Is post-induction hypotension harmful in traumatic brain injury?

Background

Taumatic brain injury (TBI) is a common presentation for emergency department (ED) and prehospital teams. If we strip it back to basic pathophysiology we can consider two phases of TBI; the primary and the secondary injury. The primary injury is the event that leads to the actual TBI following the transference of energy through brain tissue following the application of external force (impact, deceleration etc.) or through direct injury (penetrating). Prevention of primary injury is largely a public health matter; think health and safety on building sites, seatbelts, cycling helmets and so forth.

So, by the time that medical care is started, a certain amount of damage is done that (currently) cannot be reversed. Medical therapy therefore is focused on the prevention and mitigation of secondary brain injury (SBI). Imagine a cerebral contusion. If left without management it will undoubtedly continue to swell, cause mass effect and cause injury to other brain tissues that were not directly affected by the injury (at least radiographically). Secondary prevention is focused on managing intracranial pressure (ICP) and maximising substrate delivery (O₂, glucose) and clearance (CO₂). In severe TBI we try and achieve this through escalating, as required, through a set of tiered interventions – see Brain Trauma Foundation (BTF) and Seattle International Brain Injury Consensus Conference (SIBICC) guidance. These interventions generally require the patient to be intubated, ventilated and often deeply sedated with ICP and brain tissue oxygen (PbtO₂) monitoring inserted. The rationale is that in many patients with severe TBI the normal mechanisms that allow the brain to regulate its own blood flow (cerebral autoregulation) is either lost or impaired. The implication of this is that cerebral blood flow therefore becomes pressure-dependent. We don’t actually measure cerebral blood flow but instead measure perfusion pressure (CPP). In patients with impaired autoregulation, we should target a CPP of 60-70mmHg on the basis that values below this have been demonstrated to be associated with worse outcomes; similarly, an excessively high CPP may be associated with hyperaemia and worsening of oedema. From these principles it is evident that maintaining an adequate blood pressure from as close to the point of injury as possible is important. Related to this is the delivery and clearance of gases to the brain tissue. The cerebral vasculature is exquisitely sensitive to CO₂ and if left uncontrolled can cause significant vasodilatation which increases intracranial blood volume which increases ICP which decreases CPP. Therefore, early control of CO₂ is also critically important. Similarly, and hopefully obviously, the delivery of O₂ is also critical to avoid ischaemia in vulnerable tissue.

The provision of prehospital emergency anaesthesia (PHEA) is a core component of advanced PH care. The indications are myriad, but clearly a patient with severe TBI has numerous reasons why airway management might be desirable. Early control of gas exchange and supporting an adequate circulation (I’m not going to get into managing the bleeding patient with a TBI today) are essential therapies as outline above. A big question is about this can be best managed given that any rapid sequence induction can be associated with significant hypotension post-induction. To make matters worse the injured brain often triggers a significant sympathetic response as the brain tries to send signals to maximise its own perfusion; anaesthesia may obtund this response, again causing a significant degree of hypotension. Earlier this month a paper was published in JAMA Open (thanks to James Price, Kate Lachowycz, Rob Major et al.) that reviewed this topic in a post-hoc analysis of 555 TBI patients.

You can read the abstract below, but as always read the full paper and come to your own conclusions.

Abstract

Importance Preventing systemic disturbances, such as hypotension and hypoxia, is key to reducing the impact of secondary neuronal injury after traumatic brain injury (TBI). Postintubation hypotension is prevalent and may be associated with worse outcomes in patients with trauma undergoing emergency anesthesia.

Objective To investigate the association between postintubation hypotension and 30-day mortality in patients with severe TBI undergoing prehospital rapid sequence induction.

Design, Setting, and Participants This multicenter, retrospective, observational cohort study was performed between January 1, 2015, and December 31, 2022, in the East of England Trauma Network, including 3 helicopter emergency medical services (East Anglian Air Ambulance, Essex & Herts Air Ambulance, and Magpas Air Ambulance). A consecutive sample of patients (aged ≥16 years) with trauma and severe TBI who received prehospital rapid sequence induction by helicopter emergency medical services and were transported to a hospital within the East of England Trauma Network were eligible for inclusion. Severe TBI was defined as a Head Abbreviated Injury Scale score of 3 or higher. Data analysis was performed from March to May 2025.

Exposure Postintubation hypotension defined as a new systolic blood pressure less than 90 mmHg and induction of anesthesia at 10 minutes or less.

Main Outcomes and Measures The primary outcome was 30-day mortality.

Results A total of 555 patients (median [IQR] age, 48 [29-66] years; 408 [73.5%] male) were included in the final analysis; 548 (98.7%) had a blunt mechanism of injury. Within the first 10 minutes of anesthesia, 106 patients (19.1%) had postintubation hypotension, and 169 (30.5%) died within 30 days of injury (46 of 106 [43.4%] in the hypotension group and 123 of 449 [27.4%] in the nonhypotension group). After adjustment for confounders (eg, age and Glasgow Coma Scale score), postintubation hypotension was associated with increased 30-day mortality for patients with polytrauma and severe TBI (adjusted odds ratio [AOR], 1.70; 95% CI, 1.01-2.86; P = .04). For patients with isolated severe TBI who had postintubation hypotension, the odds of death adjusted for confounders (eg, age, Glasgow Coma Scale score, and Injury Severity Score) were significantly higher than for patients without (AOR, 13.55; 95% CI, 3.65-61.66; P < .001).

Conclusions and Relevance In this cohort study of patients with severe TBI who received prehospital rapid sequence induction, postintubation hypotension was associated with increased 30-day mortality. This association was strongest for patients with isolated TBI. These findings suggest the need for randomized prehospital interventional studies to reduce the incidence of postintubation hypotension in traumatic brain injury.

Tell me more about the methods.

The authors report the results of analysis of a subset of TBI patients from a larger study of PHEA in 1583 trauma patients. This was a retrospective, multicentre, observational cohort study that included patients treated by three UK HEMS teams in the East of England between 2015 and 2022. Whilst different services differ in their standard operating procedures (SOPs) each of the services delivered PHEA with a physician/critical care paramedic combination, and all the services used a standardised drug regimen of ketamine 1-2 mg/kg, rocuronium 1 mg/kg, and fentanyl 0-3 mcg/kg. Patient data was only included if a systolic blood pressure SBP was available before and after PHEA. Hypotension was defined as a SBP <90 mmHg, which is a fairly standard definition (I won’t get into this either). In many ways HEMS in the UK sets the bar for recording physiological data during critical phases of care, thanks to the ability to monitor, record, upload, and make available seemingly limitless data points. Importantly, the authors excluded patients with any pre-induction hypotension in an attempt to isolate the effect of PHEA-induced hypotension.

What were the outcomes?

The authors defined their primary outcome as the association between post-intubation hypotension within 10 minutes or less of induction and 30-day mortality. They observed this in two groups of patients; one including all patients with severe TBI (Abbreviated Injury Scale; AIS ≥3), and one that only included isolated severe TBI (AIS Head ≥3, all other regions <3).

What did they find?

The demographics of the study group are typical of a trauma population – mostly male (73.5%), mostly young-ish (median age 48), almost all blunt mechanism (98.7%). The authors have provided demographics for both the hypotensive and non-hypotensive groups. One might argue that the hypotensive group were more severely injured (Injury Severity Score; ISS median 38 vs. 30 in non-hypotensive) with a lower Glasgow Coma Scale (5 vs. 7) and with a higher AIS in other regions. That being said, the inter quartile ranges overlapped for all of these variables making it difficult to claim a definite difference.

With regards to the primary outcome, post-intubation hypotension was associated with increased mortality in the polytrauma and severe TBI group (adjusted odds ratio; AOR 1.70, CI 1.01 – 2.86). In isolated severe TBI the effect was even more pronounced with an AOR of 13.55 (CI 3.65 – 61.66). The authors also nicely presented the AORs by SBP threshold for each group. In both analyses there appears to be significant effect on mortality at pressures below 100 mmHg.

What does this mean in practice?

The authors identify some valid reasons for the increased incidence of hypotension in patients with severe TBI. They note ketamine generally has a favourable pharmacodynamic profile for the provision of PHEA, however the normal mechanism through which ketamine maintains stable haemodynamics is adversely affected in TBI. The reason for this has been postulated as being caused by brain-induced autonomic dysfunction and an inability of ketamine to initiate a sympathomimetic response in these patients. The association of hypotension and ketamine in this group of papers is becoming more widely accepted, see this recent review here. This may then be compounded by varying degrees of hypovolaemia and then the effects of positive pressure ventilation post-intubation.

This is a nice study that adds to our understanding of the effects of hypotension on severe TBI. It has been known for a long time that lower blood pressures are associated with worse outcomes in this group of patients which this study confirms and contributes to identifying the threshold that we should be wary of. Retrospective studies are excellent tools to observe trends in the data and to form hypotheses. The authors rightly acknowledge that their observed effects of hypotension warrant further investigation, a sentiment that is echoed in the BTF guidance.

How robust are the findings?

I like this paper a lot and have some reflections/questions after reading it. I am keen to emphasise that these are not criticisms of the paper, as I fully acknowledge and understand how hard collecting, cleaning, and analysing this data can be, let alone the peer review process for publication. Perhaps I could phrase my thoughts as items that I would hope a future study might answer.

It is not clear from the data in this paper how long patients were hypotensive for. The primary outcome was a new SBP <90 mmHg at 10 minutes or less of induction. I would want to know the difference and magnitude of effect of a single episode of hypotension vs a sustained episode (seconds vs minutes). In haemorrhage we can consider the ‘dose of shock’ (time spent with a blood pressure at a sub-perfusing value) as a predictor of organ failure/mortality. Can we assume the same from this study; it is almost certain that we can as it would seem biologically implausible that a seconds-long episode of hypotension would have such a magnitude of effect on a 30-day outcome. Put differently is there a ‘time x pressure’ value that equates to worse outcomes. The other data I would like to see is the percentage who had blood pressure measured with an intra-arterial catheter vs. an oscillometric method. Vasopressors were only administered in a third of the hypotensive group, it would be interesting to know the SOP with regards to managing post-intubation hypotension. It would also be useful to know the vasopressors used and how many of the patients received vasopressors proactively or reactively.

The bottom line

This paper highlights once more that we need to consider airway management in both anatomical and physiological terms. Patients with severe TBI may have elements of difficult anatomy, but they are most certainly physiologically challenging, especially when one considers that perfusion and avoidance of hypoxia and hypercapnia are critical components of TBI care. Pressure is important, but it is perfusion pressure rather than systolic pressure that we probably actually care about. This will differ from individual to individual, however it is challenging to set individualised targets in the ICU let alone pre-hospital. Technology such as the pressure regulation index (PRx) are gaining momentum in the ICU to identify the optimal perfusion pressure (CPP_opt) for TBI patients, however this requires an ICP monitor and, crucially, avoiding things that will affect the ICP/CPP such as rolling, coughing, and presumably the effects of flying on CSF fluid dynamics. We currently do not have sensitive enough tools in the pre-hospital phase to be able to accurately measure ICP or CPP and therefore have to rely on robust physiological targets.

In the case of severe TBI, I would argue that PHEA is vital and that it should be performed as early as possible to gain ventilatory control. Given the relationship between hypotension and outcome, it would seem prudent to figure out how best to avoid it; whether with fluids or vasopressors. It appears that ketamine is associated with hypotension in this group of patients and given its ubiquitous use pre-hospital would warrant a consideration about how best to anticipate and mitigate this effect. I would still rather receive ketamine than propofol for induction pre-hospital, and I daren’t discuss etomidate or thiopentone in this post! Thanks once again to the whole authorship as it is a really valuable piece in the puzzle.

Cheers

Rich