SUMMARY

Introduction. Endotracheal intubation without neuromuscular relaxation has become more common, but its impact on surgical units of teaching hospitals is unknown.

Objective. To assess the impact of avoiding neuromuscular relaxation in terms of incidence of hoarseness and sore throat in ambulatory surgery patients requiring endotracheal intubation in surgical unit of a teaching hospital.

Method. A quasi-experimental, non-randomized, double-blind study in adult patients undergoing outpatient surgery requiring endotracheal intubation. The non muscle relaxant group received lidocaine (1.5 mg kg^-1), propofol (1.5 - 2 mg kg^-1) and remifentanil 4 mcg *kg^-1 and the muscle relaxation group received the same technique and rocuronium 0.3 to 0.6 mg *kg^-1. The primary outcomes were assessed at days 1, 3 and 14.

Results. We enrolled 287patients, where 51.7 % received rocuronium. The incidence of hoarseness was significantly higher at 24 hours in the nonrelaxant group (26 % vs. 15 %, p value: 0.016) being not significant after 72 hours of follow up (0.6 % vs. 0 %; p: 0.37). We found no differences in the incidence of sore throat between the groups. All the study patients were asymptomatic at one week.

Conclusions. Avoidance of neuromuscular relaxation is associated with a transient (First 24 hours) increase in hoarseness after ambulatory surgery, but no difference in sore throat. We found no differences in the incidence of laryngeal symptoms after 72 hours. The addition of muscle relaxant to reduce the incidence of hoarseness can be justified.

Keywords: Intratracheal intubation, general anesthesia, adverse effects, methods, postoperative complications (Source: MeSH, NLM).

INTRODUCTION

Postoperative laryngeal symptoms like hoarseness and sore throat are common in surgical patients who require endotracheal intubation with a rate ranging from 14 % to 50 % which has influence on their degree of satisfaction (1-4).

The use of muscular relaxants for intubation is a commonly accepted practice, as it allows adequate intubation conditions (mandibular relaxation, relaxed vocal cords, and absence of motion or cough), which reduces laryngeal trauma and thus the likelihood of developing postoperative laryngeal symptoms. However different reports show contradicting results (5,6).

The use of muscle relaxants has risks like allergic reactions and residual postoperative neuromuscular blockade, which can lead to serious respiratory complications and the need to antagonize their action using medications that can produce postoperative nausea and vomiting (6-10). For these reasons and with the introduction of short acting medications like propofol and remifentanil that can produce adequate intubation conditions in moderate doses, many anesthesiologists in their daily practice do not use muscle relaxants for intubation (11-14).

The present study intends to describe the incidence of postoperative laryngeal symptoms (hoarseness and sore throat) when intubation is performed with and without neuromuscular relaxants in a sample of patients undergoing ambulatory surgery with a duration of less than 2 hours.

METHODS

Study design

A quasi-experimental double blind non randomized study.

Study population

Patients undergoing ambulatory surgery under general anesthesia in the ambulatory surgery unit of the University IPS (Universidad de Antioquia), in Medellin, Colombia.

Inclusion criteria

•Inclusion criteria: patients between 18 and 70 years undergoing elective ambulatory surgery with general anesthesia lasting less than 2 hours, that required endotracheal intubation.

•Exclusion criteria: patients with probable difficult airway, scheduled laryngeal or oropharyngeal surgery, laryngeal disease or sore throat at the moment of assessment, allergies to neuromuscular relaxants, pregnancy, and those with risk factors for slow gastric emptying.

Sample size

Previous studies have demonstrated overall an incidence of hoarseness and sore throat close to 20 % in patients with endotracheal intubation without the use of muscle relaxants (2). In this study is expected that a 10 % difference in symptoms be obtained.

A 95 % sensibility and a power of 80 % with a two tailed hypothesis were chosen. According to Pocock formula (15), the sample size was 308 patients expecting up to 10% patients loss.

At 24 hours, the incidence of hoarseness observed in the muscle relaxant group was 15% versus 26 % in the group without relaxation, a statistically significant difference (p = 0.016). A number needed to treat (NNT) of 6 (3-43) with muscle relaxants to avoid hoarseness was calculated, and an estimated relative risk (RR) of 1.4 (1.04-2.04, Confidence interval CI 95 %) to develop hoarseness when this medication is not used. There was no difference in the incidence of sore throat between the groups (26.9 % and 24 % p = 0.625).

At the 72 hour follow-up, three patients in the group with muscle relaxant and five patients the group without any relaxant, still persisted with sore throat, whereas two patients in the no muscle relaxant group still persisted with hoarseness. None of these values were statistically significant. After two weeks, one patient of the group without muscle relaxant persisted with mild hoarseness. There was no difference in the incidence of perioperative nausea (21.7 % and 20.7 % p = 0.97) and vomiting (13.1 % and 16.2 % p = 0.42) between the groups with and without muscle relaxant.

In a stratified analysis (comparing procedures lasting less than an hour vs. more than an hour), the incidence of hoarseness was higher in surgeries lasting more than an hour in both groups, 18% group with muscle relaxant and 29.2 % in the group without any muscle relaxant. There was a statistically significant difference in the incidence of early laryngeal symptoms in surgeries lasting more than an hour between groups (p = 0.03) (Figure 1).

28.6 % of intubations were performed by anesthesiologists and 71.4 % were performed by personnel in training. Even though most of the intubation in the muscle relaxant group were performed by personnel in training (77.6 %), surprisingly, we could not find statistically significant differences in laryngeal symptoms between these patients and those of the same group who were intubated by anesthesiologists. In subgroup analysis in the group with no muscle relaxant, of the 36 patients that developed hoarseness 16 were intubated by the anesthesiologist and 20 by the personnel in training (RR 1.45; IC 95 % 0.83-2.5). And in the group of patients with muscle relaxant out of 23 patients have developed hoarseness; six were intubated by the anesthesiologist and 17 by personnel in training (RR 1.22; IC 95 % 0.52-2.86).

DISCUSSION

Avoiding the use of neuromuscular relaxation is gradually becoming more common (11-14), for considerations regarding the risk benefit of serious adverse reactions (anaphylaxis, respiratory complications) against laryngeal symptoms (2).

The use of muscle relaxants is less desirable in procedures of less than 1 hour which was confirmed in this study in which the incidence of laryngeal symptoms was lower in this population. Avoidance of muscle relaxants decreases the risks of residual blockade and the need of reversal which increases postoperative nausea and vomiting (8). Likewise, a small proportion of our patients required reversal.

There are several variables that influence the incidence of perioperative hoarseness and sore throat (1-4). Our population was similar in terms of gender, smoking history, gastric reflux and surgical duration. The level of training can also affect the incidence of these symptoms, however in this study even though the analysis used was intention to treat; there was no association between the expertise level and the incidence of symptoms. This could be explained for two reasons: first the anesthesiologist had the opportunity of selecting the use or not of muscle relaxant which reflected their knowledge and confidence in each technique; and second the difference in the frequency of muscle relaxation use, which was much higher when the intubation was be performed by personnel in training.

However, we emphasize that the only factor that protected the patient from early hoarseness (at 24 hours), was the use of muscular relaxant. This could be related to that degree of abduction /opening of the vocal cords during intubation, the degree or percentage of opening of the cords was not measured in this study.

Our results are similar to those obtained in two randomized studies: Mencke et al (2) found a higher incidence of laryngeal symptoms in the group that received no atracurium (40 % vs. 15 % p = 0.02) on a sample of 80 patients using propofol and fentanyl to intubate when no relaxant was used; and Combes et al (17) analyzed the use or not of rocuronio with similar results (38 % vs. 26 %, p 0.05).

However, other studies differ from our results. Baillard et al (6) found no difference in laryngeal symptoms between the two groups (33 %) using sufentanil in the non relaxant group with a scheme similar to our study but with many more patients. More recently, Bouvet et al (18) found no difference in laryngeal symptoms in a group comparing the use of cisatracurium vs. placebo (26.5 % y 21.5 %, respectively, (p = 0.32). They randomized 130 patients, and the group intubated without muscle relaxant was given remifentanil 2 mcg/kg, a lower dose than ours, but with an incidence of hoarseness similar to ours. One explanation for the difference may be that they did not have personal in training in the study.

The studies that have assessed intubation conditions with remifentanil without the use of muscle relaxants (5), recommend doses of 3 to 4 mcg/kg (19). We used this opioid, as there is ample literature supporting it and because its pharmacological profile is ideal in ambulatory surgery.

It is still controversial whether or not muscular relaxation decreases the incidence of postoperative laryngeal symptoms. It is important to notice that in this study hoarseness when present, was significant within the first 24 hours, but at 72 hours it was virtually nonexistent in both groups.

Considering that the prevalence of smoking history was higher, although not a statistically significant difference, in the group that used no muscle relaxants, a subgroup analysis in which smokers were excluded an incidence of hoarseness of 15.7 % against 23.9 % was found for the group with and without the use of muscle relaxants. This exercise shows a statistically significant trend for the decrease of laryngeal symptoms with the addition of muscle relaxant (p = 0,051).

One of the limitations of the study is the bias of inclusion, as the use of muscle relaxant was determined by the criteria of the anesthesiologist. This was allowed, as the intention of this study was not to modify practice, but to understand the behavior of these variables in our institution. Albeit this consideration, our results were similar to those of randomized studies.

In summary, we can say that endotracheal intubation without muscle relaxant is related to a higher incidence of temporary hoarseness but not sore throat at 24 hours. There are no differences in the incidence of laryngeal symptoms at 72 hours. This study justifies the addition of muscular relaxant to decrease the incidence of early hoarseness in patients subject to surgery lasting more than 1 hour. To accept this practice as a routine, studies with larger samples using more rigorous methods and including cost analysis should be performed.

ACKNOWLEDGMENTS

To Fernando Montoya, for his counseling, to the interns of the Institución Prestadora de Servicios de Salud (IPS) of the Universidad de Antioquia, for their collaboration.

REFERENCES