Ketamine versus propofol for rapid sequence induction in trauma patients: a retrospective study

Rapid Sequence Induction (RSI) is used to facilitate emergency tracheal intubation to minimise the risk of pulmonary aspiration of stomach contents [1]. This is highly relevant in the trauma population, as these patients often are critically ill, unstable and non-fasting, and choosing the appropriate premedication will maximise the success and minimise complications [2].

Ketamine and propofol are two commonly used induction agents for RSI, yet they have very different pharmacodynamic profiles. Ketamine acts primarily through non-competitive blockade of N-Methyl-D-Aspartate (NMDA)-receptors [3] and is associated with adverse effects such as hallucinations and nightmares [4, 5]. Previous concerns related to increased myocardial ischemia or increased intracranial pressure [6‐8] have recently been rejected [9‐12]. Ketamine is known to increase heart rate, and blood pressure through a stimulatory effect on the sympathetic nervous system making it potentially suitable for haemodynamically unstable trauma patients [13‐17]. However, similar to propofol it may cause cardiovascular collapse in haemodynamically unstable patients [14]. Propofol acts through a Gamma-AminoButyric Acid (GABA)-mediated depression of the central nervous system but is also known to have negative inotropic effects as well as vasodilation. From this perspective, propofol may not be ideal for haemodynamically unstable trauma patients [5].

As such, it seems that no ideal induction agent exists, and at present there is no consensus on which induction agent should be preferred for RSI in the trauma population [18]. Furthermore, very few studies have compared the available drugs such as ketamine and propofol [19]. Therefore, the objective of this retrospective study was to compare 30-day mortality in trauma patients after RSI facilitated emergency intubation prehospitally or within 30 min after arrival in the trauma centre using either ketamine or propofol. We hypothesised that the use of ketamine would result in a lower 30-day mortality.

In this retrospective cohort study, we found no difference between 30-day mortality for trauma patients intubated with RSI using ketamine compared to propofol. There were statistically significant differences between the ketamine and propofol group regarding systolic blood pressure and GCS score before intubation. However, we did not observe any significant haemodynamic difference between the two agents in terms of SBP after intubation or change in SBP after intubation. No significant difference regarding hospital and ICU length of stay or duration of mechanical ventilation was identified either.

This study has several strengths. First, it is one of the largest studies comparing ketamine to propofol for RSI in the trauma population. Second, in the Capital Region of Denmark other aspects of trauma care such as the use of tranexamic acid, prehospital plasma, and spinal immobilisation are standardised. Both propofol or ketamine are recommended single induction agents in our trauma system, in combination with either suxamethonium, or rocuronium. In addition, an opioid, being either fentanyl or alfentanil, is usually given. However, the choice of anaesthetic technique, dosing and induction agents are based on an individual assessment made by the attending physician, and the retrospective study design does not allow for an analysis of why the particular induction agent was chosen in the specific situation. Third, the follow-up completeness regarding the primary and secondary outcomes was 100% adding reliability to our outcome assessment. However, the retrospective design of this study has several limitations. First, we did not use a strictly standardised drug protocol specifying drug dosage or indications for drug choice, and due to the retrospective design of the study we cannot comment on why the physicians chose one induction agent above the other. These aspects of trauma care did not change significantly in our pre- or in-hospital standard operating procedures for trauma care during the data collection period. Second, the timing of data collection may be inconsistent which made it difficult to describe and analyse all details. Third, in 76 (13.9% of) cases information about the type of induction agent was missing in the medical records. This introduces a bias through the possibility of missing not at random, possibly occurring when treating a critically ill patient without enough time to routinely collect information. Additionally, this study only includes patients who were admitted alive/under resuscitation to the hospital and does not consider the severely traumatised patients who were declared dead in the pre-hospital setting, as data on these patients are not recorded in the local trauma registry at Rigshospitalet which we used for data collection. Also, this study does not provide any data on the long-term outcome of trauma patients making it impossible to evaluate any significant outcomes other than in-hospital outcomes.

Even though the choice of induction agent is truly a small alteration in the trauma care pathway, Baekgaard et al. from 2020 [21] found 30-day mortality rate to be 20% and 30% for ketamine and propofol, respectively. We anticipated the same mortality rates and calculated a total sample size of 586 patients. Yet, our study showed mortality rates of 20.2% and 22.8% for ketamine and propofol, respectively, and we only included 548 patients in the designated period, 228 and 320 for ketamine and propofol, respectively. This introduces risk of type 2 error as illustrated by the wide confidence interval regarding 30-day mortality [0.58–1.66] and we cannot exclude a clinically meaningful difference.

We adjusted our analyses for sex, age, ISS, SBP and GCS before intubation. Yet, several pre-hospital factors with a significant role in 30-day mortality were not registered, e.g., ambulance response time and hospital arrival time, difficult airway predictors etc.

A recent systematic review compared studies where trauma patients were intubated in the prehospital setting or within 1 h of emergency department arrival using RSI with ketamine compared to other agents [19]. Four studies were eligible: one compared ketamine to thiopental and showed no difference in number of blood transfusions [24], and three compared ketamine to etomidate: A randomised controlled trial from 2009 showing no difference in 28-day mortality [25], a cohort study from 2015 showing no difference in hospital mortality [26], and finally an observational study from 2017 showing no difference in number of blood transfusions, hospital length of stay or in-hospital mortality [27]. Yet, these studies have several limitations, as they were rather small and therefore inconclusive. A study from 2020 by Baekgaard et al. [21] made similar findings with no difference in mortality comparing ketamine with other induction agents (propofol, etomidate, or midazolam). Finally, one recent multicentre study from 2021 reported no change in average SBP during RSI with ketamine compared to other agents in the adult trauma population [28].

Today, ketamine is considered safe and effective in the emergency department and the prehospital setting [4, 26, 29], and its use is increasing. Concerns regarding increased myocardial ischemia or increased intracranial pressure from intubation with ketamine have been rejected [9, 10], and it may actually be considered neuroprotective [30]. Yet, a randomised trial of induction agents or a prospective study with clearly documented standardised anaesthetic technique is needed, as no currently available evidence supports a specific agent used in RSI [10, 19]. This warranted trial should standardise the administration and dosage of opioids, choice of neuromuscular blocking agent and the maintenance sedation as potentially confounding factors.

Trauma patients intubated with ketamine did not have a lower 30-day mortality as compared with propofol. Moreover, we did not find a difference in hospital or ICU length of stay or duration of mechanical ventilation between the two induction agents. A randomised set-up is warranted to confirm our findings.