Prehospital intubation has been (arguably!) the most important skill the addition of doctors to the Auckland Westpac Rescue Helicopter has brought to helicopter prehospital care in our region.  While the paramedics in our team are highly skilled and very experienced, none of them are currently RSI qualified. St John Ambulance (the ambulance service provider for most of New Zealand, including Auckland) does have RSI-qualified paramedics, some of whom have worked on the helicopter in past years, but currently the helicopter team cannot perform an RSI without a doctor present. Intubations can be performed by paramedics without drugs however in certain circumstances, usually in cardiac arrest or severely obtunded patients.

With the addition of doctors, we have had a heavy training focus on RSI, including introducing an RSI checklist and actively involving the crewman in the RSI procedure.

This paper, from Prehospital and Emergency Care in 2007, provides an interesting insight into the cognitive processes around prehospital intubation, and the implications for training.

Rasmussen’s ‘Skills-Rules-Knowledge’ framework is a concept that can be applied to many difficult/complex processes. The elements are:

Skills-based processing – task performance that can become ingrained, and performed without conciously thinking about it (for example chest compressions in CPR)

Rules-based processing – when an action or series of actions is executed in response to a certain situation, for example ‘if the patient has no pulse – start CPR’. Rules based processing requires explicit knowledge, and can become difficult if the cues for a certain action are unclear.

Knowledge-based processing – a process that can be applied when rules-based processing fails or is not appropriate. For example, in the setting of cardiac arrest – rules-based processing instructs CPR to begin, knowledge-based processing might allow a decision like “the down-time is over 30 minutes, CPR is inappropriate’.

These three elements are highly inter-related – frequently used knowledge-based and rules-based process may, with experience, become skills-based processes.

With regards to prehospital intubation (paramedic performed), the authors then make the following points:

• although intubation is often taught as a skills-based process, on closer examination it relies heavily on knowledge-based processing
• recognition of the need for intubation should theoretically be rules based, however some of the cues involved (assessment of respiratory effort, degree of compromise etc) are challenging and the interpretation may vary between individuals, and there are few clear ‘rules’ for intubation in the prehospital setting, hence it is a knowledge-based process
• identifying a difficult airway is a knowledge-based process
• selection of intubation technique (position, equipment) are knowledge-based, the only clear ‘rules’ apply to the sizing of paediatric endotracheal tubes
• techniques for assessing whether intubation has been successful all aim to provide a rules-based ‘yes/no’ answer as to whether the tube is in the trachea, but all are fallible under certain circumstances. Confirmation of tube position is therefore also knowledge based

Amongst other conclusions (including the need for developing pre-defined “action rules” for common clinical scenarios; developing training that places intubation in the wider context of airway management; and further examining the cognitive processes around paramedic intubation), the authors also ask whether intubation is too cognitively difficult for a paramedic in the prehospital environment, and whether less complex devices would be more appropriate.

So how does this relate to our service?

Given our three clinician model (doctor, paramedic, medically-trained crewman), plus often other ambulance staff, it would be (hopefully!) difficult to argue that a pre-hospital RSI is too cognitively difficult for us.

Thinking about an RSI in terms of skills-based/rules-based/knowledge-based processes is useful however.

The most common way we practise RSI is currently some sort of ‘RSI drill’, involving a team of three. It includes running through our checklist, following by executing the RSI procedure as a simulation. This most closely represents a skills-based procedure. Individual elements of the sequence (laryngoscopy, passing a tube over the bougie etc) also represent skills-based processes. With our simulation lab now up and running (plus 3 new mannequins) we have the opportunity to practise skills-based elements of RSI as often as we wish.

Based on collective experience however (both from clinical practice and training) the most difficult elements of RSI are likely to be the knowledge-based processes – whether an RSI is indicated in certain situations, whether modifications to a standard RSI are required (for example DSI or neuroprotective RSI), and contingencies in the event of difficulties or failure. Given a distinct lack of evidence in many of these areas decisions rely heavily on clinician judgement and experience.

For me, the most important thing to come out of reading the paper above was the idea that we can and should be generating specific rules-based processes tailored to the clinical scenario we are dealing with. In the setting of an RSI, this is crucial for when things are not going to plan. The NAP4 audit of airway failures from the UK found that one of the main contributing factors to ED airway failures was ‘a failure to plan for failure’.

While we have a standard RSI checklist and a sequence of events to follow in the event of a failed airway, we need to (and hopefully already do) go beyond this to consider patient specific elements (for example ‘this patient’s airway burns means a higher chance of failed laryngoscopy and need for surgical airway, I will declare failure after x seconds’, or ‘I expect this patient to desaturate very fast, actions when saturations reach 90% will be x,y,z..)

Anticipating and voicing patient-specific elements and (using the terminology from the study) actively regarding them as rules-based processes should, provided they are generated PRIOR to the proverbial hitting the fan, provide an extra level of safety and decisiveness whan an RSI is not going to plan. With a failing RSI and an increasingly unwell patient, I suspect we would all find it easier to follow a clear, pre-defined, and overtly stated rule or two (that we and our team had tailored for the specific scenario) rather than trying desperately to apply knowledge-based processing too late and under extremely stressful circumstances…

What ‘rules’ would you generate for intubating this child?

The full text pdf for this article is here (secure area limited to ADHB staff – ADHB has subscription access to this journal via the Philson Library at the University of Auckland School of Medicine)