METHODS

This was a multicentre, prospective observational study running from 11 January 1999 to 10 January 2001 in seven Scottish urban teaching hospitals. Every patient on whom endotracheal intubation was attempted in the participating emergency departments was eligible for the study. Patients who had been successfully intubated before arrival at hospital by either ambulance paramedics or medical staff were specifically excluded from the study.

A form was completed by the intubating doctor immediately after the intubation attempt wherever possible. Data were collected on patient age and sex; indications for intubation; drugs used to facilitate intubation (if any); number and details of each attempt before successful intubation (including the Cormack-Lehane grade¹⁵); grade and specialty of intubating doctor; immediate complications and physiological data before and after intubation (within 15 minutes).

Investigators in each centre checked the resuscitation room log books regularly to ensure that all eligible patients had been included. When a form was not completed prospectively, investigators completed a form retrospectively with reference to the emergency department records and, if necessary, personal contact with the person who performed the procedure was made. Occasionally it was not possible to complete forms retrospectively because of lack of notes or difficulty in identifying the personnel involved in the resuscitation.

Completed forms were returned to the Central Office of the Scottish Trauma Audit Group (STAG)¹⁶ for data entry and analysis using the SPSS database (v.9).¹⁷ The χ² test was used to compare categorical variables. The Mann-Whitney U test was used for non-parametric data. Statistical significance was defined as p<0.05.

As this was an observational study, no restrictions were placed on the drugs used for RSI or any other method for facilitating endotracheal intubation. Participants were free to use whatever technique and drugs they chose for each individual patient. Therefore, no allocation of patients to one specialty or another was made on account of the study and no changes were made to normal emergency airway practice in the participating departments.

Only adult patients on whom RSI was attempted in the emergency department were included in the final analysis. Patients were excluded from the analysis if they met one or more of the following criteria: non-traumatic cardiac arrests (n=593); intubation attempt without drugs (including cardiac arrest following trauma) (n=238); intubation attempt with sedative, anaesthetic or neuromuscular blocking drug only (n=47); inhalational induction of anaesthesia (n=11); or age less than 13 (n=44).

Physiological compromise was defined as the presence of one or more of the following immediately before any attempt at intubation: systolic blood pressure less than 90 mm Hg; Glasgow Coma Score (GCS) <9; respiratory rate less than 10 per minute or greater than 29 per minute; oxygen saturation <90% (measured by pulse oximetry, regardless of inspired oxygen concentration).

RESULTS

A total of 1631 patients met the study entry criteria. Altogether 735 patients fulfilled the criteria for RSI and were therefore eligible for analysis. Table 1 gives patient details for the two groups (anaesthetists and emergency physicians). Table 2 compares the success rates for RSI; grades of laryngoscopy; details of physiological compromise, proportion of patients intubated within 15 minutes of arrival in the emergency department and a summary of the number of patients with immediate complications. Table 3 details the number of immediate complications (not patients) seen in each group. A single surgical airway was required by each specialty for failed intubation after RSI.

There were no significant differences in the pattern of results between those hospitals where emergency physicians regularly intubated and those where they did not. In patients requiring more than one attempt at intubation, the second attempt was usually performed by the same specialty that had initiated the process. However, in 33% (20 of 61) of intubations initiated by emergency physicians, subsequent attempts were performed by anaesthetists and in 7% (2 of 29) of anaesthetist initiated intubations, subsequent attempts were performed by emergency physicians.

It was not possible to accurately identify the exact proportion of forms that were completed retrospectively, but our best estimate is around 90 RSI forms (12%). The majority of these forms were completed within two or three weeks of the RSI occurring.

DISCUSSION

This study has shown that, for RSI in the emergency department, anaesthetists have a higher initial intubation success rate and obtain more grade I and II views at laryngoscopy than emergency physicians. This may be attributable to superior technical ability of anaesthetists in optimally positioning patients and use of intubating equipment or because of the more urgent nature of more of the emergency physician intubations. It is probable that the higher initial success rate is related to the higher proportion of good views obtained at laryngoscopy.

Both anaesthetists and emergency physicians seem to be capable of performing RSI in the emergency department with comparable immediate complication rates. Emergency physicians tended to have a higher proportion of patients who experienced immediate complications compared with anaesthetists. This did not reach statistical significance, but the study is not sufficiently powered to conclude that there is definitely no difference between the groups. The majority of these complications were either recognised oesophageal intubations or hypotensive episodes. The complications overall noted in the study are comparable to another study of intubation in the critically ill.¹⁴ It is apparent from table 1 that emergency physicians and anaesthetists are performing RSI on differing populations of patients and this may affect the immediate complications observed.

This study has also shown that emergency physicians are performing RSI on a higher proportion of patients with physiological compromise and a higher proportion of patients within 15 minutes of arrival in the emergency department. It is probable that these patients are so ill that the emergency physician has considered it necessary to intervene immediately to save life and reduce morbidity. It may also be related to the immediate availability of the emergency physician within the emergency department, while there may also be a variable delay between calling the anaesthetist to the emergency department and their arrival.

Notwithstanding the results of this study it is clear that trainees in anaesthesia and emergency medicine need to receive adequate training in resuscitation including advanced airway techniques; this has to include RSI and the management of the difficult or failed airway. This may entail experience in the emergency department, intensive care unit and operating theatre for both specialties along with dedicated simulator training.

The lack of internationally accepted definitions of complications of RSI means that studies of emergency airway management (in any setting) cannot be compared on an equivalent basis. Standardised definitions would permit national and international comparisons in the future.

In the UK, there is a role for both emergency physicians and anaesthetists in the management of acute airway compromise in the emergency department. Close collaboration on training and service models and delivery is essential.

Contributors

Colin Graham: original idea, literature search, study design, proforma design, prospective and retrospective data collection, on site coordination, interpretation of results, principal investigator, writing the paper, coordination of coauthors. Diana Beard: study design, protocol, proforma design and printing, retrospective data collection, cleaning, analysis and reporting of data, co-ordination of the study, writing of results. Angela Oglesby: data collection, on site coordination, data entry, form re-design, contribution to the paper. Shobhan Thakore: original idea, study design, form design, data collection, on site coordination, subset analysis, contribution to the paper. Jacqueline Beale: form re-design, data collection, on site coordination, data entry. Jocelyn Brittliff: data collection, on site coordination, form re-design. Michael Johnston: study design, proforma design, interpretation of results, revision of paper, project supervision. Dermot McKeown: liaison with the Scottish Board of the Royal College of Anaesthetists, proforma redesign, interpretation of results, reporting on complications, revision of the paper, project supervision. Tim Parke: study design, form design, interpretation of results, revision of paper, project supervision. Colin Graham is the guarantor for the study.