Patients with dental cellulitis may have laryngeal deviation and oedema, increased risk of bleeding due to inflammation and limitation of mouth opening, all leading to difficulty in airway management. If there is posterior mandibular localisation, constant trismus exacerbates the difficulties.1 In the case of dental cellulitis, general anaesthesia is indicated when local anaesthesia for the removal of the responsible tooth is impossible. This is likely to be the case when there is extensive cellulitis, particularly with extension below the mylohyoid muscle, allowing access to the submandibular space, the neck and the mediastinum. Few studies have examined airway management of patients with dental cellulitis.1,2 One assessed a novel system for anticipated difficulty in airway management in dental cellulitis and 69% of patients were intubated using fibre-optic bronchoscopy.2 Although clinical data and scientific evidence are scarce, fibre-optic tracheal intubation has been recommended in patients with dental cellulitis, especially for those with mandibular localisation or abscess of the floor of the mouth.1 Fibre-optic intubation without neuromuscular blocking agents is a safe procedure for visualisation of the vocal cords and its routine use is justified in the hands of experienced anaesthesiologists.3,4 Nevertheless dental cellulitis and its consequences markedly enhance the difficulties of fibre-optic tracheal intubation. The equipment is expensive5 and maintenance time consuming (preparation, cleaning and sterilisation), and appropriate training is difficult, particularly with regard to the management of difficult cases.
Our hospital is the referral centre for maxillofacial surgery for a Paris district with a catchment area of 12 million, accepting emergency and the most severe dental cellulitis cases. The airway management of severe dental cellulitis is consequently a frequent problem for our department. Because experienced anaesthesiologists in our team have suggested that fibre-optic intubation might be unnecessary in patients with dental cellulitis and chose not to use it, we designed an intubation procedure in accordance with their routine practice, and according to the difficult airway algorithm of the American Society of Anesthesiologists and French Society of Anaesthesia and Intensive Care.6,7 Our algorithm (Fig. 1) proposes direct laryngoscopy as the first step of airway management once those with predicted difficult mask ventilation or difficult intubation have been excluded. Consequently, fibre-optic intubation was limited to those considered to be at high risk of failed intubation (difficult intubation risk factor per se or unable to tolerate the supine position) or difficult ventilation before taking into account the presence of dental cellulitis. Our study attempts to evaluate the efficacy and safety of this algorithm for the airway management of patients with dental cellulitis with an expected low incidence (<5%) of difficult mask ventilation.
Material and methods
We performed a prospective observational single centre study between February 2008 and February 2009. Ethical approval was obtained from our ethical committee (Comité pour la Protection des Personnes Groupe Hospitalier Pitié-Salpêtrière, Paris, France) (Chairperson Golmard JL). Because the routine procedure was not modified by the research protocol, informed consent was waived. In accordance with French legislation, patients were informed about the procedure and could refuse the inclusion of their data in the analysis.
All patients suffering from dental cellulitis and requiring emergency surgery were included, except pregnant women and those under 18 years. The emergency anaesthesia team rotated every day and comprised senior anaesthesiologists (who had worked in the unit for 5 years or more), junior anaesthesiologists (with a minimum of 3 years of professional experience in anaesthesiology out of their 5 year training programme) and nurse anaesthetists. If a nurse anaesthetist or junior anaesthesiologist could not intubate at the first attempt, a senior anaesthesiologist always stepped in.
On arrival in the operating theatre, intravenous access and routine monitoring, including heart rate, non-invasive arterial blood pressure, ventilatory frequency, and peripheral oxygen saturation, was applied. Equipment for emergency ventilation via the cricothyroid route was always ready. The maxillofacial surgeon was present during the induction of anaesthesia to perform emergency tracheotomy immediately if necessary. The procedure began with checking tolerance of the supine position even if there was evidence of airway obstruction (difficulty in swallowing). If the patient did not tolerate the supine position (dyspnoea and/or pharyngeal obstruction), had predictable difficult mask ventilation as defined by Kheterpal et al.8 or had a predicted difficult intubation as defined by Arné et al.,9 awake fibre-optic tracheal intubation was performed. We chose the criteria of Arné et al.9 because they were based on a population undergoing ear, nose and throat surgery. If the patient tolerated the supine position without a predictive factor for difficult mask ventilation and difficult intubation, local anaesthesia (three oral sprays with 30 s of gargle and three nasal sprays in each nostril of lidocaine 2%) and a 4 min pre-oxygenation procedure was performed. Bispectral index (Aspect Medical Systems Inc., Natick, Massachusetts, USA) was used to monitor depth of anaesthesia which was induced using intravenous propofol10 through an automated syringe driver (concentration objective mode, Schnider model, Fresenius Vial, Le Grand Chemin, France) until a bispectral index score (BIS) between 30 and 50 was obtained.
The procedure is illustrated in Fig. 1. We used single-use metal Macintosh blades for laryngoscopy (Comepa industries, Bagnolet, France) to minimise the rate of tracheal intubation failure in emergency surgery.11,12
Trismus was defined as being unable to open the mouth normally because of sustained contraction of the jaw muscles in reaction to pain and local inflammation.
To calculate the difficult mask ventilation risk score from Kheterpal et al.,8 we noted the following variables during the pre-anaesthetic consultation: age, size, weight, presence of beard, jaw protrusion and history of snoring and sleep apnoea. A threshold of n value of 2 criteria was defined as predictable difficult mask ventilation. To calculate the difficult intubation risk score from Arné et al.,9 we noted the following variables during the pre-anaesthetic consultation: a history of difficult intubation, disorders associated with difficult intubation (cervicofacial surgery, cervicofacial radiotherapy, retrognathia), mouth opening, thyromental distance and mobility of the head and neck. It was impossible to include the Mallampati class in the two scores13 because of the very high incidence of preoperative trismus. For the same reason, we decided to exclude mouth opening from the Arné score.9 Others variables collected during the pre-anaesthetic consultation included the duration of the dental abscess, the teeth causing the infection, localisation of cellulitis [anterior (31–35, 41–45) or posterior (36–38, 46–48) mandibular, maxillary] and the extension of the cellulitis to the submandibular space. During the induction of anaesthesia the following variables were noted: the maximal effect site concentration of propofol and the total dose administered until intubation (time to inflation of the cuff), the BIS at intubation and the Cormack–Lehane scale.14 All data were recorded prospectively on a special sheet.
The primary endpoint was the proportion of difficult mask ventilation which was expected to be less than 5%. We choose this end point because, in these emergency conditions, the key issue is to avoid the life-threatening association of difficult intubation and difficult mask ventilation, although this situation is very rare.15 The anaesthesiologist rated mask ventilation as difficult when it was considered that the difficulty was clinically relevant and potentially hazardous. Specifically these were as follows: when mask ventilation had to be maintained for a longer time because an unassisted anaesthesiologist was unable to maintain SpO2 of at least 93%; if there was a major gas flow leak around the face mask, requiring a gas flow greater than 15 l min−1 and the oxygen flush valve had to be used more than twice; when there was no perceptible chest movement; or when there was no expired carbon dioxide curve on the capnograph; and when it was necessary to perform two-handed mask ventilation, and a change of operator was required, as previously reported.16
The secondary end points were the number of complications of laryngoscopy (decrease in peripheral oxygen saturation (SpO2) below 93%, low systolic arterial blood pressure (<90 mmHg), bradycardia (<60 beats/min) and breach of the abscess); the incidence of persistent trismus after induction of anaesthesia; the number of failures of intubation at the first attempt; and the number of tracheotomies required for airway management. The latter was expected to be less than 10%.
We estimated that the maximum acceptable incidence of difficult mask ventilation should be 5%. Assuming this value as the upper limit of the corresponding 95% confidence interval (CI), we calculated that 125 consecutive patients (approximately 1-year recruitment) would allow a maximum of one patient with difficult mask ventilation.
Normality of the distribution of variables was assessed using the Kolmogorov–Smirnov test. Data are expressed as mean ± SD, median (interquartile range) or percentage with their associated 95% CI. Comparison of means was performed using the Student's t-test, comparison of medians using the Mann–Whitney U-test and comparison of percentages using Fisher's exact test. To analyse factors associated with failure at the first attempt, a multivariate analysis was performed using backward logistic regression and odds ratios were calculated with their 95% CI. To limit over fitting, only significant (P < 0.10) variables in the univariate analysis were included in the multivariate analysis. The calibration (Hosmer–Lemeshow test) and discrimination (C-statistics) of the multivariate model were assessed. To validate this model, we performed a 10-fold cross validation which is recognised as one of the most efficient methods available.17 All P values are two-tailed and a P value of less than 0.05 was considered significant.
One hundred and twenty-seven consecutive patients were eligible and were included. Baseline characteristics are presented in Table 1.8,9 All patients had preoperative trismus.
None fulfilled the criteria for anticipated difficult mask ventilation or anticipated difficult tracheal intubation and one did not tolerate the supine position. This unique patient (1%, 95%CI 0–4%) was successfully intubated using awake fibre-optic intubation.
None of the remaining 126 experienced difficult mask ventilation (0%, 95% CI 0–3%). One hundred and seventeen (93%, 95% CI 87–97%) were successfully intubated at the first attempt via the nasotracheal route, 54 (43%, 95% CI 35–52%) by a senior anaesthesiologist, 32 (25%, 95% CI 18–34%) by a junior anaesthesiologist and 31 (25%, 95% CI 17–33%) by a nurse anaesthetist. Among the remaining nine patients, seven were successfully intubated at the second attempt by the orotracheal route, two of them requiring a molar hold because of persistent trismus after general anaesthesia. Tracheotomy was performed in the last two patients. The study flowchart is shown in Fig. 2.
Patients whose intubation failed at the first attempt required more propofol, as the airway procedure was obviously longer, but the estimated concentration of propofol was not significantly different when compared with patients whose intubation was successful at the first attempt (Table 2). Oxygen desaturation also occurred more frequently in patients whose intubation failed at the first attempt (Table 2), but no other complication was recorded. In the multivariate analysis, only two variables were significantly associated with a failed intubation at the first attempt: presence of retrognathia and extension of cellulitis to the floor of the mouth (Table 3).
Two tracheotomies (2%, 95% CI 0–6%) were performed preoperatively according to our algorithm because of difficult airway management and two were performed immediately following surgery because of wide surgical debridement causing major oedema. All these cases had a posterior mandibular abscess with an extension to the oral floor. The former two patients had a disorder associated with difficult intubation (cervicofacial radiotherapy and retrognathia, respectively), but had no other difficult intubation risk factor.
Our study shows that direct laryngoscopy is safe and permits appropriate airway management in most patients with dental cellulitis. The use of fibre-optic intubation can be confined to a small proportion of patients easily identified before anaesthesia.
Difficult tracheal intubation was anticipated in all our patients: approximately 40% had an extension to the submandibular space and others had wide extension. Although evidence from clinical studies is sparse,1,2 awake intubation using fiberoscopy is usually recommended,18–20 but support for this recommendation is relatively poor. Complications19 and failure20 of fibre-optic intubation have been reported especially in emergency situations.18 Benumof et al.21 recommended awake direct laryngoscopy with appropriate local anaesthesia to decide whether tracheal intubation by direct laryngoscopy or by fiberoscopy was the best procedure. However, as it requires total cooperation from the patient, awake direct laryngoscopy is difficult to perform under emergency conditions. It may also compromise further fibre-optic intubation because of oedema and bleeding and may be associated with severe complications, including anoxic cardiac arrest, if performed by an anaesthesiologist with only a modest training in this procedure. This may be more likely when facing a difficult case or when an unexpected complication (bleeding, laryngospasm) occurs.
Part of the difficulty of tracheal intubation in patients with dental cellulitis is related to limited mouth opening caused by both trismus and the cellulitis itself. We observed that 49% of our patients had a mouth opening of less than 20 mm which is usually considered to predict a very difficult tracheal intubation.9 However, induction of anaesthesia and muscle relaxation relieved the trismus in most of our patients (96%) and appropriate mouth opening could be obtained in the rest by using a molar hold. Thus, we can conclude that mouth opening is not a good predictor of difficult tracheal intubation in patients with dental cellulitis. This is the reason why we did not use the Mallampati score or any other criteria related to mouth opening in our preoperative assessment of expected difficult mask ventilation8 or tracheal intubation (Table 1).9
We decided to perform fibre-optic intubation not only in those patients who fulfilled the criteria for difficult mask ventilation and/or difficult tracheal intubation but also in those who did not tolerate the supine position. We considered that these patients were at a very high risk of intubation failure and that they justified the systematic use of fibre-optic intubation. Indeed, fibre-optic intubation is recommended by many authors as soon as airway obstruction is apparent.18 We cannot dismiss the hypothesis that direct laryngoscopy might also have been successful in patients who cannot tolerate the supine position. However, this concerns only a few patients (4%) and we think that a precautionary principle should be applied until new information is available in this subgroup. Only a large multicentre study will answer this question. We decided not to use other devices such as video-laryngoscopes because we wished to observe the performance of standard laryngoscopes in a technique as basic as nasotracheal intubation. It is possible that new devices could help as an alternative to the fibre-optic technique when classic nasotracheal intubation fails, but this remains to be demonstrated.
Intubation failed at the first attempt in a small proportion of our patients (7%). This proportion is close to that (3–6%) observed in emergency conditions requiring rapid sequence induction with no predictable difficult intubation.11 In these patients, the total dose of propofol was significantly higher, probably because airway control took longer, and associated desaturation occurred more frequently (Table 2). However, these decreases in SpO2 were not severe and without significant consequences as shown by the absence of hypotension and bradycardia (Table 2). In the multivariate analysis, two variables were significantly associated with failure of intubation at the first attempt, retrognathia and the extension of the cellulitis to the submandibular space. In contrast, the posterior location of the causal tooth was not a risk factor (Table 3).
Only two patients required preoperative tracheotomy for airway management. These patients could be ventilated but were impossible to intubate using standard laryngoscopy. Nasotracheal intubation failed, and also orotracheal intubation using an Eschman stylet. We then decided not to attempt fibre-optic intubation because of concern that bleeding and oedema, induced by several attempts at laryngoscopy, might complicate the procedure.22 In such cases, we preferred a rapid decision to perform tracheotomy for safety reasons. However, further studies are needed to develop an appropriate algorithm in these patients in order to avoid tracheotomy for airway management completely, if possible.
Our study has some limitations. First, it was performed in emergency conditions and our patient group should be considered at high risk of failed intubation. Even so, our results could probably be safely applied to patients in whom surgery is not considered urgent. Second, junior anaesthesiologists and nurse anaesthetists were recruited to participate in the study in association with senior. This was desirable because they are representative of the intubation providers in academic emergency and critical care departments. Supervision of intubation by a senior anaesthesiologist has been recently been shown to significantly decrease failure rate.23 Third, our study was conducted in adult patients and may not apply to paediatric patients. Fourth, several important safeguards present during the study deserve emphasis: at least one of the senior anaesthesiologists was experienced in fibre-optic intubation; the surgeon was present in the operating room before induction of anaesthesia, to perform a rapid tracheotomy if required; emergency equipment to perform trans tracheal oxygenation and ventilation and retrograde tracheal intubation was available.24,25
In conclusion, most patients with dental cellulitis may be intubated using direct laryngoscopy even if mouth opening is limited. Indications for fibre-optic intubation should be restricted to a small group of patients with other causes of expected difficult intubation or difficult mask intubation, and also those do not tolerate supine position.
Support was provided solely by departmental sources.
The authors would like to thank Dr David Baker, DM, FRCA, Emeritus Consultant Anaesthesiologist, CHU Necker – Enfants Malades, Paris for reviewing this paper.
Financial support and sponsorship: none.
Conflict of interest: none.
1. Reyford H, Boufflers E, Baralle MM, et al. Dental cellulitis
and tracheal intubation [in French]. Ann Fr Anesth Reanim
2. Darshane S, Groom P, Charters P. Responsive contingency planning: a novel system for anticipated difficulty in airway management
in dental abscess. Br J Anaesth
3. Heidegger T, Starzyk L, Villiger CR, et al. Fiberoptic intubation and laryngeal morbidity: a randomized controlled trial. Anesthesiology
4. Ovassapian A. The flexible bronchoscope. A tool for anesthesiologists. Clin Chest Med
5. Ahmad M, Dweik RA. Future of flexible bronchoscopy. Clin Chest Med
6. Langeron O, Bourgain JL, Laccoureye O, et al.
Difficult airway algorithms and management [in French]. Ann Fr Anesth Reanim
7. American Society of Anesthesiologists Task Force on Management of the Difficult Airway. Practice guidelines for management of the difficult airway: an updated report by the American Society of Anesthesiologists Task Force on Management of the Difficult Airway. Anesthesiology
8. Kheterpal S, Han R, Tremper KK, et al. Incidence and predictors of difficult and impossible mask ventilation. Anesthesiology
9. Arné J, Descoins P, Fusciardi J, et al. Preoperative assessment for difficult intubation in general and ENT surgery: predictive value of a clinical multivariate risk index. Br J Anaesth
10. Symington L, Thakore S. A review of the use of propofol for procedural sedation in the emergency department. Emerg Med J
11. Amour J, Le Manach YL, Borel M, et al. Comparison of single-use and reusable metal laryngoscope blades for orotracheal intubation during rapid sequence induction of anesthesia: a multicenter cluster randomized study. Anesthesiology
12. Amour J, Marmion F, Birenbaum A, et al. Comparison of plastic single-use and metal reusable laryngoscope blades for orotracheal intubation during rapid sequence induction of anesthesia. Anesthesiology
13. Mallampati SR, Gatt SP, Gugino LD, et al. A clinical sign to predict difficult tracheal intubation: a prospective study. Can Anaesth Soc J
14. Cormack RS, Lehane J. Difficult tracheal intubation in obstetrics. Anaesthesia
15. Kheterpal S, Martin L, Shanks AM, Tremper KK. Prediction and outcomes of impossible mask ventilation. Anesthesiology
16. Langeron O, Masso E, Huraux C, et al. Prediction of difficult mask ventilation. Anesthesiology
17. Molinaro AM, Simon R, Pfeiffer RM. Prediction error estimation: a comparison of resampling method. Bioinformatics
18. Walls RM, Murphy MF. Manual of emergency airway management
. 3rd ed. Wolters Kluwer Health; 2008. pp. 149–158.
19. Kalfon F, Dubost J. Use of fiberoptic bronchoscope for difficult intubation in maxillofacial surgery [in French]. Ann Fr Anesth Reanim
20. Afilalo M, Guttman A, Stern E, et al. Fiberoptic intubation in the emergency department: a case series. J Emerg Med
21. Benumof J, Hagberg CA. Benumof's airway management
. Mosby Elsevier; 2007. pp. 891.
22. Riou B, Langeron O, Guesde R. Intubation trachéale difficile. Rev Med Suisse Romande
23. Lundstrøm LH, Møller AM, Rosenstock C, et al. High body mass index is a weak predictor for difficult and failed tracheal intubation: a cohort study of 91 332 consecutive patients scheduled for direct laryngoscopy
registered in the Danish Anesthesia Database. Anesthesiology
24. Combes X, Le Roux B, Suen P, et al. Unanticipated difficult airway in anesthetized patients: prospective validation of a management algorithm. Anesthesiology
25. Barriot P, Riou B. Retrograde tracheal intubation in trauma patients. Crit Care Med
Keywords:© 2012 European Society of Anaesthesiology
airway management; dental cellulitis; laryngoscopy