Preoperative airway assessment - experience gained from a multicentre cluster randomised trial and the Danish Anaesthesia Database.

Difficulties with airway management in relation to general anaesthesia have been a challenge for the anaesthesiologist since the birth of anaesthesia. Massive landmark improvements have been made and general anaesthesia is now regarded as a safe procedure. However, rare, difficult airway management still occurs and it prompts increased risk of morbidity and mortality - especially when not anticipated. Several preoperative risk factors for airway difficulties have been identified, yet none have convincing diagnostic accuracy as stand alone tests. Combining several risk factors increase the predictive value of the test and multivariable risk models have been developed. The Simplified Airway Risk Index (SARI) is a predictive model developed for anticipation of a difficult direct laryngoscopy. However, neither the diagnostic accuracy of the SARI nor of any other model has been tested prospectively and compared with existing practice for airway assessment in a randomised trial setting. The first objective of this thesis was to quantify the proportion of unanticipated difficult intubation and difficult mask ventilation in Denmark. The second objective was to design a cluster randomised trial, using state of the art methodology, in order to test the clinical impact of using the SARI for preoperative airway assessment compared with a clinical judgement based on usual practice for airway assessment. Finally, to test if implementation of the SARI would reduce the proportion of unanticipated difficult intubation compared with usual care for airway assessment. This thesis is based on data from the Danish Anaesthesia Database (DAD). Paper 1 presents an observational cohort study on 188,064 patients who underwent tracheal intubation from 2008 to 2011. Data on the anaesthesiologists' preoperative anticipations of airway difficulties was compared with actual airway management conditions, thus enabling an estimation of the proportion of unanticipated difficulties with intubation and mask ventilation. Papers 2 and 3 outline the methodology and the pre-trial calculations and considerations leading to the DIFFICAIR trial described in Paper 4. The trial was designed to randomise anaesthesia department to either thorough education in, and subsequent use of the SARI for preoperative airway assessment or to continue usual care. Registration of the SARI in DAD was made mandatory in SARI departments and impossible in usual care departments. Conditions regarding anticipation of difficulties and actual airway managements were recorded as for Paper 1. DAD data made it possible to estimate an appropriate sample size, considering the between cluster variation, and to construct a stratification variable based on 2011 baseline values of the primary outcome used in the DIFFICAIR trial. Paper 1 revealed that 1.86% of all patients who were intubated, but not planned for advanced intubation techniques (e.g. video laryngoscopy), were unanticipated difficult to intubate. However, 75 to 93% of all difficult intubations were unanticipated. Furthermore, 94% of all difficult mask ventilations were unanticipated. In Paper 4, 59,514 patients were included in the primary analyses. The proportion of unanticipated difficult intubations was 2.38% (696/29,209) in SARI departments and 2.39% (723/30,305) in usual care departments. The adjusted odds ratio was 1.03 (95% CI: 0.77-1.38), p = 0.84. No significant differences were detected in other adjusted outcome measures and neither a 58% increase in patients anticipated to have intubation difficulties nor an 84% increase in patients scheduled for advanced intubation techniques in SARI departments reached statistical significance, p = 0.29 and p = 0.06 respectively. The papers constituting this thesis demonstrate that at high proportion of airway management difficulties are unanticipated. In a cluster randomised trial it was not possible to reduce the proportion of unanticipated difficult intubation in daily clinical practice by implementing a systematic approach for airway assessment compared with usual care. However, implementation of the SARI may increase the anticipation of intubation difficulties and it may change practice towards advanced intubation techniques. This thesis underlines the continued challenge anaesthesiologists face in predicting airway management related difficulties.