated patients, dexmedetomidine was associated with increased cost savings because of the reduction in ICU length of stay and the degree of monitoring and management. 36 Second, according to the product monograph, dexmedetomidine should be used as a slow induction bolus (1 μg/kg) given over 10 minutes to avoid adverse events such as bradycardia. 37 All studies that were included in this NMA utilized this suggested protocol of slow bolus of dexmedetomidine over 10 minutes for anesthesia induction before LMA placement. This required time for dexmedetomidine delivery compared with faster anesthetics, such as 30 seconds for remifentanil, may increase total anesthesia time and reduce the number of patients who can be seen in a clinical day. However, with optimal case organization and scheduling, the additional time required for dexmedetomidine use may be negated. Overall, the use of propofol + dexmedetomidine for LMA placement may prove to be an effective and efficient anesthetic combination for LMA insertion. Only through increased use and further clinical research can the full benefits of propofol + dexmedetomidine be appreciated.
The large number of included studies from Asia and the few from North America may be indicative of the LMA’s popularity in each of the respective regions. In Europe and Asia, the LMA has become highly favored and is used more commonly over tracheal intubation, unlike in the US, where tracheal intubation is still the most used airway device. 38–40 Only 2 studies reported the use of second-generation LMAs. Our subgroup analysis of LMA type did not demonstrate any differences, mainly because of the lack of reporting. It is possible that with increased use of second-generation LMAs, the incidence of adverse events may be reduced. 41,42
Inconsistency in what outcomes constituted successful LMA placement was notable. The most common reported set of outcomes was that of the modified scheme of Lund and Stovner, 43 examining mouth opening, ease of insertion, swallowing, coughing or gagging, head or limb movement, and laryngospasm. Another set of outcomes reported was those used by Muzi and colleagues, 44 which included jaw mobility, coughing, movement, spontaneous ventilation, breath holding, and lacrimation. Most of the included studies did not cite an existing set of LMA outcomes, and our NMA required grouping of outcomes into similar adverse event categories to permit analysis and comparison. Our study limited outcomes to only those that encompassed LMA insertion after GA induction and not outcomes that followed GA, such as patient satisfaction, sore throat, or incidence of postoperative nausea and vomiting (PONV). The meta-analysis by Joo and Perks 45 suggested that induction of GA with propofol for LMA insertion may be favored over sevoflurane due patient satisfaction and less frequent PONV. Further investigation is required to determine how the choice of anesthetics for LMA insertion may affect patient satisfaction and recovery.
A limitation of this NMA was that combining similar anesthetics into groups resulted in 3 one-arm studies without a comparator, preventing inclusion in the meta-analysis. By grouping anesthetics together, we made robust class comparisons at the expense of interdrug comparisons, such as alfentanil vs fentanyl. The ability to make comparisons is dependent on the sets of interventions being similar enough to be combined. 21 The heterogeneity analysis in this study demonstrated significant inconsistency with low to considerable heterogeneity; hence, caution is required for interpretation of the study’s results. Another limitation of the included studies was the lack of standardized dosing of anesthetics used. Propofol was used in doses that ranged from 1.5 mg/kg to 3.0 mg/kg. Higher doses of propofol are associated with hypotension, cardiorespiratory depression, and prolonged apnea. 5,41,42 Therefore, combining propofol with another anesthetic may be beneficial only if the propofol dose is reduced appropriately. Furthermore, a limitation of the included studies was the lack of preoperative standardization regarding the use of anticholinergics, sedatives, and anxiolytics. Though our subgroup analysis did not show any conclusive influence on the use of premedication with LMA insertion, we were limited by the number of studies that reported their use. The use of a benzodiazepine, such as midazolam, works synergistically with opioids and may potentially prolong apnea time following induction if an opioid is used. 46 Similarly, anticholinergics have been shown to improve the success of tracheal intubation by reducing oral secretions and preserving hemodynamic stability. 47,48 Likewise, antihistamines (eg, hydroxyzine) can be used as premedication to reduce anxiety, pruritus, and PONV as well as reduce oral secretions through their antisialagogue properties. 49 By reducing oral secretions, anticholinergics and antihistamines may potentially reduce the risk of laryngospasm during induction. 50 It is unclear how the role of premedication affects the success rate of LMA insertion, as different types of premedication can provide beneficial or unfavorable conditions for LMA insertion. Our subgroup analysis of premedications did not demonstrate a difference. If further study on anesthetics for LMA insertion is undertaken, we recommend that premedication be omitted for clear, uninfluenced comparison of the anesthetics. Finally, caution should be taken with interpretation of the SUCRA-ranked results. SUCRA does not take into consideration the magnitude of differences in effects between treatments. 51 Differences in top-ranked treatments may not be clinically appreciable. Likewise, this NMA weighed all outcomes equally during the overall SUCRA ranking. Some outcomes may be more relevant for the clinician in their setting and should be considered when selecting an anesthetic for LMA insertion given their respective clinical circumstances.
To the authors’ knowledge, a strength of this study is that it is the first NMA to compare and rank all anesthetic combinations used for GA induction. The NMA method allows for direct and indirect comparisons of multiple treatments when head-to-head comparisons are not always available in the literature. In the studies that met our inclusion criteria, 48 different anesthetics and anesthetic combinations were identified. This study demonstrates that the comparison of different anesthetic combinations with an NMA is feasible and effective.
CONCLUSION
Our systematic review and meta-analysis of anesthetics used for GA induction and LMA insertion demonstrated that all anesthetic combinations, except for those used with thiopental, reduced adverse outcomes in comparison with propofol alone. In an NMA using cumulative SUCRA-ranked scoring, very low to moderate quality evidence suggests that the combination of propofol + dexmedetomidine is the most effective anesthetic combination for minimizing adverse outcomes during LMA insertion following GA induction according to the current literature.