Neuromuscular blockade and airway management during endotracheal intubation in Brazilian intensive care units: a national survey

Objective To describe the use of neuromuscular blockade as well as other practices among Brazilian physicians in adult intensive care units. Methods An online national survey was designed and administered to Brazilian intensivists. Questions were selected using the Delphi method and assessed physicians’ demographic data, intensive care unit characteristics, practices regarding airway management, use of neuromuscular blockade and sedation during endotracheal intubation in the intensive care unit. As a secondary outcome, we applied a multivariate analysis to evaluate factors associated with the use of neuromuscular blockade. Results Five hundred sixty-five intensivists from all Brazilian regions responded to the questionnaire. The majority of respondents were male (65%), with a mean age of 38 ( 8.4 years, and 58.5% had a board certification in critical care. Only 40.7% of the intensivists reported the use of neuromuscular blockade during all or in more than 75% of endotracheal intubations. In the multivariate analysis, the number of intubations performed monthly and physician specialization in anesthesiology were directly associated with frequent use of neuromuscular blockade. Etomidate and ketamine were more commonly used in the clinical situation of hypotension and shock, while propofol and midazolam were more commonly prescribed in the situation of clinical stability. Conclusion The reported use of neuromuscular blockade was low among intensivists, and sedative drugs were chosen in accordance with patient hemodynamic stability. These results may help the design of future studies regarding airway management in Brazil.

Objective: To describe the use of neuromuscular blockade as well as other practices among Brazilian physicians in adult intensive care units.
Methods: An online national survey was designed and administered to Brazilian intensivists. Questions were selected using the Delphi method and assessed physicians' demographic data, intensive care unit characteristics, practices regarding airway management, use of neuromuscular blockade and sedation during endotracheal intubation in the intensive care unit. As a secondary outcome, we applied a multivariate analysis to evaluate factors associated with the use of neuromuscular blockade.
Results: Five hundred sixty-five intensivists from all Brazilian regions responded to the questionnaire. The majority of respondents were male (65%), with a mean age of 38 ± 8.4 years, and 58.5% had a board certification in critical care. Only 40.7% of the intensivists reported the use of neuromuscular blockade during all or in more than 75% of endotracheal intubations. In the multivariate analysis, the number of intubations performed monthly and physician specialization in anesthesiology were directly associated with frequent use of neuromuscular blockade. Etomidate and ketamine were more commonly used in the clinical situation of hypotension and shock, while propofol and midazolam were more commonly prescribed in the situation of clinical stability.
Conclusion: The reported use of neuromuscular blockade was low among intensivists, and sedative drugs were chosen in accordance with patient hemodynamic stability. These results may help the design of future studies regarding airway management in Brazil. facilitate airway visualization and reduce procedurerelated complications. (8) In the operating room, several clinical trials and a systematic review have indicated the superiority of the routine use of NMB during airway management. (9) However, in the ICU setting, this recommendation of routine use of NMB during EI is supported only by the evidence from the operating room and observational studies. (9)(10)(11) Moreover, bedside practice is highly variable among intensivists, and the use of NMB in EI varies from 20 to 90% in different cohorts. (12)(13)(14) Therefore, considering the paucity of data regarding the use of NMB during EI, as well as other airway management practices used in the ICU, we decided to perform a national survey among adult intensive care physicians in Brazil.
The objective of this survey was to describe the use of NMB during orotracheal intubation in the ICU as well as intensivist perception regarding the use of NMB. As secondary outcomes, we decided to evaluate factors associated with frequent use of NMB during EI and to describe common practices during EI and airway management.

METHODS
The study protocol was approved by the Research Ethics Committee of Faculdade de Medicina da Universidade de São Paulo (number 14637519.2.0000.0065). Participation in the survey was strictly voluntary.
The questionnaire was designed using an informal Delphi process among all authors. One of the authors designed the survey and was responsible for facilitating communication between authors. An initial draft was distributed to all study authors, and responses were compiled into a new draft. Subsequently, the responses were once again reviewed by all authors, and this process was repeated until a consensus was reached with the final questionnaire.
This was an adult ICU physician-centered survey and consisted of 28 questions (Table 1S -Supplementary material). The survey aims were to describe demographic characteristics of the participants and their usual practice with airway devices; to describe the availability of airway resources at the participants' workplace; to evaluate selfreported use of NMB during EI and perception of the possible benefits associated with the use of NMB; to study which drugs are more commonly prescribed during EI in patients with and without hemodynamic instability, and describe physicians' strategies in a "do not intubate, do not ventilate" scenarios using clinical vignettes.
Most of the questions in this survey allowed only a single answer. However, participants were able to give multiple answers to questions regarding sedative agents and neuromuscular blockers commonly used during EI so that some percentages could be greater than 100%.
This study was conducted with logistics support from AMIBnet (the Brazilian network of research in ICUs), and the survey was sent to several email lists and those subscribed to the AMIBnet mailing list. Respondents were asked to complete the survey and invited to forward it to a colleague. Reminders were sent after 1 month.
Our primary outcome was to evaluate the use of NMB during EI and physicians' perceptions regarding possible benefits of the use of NMB during airway management. Secondary outcomes included evaluation of factors associated with the use of NMB during EI, the description of drugs commonly used during EI in patients with and without hemodynamic instability, available airway resources in respondents' ICUs and the respondents' demographic data.

Statistical analysis
Considering the population of 10,000 ICU physicians in Brazil in accordance with a previously reported census from Associação de Medicina Intensiva Brasileira (AMIB), we estimated that a sample size of 370 responses would be enough to represent the intended population with a confidence level and a confidence interval of 95% and 5%, respectively.
Continuous data were reported as the mean (standard deviation) and median (25th percentile, 75th percentile) as appropriate. Categorical variables were presented as absolute numbers and percentages.
We used a logistic regression model to assess the association between regular use of NMB in EI and independent variables as follows: the response to the question "How often do you use neuromuscular blockade during EI"? was categorized "as frequent" use and "infrequent" use and adopted as the dependent variable. The Likert scale was used for questionnaire responses; never used, rarely used and regularly used were categorized as "infrequent" use, and frequently and always used were categorized as "frequent use". We included variables associated with previous experience in critical care, such as board certification in critical care, average number of EIs performed monthly, time since graduation, last medical residency and any previous difficult airway course, as independent variables. The number of EIs was categorized as lower or higher than 3 EIs/month for the analysis. We chose the independent variables to be included in the model based on clinical relevance. We determined the model with the best fit using the Bayesian information criteria and the final model calibration was determined by the Hosmer-Lemeshow test using deciles and the Gruppo Italiano per la Valutazione Degli Interventi in Terapia Intensiva (GiViTi) calibration belt. We respected the minimum of one variable to 10 outcomes proportion in model building.
A two-sided p value < 0.05 was considered statistically significant. Odds ratios (OR) and 95% confidence intervals were calculated for associated measurements. The commercial statistical software package used was STATA version 15.1.

Sample characteristics
Six hundred and twenty-four physicians responded to the survey. However, after removal of incomplete and duplicated responses, 565 remained for final analysis. All regions from Brazil were properly represented in this survey ( Figure 1S -Supplementary material). The surveyed workforce was predominantly male (65%), with a mean age of 38 ± 8.4 years. Sixty-eight percent previously attended a difficult airway course, and 403 (71%) reported feeling confident or very confident in managing a difficult airway situation. Physician demographic data are described in table 1. Most responders worked predominantly in a private ICU (52.4%) with a medical-surgical profile (62.7%). Rescue devices such as supraglottic devices were not available in all ICUs ( Table 2).

Use of neuromuscular blockade and factors associated with its use
Only 40.7% of the intensivists reported the use of NMB during all or more than 75% of EI procedures. Thirty-seven percent of the physicians reported using NMB in less than 25% of the procedures. Ninety-one percent of responders described that the use of NMB facilitates the visualization of the vocal cords during EI. In contrast, 22.3% reported that its use could increase the procedural risk.
In the multivariate analysis, the number of EIs performed monthly and physician specialization in anesthesiology were directly associated with frequent use of NMB during EI. In contrast, time (in years) since graduation was inversely associated with frequent use of NMB (Table 3).

Attitude towards a "do not intubate, do not ventilate" scenario
In a "do not intubate, do not ventilate" scenario, 82% of the responders reported the use of a supraglottic device as a rescue strategy. Nine percent and 7.8% reported a new attempt of direct laryngoscopy and urgent cricothyroidotomy as rescue therapy, respectively.

Attitudes on the use of drugs during airway management
Etomidate and ketamine were the most commonly prescribed drugs in the clinical scenario of shock with   (7) ICU -intensive care unit. Results expressed as n (%) or means (± standard deviations). the need for vasoactive drugs, while propofol and midazolam were more frequently used in the clinical scenario without hypotension (Figure 1). Sixty-seven percent and 72% of responders reported fentanyl use in the clinical scenario with and without shock, respectively. Succinylcholine and Rocuronium were the most commonly reported NMBs for EI (82.1 and 46.1% of the intensivists, respectively).

DISCUSSION
The main findings of this survey are summarized as follows: Brazilian critical care physicians report a low use of NMB during EI in the ICU; not all ICU physicians reported the use of supraglottic airway devices as a first-line rescue strategy in a "do not intubate, do not ventilate" scenario; and fentanyl is widely used during EI, and the choice of sedative drugs may vary in accordance with patient hemodynamic status. Taken as a whole, these data suggest there is substantial room for improvement in intubation practices and suggest clinical equipoise for clinical trials assessing the use of neuromuscular blockers and fentanyl in ICU intubations.
The finding that most intensivists do not use NMB regularly during EI in the ICU is not in accordance with published guidelines recommending the routine use of NMB. (5,6) Although a recently published meta-analysis of 34 trials evaluating the impact of NMB in tracheal intubation found that avoidance of NMB was associated with increased risk of patient discomfort, airway injury and difficult laryngoscopy, (8) we must highlight that current evidence is focused mainly on the operating room. Data on emergency departments and critical care units are still sparse. Two observational cohort studies found a reduction in procedure-related complications with the use of NMB during EI in the ICU. (11,15) Similarly, in a propensity matched analysis, the use of NMB improved first attempt success from 69.5% to 80.9% in a cohort of critically ill patients. (10) However, to date, no randomized clinical trial has evaluated the use of NMB in the critical care setting, which could contribute to clinicians' attitudes towards its use, especially non-anesthetists and physicians who do not intubate frequently.
Our results differ from those of previously published studies. In an Australian and New Zealand survey regarding airway management, most physicians reported the use of rapid sequence intubation in association with NMB in all or in the majority of EI in the ICU. (16) Similarly, in a multicenter national cohort in Scotland, NMB was withheld in only 8% of the EI. (13) In this survey, although 91% of responders reported that the use of NMB facilitates EI in the ICU, 22% also reported that its use could increase procedure-associated risk, suggesting a fear of a "do not intubate, do not ventilate" scenario and its consequences. A possible explanation for this difference is the fact that intensive care physician training is associated with anesthesia departments in several countries and, therefore, practices from the operating room may be more easily transposed to the ICU. In Brazil, most intensivists are not anesthetists by training, and in our survey, only a minimum proportion of responders were anesthetists. Despite the small number of responses, anesthesia training was directly associated with frequent use of NMB in the multivariate analysis. Additionally, a higher number of EIs performed monthly was directly associated with the use of NMB, suggesting that the more experienced the physician, the higher the probability of using NMB.
In regard to sedative drugs and analgesics used during EI, we found that etomidate and ketamine were more commonly used in the clinical situation of hypotension and shock. On the other hand, propofol and midazolam were more commonly prescribed in the situation of clinical stability. This is endorsed by previous recommendations in which the choice of induction drug should be dictated by patient hemodynamic status. (5) Moreover, we found a higher report of succinylcholine use than rocuronium use in our survey. Although some recommendations suggest that rocuronium should be preferred over succinylcholine in EI, a randomized controlled trial found both drugs to be equivalent in the ICU. (17) To our knowledge, this is the first large survey regarding airway management, the use of sedatives and NMB in Brazil. Furthermore, these results may help the design of future research regarding airway management in Brazil along with the standardization of practices to allow a safer procedure. However, our study has several limitations. First, a selection bias is always a possibility in a survey, and we did not assess the nonresponse rate. It is possible that nonresponders had a different prescribing practice. We reached a highly qualified sample with more than half of the responders having board certification, which may not represent the intended population. However, we reached a response rate higher than initially expected, which could minimize these risks. Second, since patient medical records were not assessed, a certain degree of recall bias may have influenced physicians' responses. Third, most responses were from the Southeast region of Brazil, which may limit the generalizability of the data. However, all regions were properly represented in accordance with the national ICU physician's registry.

CONCLUSION
The reported use of neuromuscular blockers is low among intensivists, and almost a quart of physicians raised the concern of increased risk with the use of neuromuscular blockers. Fentanyl use was not determined by the patient's hemodynamic status, and sedative drugs were chosen according to the patient's hemodynamic status. The availability of different airway devices, as well as physicians' knowledge and confidence regarding difficult airway management, was heterogeneous.

AUTHORS' CONTRIBUTIONS:
P.V. Mendes and B.A.M.P. Besen had full access to all data and take responsibility for the writing of the manuscript, the integrity of the data and the accuracy of the final analysis. J.G.R. Ramos, F.H. Lacerda and L.U. Taniguchi contributed substantially to the study design, data acquisition, and analysis and interpretation of the results. All authors read and approved the final manuscript.