Comparison of Esketamine and Propofol for Prehospital Emergency Anaesthesia in Patients With Traumatic Brain Injury—A Retrospective Observational Study

ABSTRACT

Background

Prehospital emergency anaesthesia and controlled ventilation are cornerstones in the treatment of traumatic brain injury patients. Currently, there is a lack of consensus regarding the choice of anaesthetic used during the induction of anaesthesia. Esketamine has gained significant popularity, but its use has been limited in patients with traumatic brain injury due to fears of increased intracranial pressure. A protocol for anaesthesia was implemented at a Finnish helicopter emergency medical services (HEMS) unit during 2015 that mandated the use of esketamine over propofol for most patients.

Methods

We performed a retrospective cohort study to evaluate the differences in mortality and physiology in patients with traumatic brain injury, intubated using either propofol or esketamine. We collected data on patients treated by a single HEMS unit in Finland between January 2014 and December 2021. Our primary outcome was mortality before hospital discharge, and our secondary outcome was physiological stability, defined as the frequency of hypotension (systolic blood pressure ≤ 90 mmHg or a decrease of ≥ 10%) after intubation. Controlling for confounders was done through a logistical regression analysis.

Results

We identified a total of 366 patients, 301 of whom were treated with esketamine and 65 with propofol. There was no significant difference in mortality between the esketamine and propofol groups (odds ratio 0.598, 95% confidence interval 0.281–1.272). The decrease in blood pressure after intubation was greater in the propofol group (absolute change −37.3 vs. −12.4 mmHg, 95% confidence interval −38 to −15 mmHg), but because the initial blood pressure was also higher, both groups had similar postintubation physiology.

Conclusion

In this study, we found no significant difference in mortality, but we found significant differences in the haemodynamic responses between esketamine and propofol, with a slightly favourable haemodynamic profile in patients treated with esketamine. The protocol implemented in 2015 heavily shifted the choice of anaesthetics from propofol to esketamine, but there was a distinct resurgence in propofol use during the study period.

Editorial Comment

Patients with severe head injuries can receive advance intensive care including anaesthetic drugs in the field for intubation, when prehospital teams have this competence. This retrospective analysis presents a comparison of outcomes for traumatic brain injury cases who received either propofol or esketamine in the prehospital setting, as part of a time‐interrupted series where drug choice was changed. While possibly differing in circulatory effects, the two drugs were not associated with a difference in mortality before hospital discharge.

1. Introduction

Prehospital emergency anaesthesia (PHEA), facilitated with an anaesthetic and a neuromuscular blocking agent, is a crucial procedure performed on critically ill patients to secure the airway and ensure proper ventilation and oxygenation. In patients with traumatic brain injury (TBI), PHEA allows for controlled normovenvtilation to control intracranial pressure, a cornerstone of TBI management [1]. The safety of the procedure is critical and influenced by many factors, including the choice of drugs for induction of anaesthesia. Safety is further emphasised in the prehospital environment, where swift actions are required in challenging environments, often with insufficient patient information available. Thus, standardised protocols for choosing induction drugs are used in many services. These drugs commonly include ketamine, etomidate, propofol, or midazolam as an anaesthetic often combined with fentanyl [2]. Due to the lack of evidence, there is currently a wide variety in PHEA practices, including the choice of anaesthetics [3, 4].

To maintain adequate cerebral perfusion pressure during and after PHEA, the haemodynamic effects of the anaesthetic are important. Propofol is the most‐used anaesthetic drug, especially in haemodynamically stable patients, while ketamine has recently gained popularity [4, 5, 6]. Etomidate and midazolam, on the other hand, have lost favour among many providers—etomidate due to fears of adrenal dysfunction [7, 8], and midazolam due to its poor haemodynamic properties [9, 10]. Ketamine is regarded as the most stable option haemodynamically, but concerns over its effect on intracranial pressure have reduced the drug's usage in patients with TBI [11]. The recent rapid shifts in the choice of anaesthetic and the disparity in the preferred drugs between different providers signify a lack of consensus and high‐quality evidence on the safety of these drugs [3]. The studies published to date comparing the mortality rates between ketamine and propofol in PHEA are few and have thus far found no definitive answers [12, 13, 14, 15].

A protocol for PHEA was implemented in a single Finnish helicopter emergency medical services (HEMS) unit in 2015, significantly shifting the choice of anaesthetic from propofol to esketamine [16]. This allowed for comparisons of outcomes between the drugs. In this study, we focus on patients with TBI because they may be most prone to the potentially harmful pharmacokinetic and ‐dynamic differences between these drugs. This study aims to compare mortality before university hospital discharge and physiological stability, defined as the frequency of hypotension (systolic blood pressure [SBP] ≤ 90 mmHg or a decrease of ≥ 10%) after intubation in patients with TBI undergoing PHEA with either propofol or esketamine.

Our hypothesis is that esketamine is similar to propofol in outcome when used for PHEA.

2. Methods

2.1. Study Design

This was a retrospective cohort study analysing patients of a single HEMS unit between January 2014 and December 2021. A review by the ethical committee was not necessary as registry‐based retrospective studies do not require one according to Finnish law (Medical Research Act 488/1999, amendments 295/2004, 794/2010). Approval of the study protocol and access to patient records was granted by the Helsinki University Hospital (HUS/305/2022). We followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines in the reporting of the study [17].

2.2. Setting

This study investigated patients treated by a single HEMS unit (FinnHEMS 10) in Finland. Located in the city of Vantaa, the FinnHEMS 10 operates in an area of approximately 10,000 km [2] in southern Finland. It is the busiest HEMS base in Finland, with nearly 1.3 million people in its operating area and around 3000 dispatches each year [18]. Nationwide, there were five other HEMS bases that operated in Finland around the time of the study [19]. The FinnHEMS 10 operates in a three‐person team: a physician, who is usually a senior anaesthesiologist working full‐time in HEMS; an HEMS crew member who is a firefighter or a paramedic; and a pilot with special training in assisting with prehospital critical care.

2.3. The PHEA Procedure

During the study period, there was significant development in the way PHEA was performed in the unit. Until 2014, the process was not standardised but rather was dictated purely by the physician on call. This changed in 2015 when a new PHEA protocol was introduced in an attempt to standardise practices and improve teamwork [20]. The protocol was implemented over a 1‐year period and used throughout the rest of the study period. The drugs available did not change with the introduction of the PHEA protocol and included fentanyl, propofol, esketamine, midazolam, and rocuronium. Their respective proportions, however, changed drastically. Before the implementation, propofol was the preferred anaesthetic for nearly all patients, while the new protocol mandated the use of esketamine (1 mg.kg¯¹) as the standard choice for most patients. Additionally, administration of rocuronium (1 mg.kg¯¹) was made mandatory, and fentanyl was to be given 2–3 min before induction if increased intracranial pressure was suspected. According to the new protocol, propofol should be used for induction only in special indications—namely, for patients with status epilepticus, or at the discretion of the physician for a patient with isolated TBI and particularly high blood pressure [20]. The new protocol also mandated the routine use of a C‐MAC video laryngoscope with a bougie as an adjunct, whereas before the implementation, physicians were able to choose between these and a traditional Macintosh laryngoscope and a preinserted stylet. It was already known at the time of the study that the protocol implementation also increased the success rate of the procedure, but based on a previous study [21] showing that improved first‐pass success rate does not seem to affect mortality, we deemed that this would not create problems for interpreting our results. During the study period, the HEMS unit used norepinephrine (infusion and boluses) and phenylephrine (prefilled syringe for boluses) for prevention and treatment of hypotension after induction of anaesthesia.

2.4. Participants

We gathered data on all patients from the study period who underwent drug‐facilitated intubation after having experienced a TBI. We defined this cohort as all patients who were diagnosed with any of the TBI‐related ICD‐10 codes S06.0–S06.9 during their hospital stay. We excluded patients who were intubated by providers other than HEMS, those intubated during cardiac arrest, and those intubated using none or a combination of both anaesthetics.

2.5. Data Source

We extracted data from the Finnish HEMS database [19], an electronic database used to track all Finnish HEMS missions. The details of the missions are submitted by the physician or the paramedic on call; these include data on airway management that is recorded according to an international standardised template [22]. Additional data, including hospital stay diagnoses, dosing of induction drugs, and arterial blood gas results, were collected through manual chart review of both prehospital and in‐hospital electronic patient records. These additional data are primarily collected by junior researchers not working in the clinical operations in the HEMS. In cases of unclear interpretation, a senior researcher was consulted.

2.6. Variables

The collected data included baseline characteristics like age, sex, patient category, and vital signs recorded when the HEMS unit arrived on scene. We further gathered data on the proportion and dosage of drugs given to patients for PHEA as well as the success rate of intubation and vital signs after airway management. We lastly collected the ICD‐10 codes assigned to the patients during their hospital stay and determined whether the patients had died before discharge from the university hospital. The primary outcome was death before discharge from the university hospital. This was mainly used for practical reasons related to electronic patient records as this was the point until which we reliably had data on all patients. For our secondary outcome, we chose postintubation physiological stability because it has been shown to influence mortality [23]. Physiological stability was defined as the frequency of hypotension (SBP < 90 mmHg or decrease of ≥ 10%) occurring after intubation. This definition followed the international consensus on quality indicators of prehospital anaesthesia [24]. We also included other definitions of haemodynamic changes to adequately report the effects of the drugs (Data S1). To correct for possible differences in the subgroups regarding the primary outcome, we controlled for the effects of age, sex, injury severity, vital signs, and Glasgow Coma Score (GCS) at HEMS arrival in our statistical analysis. These parameters were chosen based on a combination of literature review and expert opinion [25, 26]. The injury severity was accounted for by calculating the ICD‐10 code‐based injury severity score (ICISS) [27]. We used all of the ‘Injury, poisoning and certain other consequences of external causes (S00‐T98)’ ICD‐10 codes when calculating the ICISS score. In patients with multiple diagnoses in the ‘Intracranial injury (S06)’ category, all of the diagnoses were included in the calculation.

2.7. Statistical Methods

Baseline characteristics, operational data, and demographics are presented separately for both drug groups as well as for the total cohort. Categorical data are presented as numbers and percentages and compared using a chi‐square or Fisher's exact test. Continuous data are presented as means with standard deviation or median and interquartile range, depending on the distribution of the data, and comparisons of continuous data are done using an independent t test as well as the Mann–Whitney U test. Distribution was assessed visually with the help of histograms. A p < 0.05 was considered statistically significant. For the mortality analysis, we constructed a multivariable logistic regression model to evaluate the association between anaesthetic (esketamine versus propofol, propofol as reference) and in‐hospital mortality. No stepwise methods were used, and all covariates were entered into the model simultaneously. Missing data regarding either any covariate or outcome resulted in the case being excluded from the final analysis. The overall accuracy of the model was recorded. The goodness of fit of the model was assessed using the Hosmer and Lemeshow test, and evaluation of outliers using Cook's distance and leverage values. Multicollinearity was assessed using variance inflation factor (VIF) and tolerance levels. To estimate the degree of uncertainty in our main outcomes, we calculated the Hodges–Lehmann median difference with a 95% confidence interval or Agresti–Caffo mean difference with 95% confidence intervals depending on the nature of the outcomes, Hodges–Lehmann for continuous variables and Agresti–Caffo for categorical variables. To assess for possible differences between patients treated either before or after the protocol implementation, we performed sensitivity analysis comparing the groups regarding key variables. All analyses were performed using SPSS Statistics for Mac, version 28 (IBM Corp., Armonk, NY, USA).

3. Results

3.1. Study Size and Baseline Characteristics

During the study period, the HEMS unit provided PHEA for a total of 2203 patients. Of these, 366 patients were included in the study: 65 patients in the propofol group and 301 in the esketamine group. Propofol was used in 25 (38.5%) patients before or during, and 40 (61.5%) patients after implementation of the PHEA protocol. Patient selection has been visualised as a flowchart in Figure 1.

FIGURE 1.

Flowchart of patient selection. PHEA, prehospital emergency anaesthesia.

The comparison of baseline characteristics is presented in Table 1 and the breakdown of TBI‐related ICD codes in Data S1. Initial GCS and total ICISS were similar in both groups. However, there were some significant differences between the groups. The patients in the propofol group were older (59 vs. 47 years) and more frequently female (31% vs. 18%). Furthermore, they were also characterised by a higher initial blood pressure (161 mmHg vs. 138 mmHg). In the sensitivity analysis, the patients treated prior to the PHEA protocol were older and had a lower FPS rate, but all other characteristics, including mortality, were similar compared to the patients treated after the implementation of the PHEA protocol.

TABLE 1.

Baseline characteristics and preintubation vitals of patients intubated after traumatic brain injury. Categorical data are presented as n (%) and continuous data as mean (SD) or as median (25th–75th percentiles).

Variable	n (missing)
Propofol	Missing	Esketamine	Missing	p
Age, years	366 (0)	59 (41–76)	0	47 (30–62)	0	< 0.001
Heart rate	358 (8)	92.47 (28.84)	1 (1.5)	92 (26)	7 (2.3)	0.88
Systolic blood pressure, mmHg	350 (16)	161 (38)	4 (6.1)	138 (35)	12 (4.0)	< 0.001
Oxygen saturation, %	346 (20)	96 (93–98)	5 (7.7)	97 (93–99)	15 (5.0)	0.14
GCS	365 (1)	5 (3–7)	1 (1.5)	5 (3–8)	0	0.58
ICISS	366 (0)	0.83 (0.71–0.84)	0	0.76 (0.66–0.84)	0	0.12
Sex, male	364 (2)	45 (69)	0	245 (82)	2 (0.7)	0.027

Abbreviations: GCS, Glasgow coma scale; ICISS, ICD‐10 code‐based injury severity score.

3.2. Prehospital Treatment

All parameters of prehospital treatment are presented in Table 2. The choice of anaesthetic changed significantly through the study period as propofol use fell dramatically after protocol implementation in 2015 but has since slowly increased up until 2021. This annual change in the proportional use of drugs is visualised in Figure 2. There was a statistically significant difference in the proportion of fentanyl used as premedication, with the propofol group receiving it more often and in higher doses. Patients in the propofol group were also more likely to receive vasoactive drugs during the prehospital care. The first‐pass success rate was significantly higher in the esketamine group.

TABLE 2.

Parameters of treatment in prehospital emergency anaesthesia and before hospital handover. Categorical data are presented as n (%) and continuous data as mean (SD).

Variable	Total sample
Propofol	Missing	Esketamine	Missing	p
Propofol dose, mg	65	72 (30)	—	0 (0)	—	—
Esketamine dose, mg	301	0 (0)	—	70 (21)	—	—
Rocuronium dose, mg	360	80 (50–100)	2 (3.1)	100 (80–100)	4 (0.1)	< 0.001
Fentanyl dose, mg	304	0.25 (0.095)	2 (3.1)	0.225 (0.085)	60 (20)	0.017
On‐scene time, min	361	27 (13)	1 (1.5)	24 (10.68)	4 (1.3)	0.69
FPS	366	59 (92)	0	298 (99)	0	0.005
Use of rocuronium	360	63 (97)	2 (3.1)	297 (99)	4 (0.1)	0.29
Use of fentanyl	304	63 (97)	2 (3.1)	241 (80)	60 (20)	< 0.001
Use of vasoactives	261	21 (32)	8 (12)	26 (8.6)	97 (32.2)	< 0.001
Invasive blood pressure monitoring	366	13 (20)	0	69 (22.9)	0	0.61

Abbreviation: FPS, first pass success.

FIGURE 2.

Yearly proportional use of propofol and esketamine in patients with traumatic brain injury undergoing prehospital emergency anaesthesia. The standard operating procedure for prehospital emergency anaesthesia was implemented during 2015.

3.3. Mortality

There was no statistically significant difference in mortality in the logistic regression model (confidence interval 0.281–1.272 when using propofol as reference). The unadjusted mortality rate was significantly lower in the esketamine group (28.9% vs. 43.8%, p = 0.021, mean difference 14.8% with 95% CI of 1.8% to 27.9%). The results of the logistic regression analysis are presented in Table 3. A total of 331 (90.4%) cases had complete data on all covariates and mortality and were included in the final model. The model had an accuracy of 78.9% with a Hosmer and Lemeshow test p value of 0.360 that indicated a good fit. Analysis of leverage value, Cook's influence, and standardised residuals revealed some outliers; however, in sensitivity analysis excluding these cases, the findings were the same. Tolerance and VIF indicated a low possibility of multicollinearity in the model.

TABLE 3.

Logistical regression analysis of different variables affecting mortality before university hospital discharge.

Variable	Odds ratio	95% confidence interval	p
Esketamine, propofol as reference	0.598	0.281–1.272	0.182
Age	1.055	1.037–1.074	< 0.001
Sex	1.941	0.968–3.889	0.62
Total ICISS	0.003	0.000–0.41	< 0.001
Vital signs before induction
Pulse	0.995	0.984–1.007	0.435
Blood pressure	1	0.991–1.008	0.927
Oxygen saturation	0.952	0.920–0.985	0.005
GCS	0.727	0.642–0.823	< 0.001

Abbreviations: GCS, Glasgow coma scale; ICISS, ICD‐10 code‐based injury severity score.

3.4. Physiological Stability

The changes in SBP are presented in Figure 3. The decrease in blood pressure after intubation was significantly larger in the propofol group (absolute change −37.3 vs. −12.4 mmHg, median difference −26 mmHg with 95% CI of −38 to −15). All vital signs were similar between the two groups in the postintubation measurements. Vital signs were still similar at hospital handover, except for the SBP, which was higher in the propofol group. In the propofol group, there were significantly more decreases in blood pressure of 10% and 20% (67.8% vs. 40.9%, and 54.2% vs. 27.8%, respectively, mean differences 26.9% and 26.5% with 95% CI of 13.1%–39.3% and 12.6%–39.7%). Further, there were no significant differences in the number of increases in blood pressure of either 10% or 20% as well as no significant differences in the amount of new hypotension (defined as postintubation SBP ≤ 90 mmHg with preintubation SBP ≥ 90 mmHg) or new hypertension (defined as postintubation SBP ≥ 160 mmHg with preintubation SBP ≤ 160 mmHg). All postintubation vitals and vitals at handover are presented in Data S2.

FIGURE 3.

Systolic blood pressure during different phases of the prehospital treatment. The line charts represent mean values. The error bars correspond to the 95% prediction interval for systolic blood pressure of individual patients.

4. Discussion

4.1. Main Findings

This study showed no differences in adjusted mortality, but a more stable physiology with esketamine when compared to propofol during PHEA of patients with TBI. However, the rate of hypotension postintubation was similar in both groups. The difference was probably caused both by the pharmacodynamic differences between the agents as well as by the fact that the propofol group had a higher baseline blood pressure, stemming from the new protocol recommending propofol for severely hypertensive patients. There have been concerns with the drastic change from propofol to esketamine and whether the use of esketamine in patients with TBI might worsen outcomes. According to this study, these concerns might be unwarranted.

4.2. Anaesthetics

The protocol implemented in the study unit heavily shifted the choice of anaesthetics from propofol to esketamine. This new protocol is similar regarding the use of ketamine/esketamine to those implemented in other HEMS systems [28, 29]. Despite this trend, there have been concerns over the choice of anaesthetics. The cautious attitude toward the use of ketamine could explain why the drug, which was initially used in almost all cases of TBI after implementation, has since lost some popularity and a new increase in the use of propofol has been seen in the studied system. This is in contrast with a British HEMS system, tending to mainly trauma patients; their PHEA protocol allows solely the use of ketamine, regardless of patient pathology [30].

The two drugs have a quite different pharmacodynamic response. Esketamine, acting primarily through noncompetitive blockage of NMDA receptors, is known to increase heart rate and blood pressure through a stimulatory effect on the sympathetic nervous system, making it potentially suitable for unstable patients. Its effects are similar to its racemic version, ketamine, though requiring lower doses. Initially, there were concerns over its potential effect on increasing intracranial pressure [31, 32], which resulted in avoidance in cases of suspected TBI. However, in recent years, several studies [33, 34] have questioned this effect. Ketamine has since gained significant popularity despite these concerns [5, 35]. Propofol, acting primarily through GABA‐mediated depression of the nervous system, has been commonly used in all patient categories. It has been shown to decrease intracranial pressure and therefore might be suitable in patients with suspected TBI and severe hypertension [36]. However, it is a known negative inotrope with vasodilatory properties [37] which creates a risk of postintubation hypotension and thus might make it less desirable to use on the haemodynamically unstable.

4.3. Outcome

In the multivariate model, there was no statistically significant difference in mortality, although a possible trend of esketamine being more beneficial was seen. The lower number of patients treated with propofol might have been a limiting factor, decreasing chances of a significant result. All in all, while our data does not give reason to support the historical hesitance toward using esketamine [38], it also does not give a clear answer to whether there is a difference in mortality between using it over propofol in this patient group. Any further conclusions should be drawn with caution due to limitations in sample size and study design.

4.4. Physiological Stability

Studies regarding blood pressure control of patients with TBI have determined both hypotension and hypertension to be associated with worse outcomes [39], with a potential ‘safe range’ of SBP around 130–180 mmHg [40]. The postintubation SBP was nearly identical between the groups, and while the optimal SBP range is probably individual for each patient, the average SBP of 127 mmHg in our study suggests that many patients were hypotensive. This might be especially true for the esketamine group because their average SBP further decreased during the time from postintubation to hospital handover. This difference could, however, result from the esketamine group having more polytrauma and hence more unstable blood pressure or from the propofol group receiving more vasoactive drugs to support their blood pressure. It is unclear if this excess hypotension in the esketamine group might hide potential benefits regarding mortality. Nonetheless, these results might suggest that the patients in our study could benefit from more active support of haemodynamic stability.

4.5. Strengths and Limitations

A strength of this study is that it has one of the largest published samples of TBI patients undergoing PHEA to date. However, it also has several limitations. First, the low amount of data from before the protocol implementation limits the size of the propofol group and thus the power of the study; the difference in mortality would have to be quite substantial to be statistically significant with this sample size. Second, this was a retrospective study, which makes it susceptible to several biases, especially selection bias, because despite the protocol implementation, the physicians chose anaesthetics according to their assessments of who would benefit from either drug. Third, data on vital signs were available only on three time points for each patient, and thus we were unable to describe the complex changes in haemodynamics after induction of anaesthesia. Furthermore, the data source did not include the data on sedation used during transportation, which may have a significant effect on blood pressure at hand‐over. Fourth, while ICISS is a useful tool in assessing injury severity, its accuracy might be limited in patients with TBI [41]. Finally, using mortality as our primary outcome may not be optimal, as it does not reveal anything about recovery or neurological outcome. Mortality before university hospital discharge also leaves out potential deaths happening after transferral to other hospitals.

5. Conclusion

The implementation of a protocol for PHEA significantly increased the use of esketamine over propofol, but there was a distinct resurgence in propofol use during the later parts of the study period. We conclude that the use of esketamine does not seem to result in a difference in adjusted mortality but seems to result in a slightly favourable haemodynamic profile when used in PHEA of patients with TBI. However, our study has several limitations which might affect our results even despite our efforts to mitigate them. Hence, conclusions regarding differences in mortality could not be drawn. Further studies are justified on the subject, preferably with a randomised design.

Author Contributions

Study design was planned by all authors. J.N. and H.L. prepared all illustrations. The first manuscript draft was prepared by J.L. J.N. and H.L. made important edits to the text. All authors have read, edited and accepted the last version of the manuscript.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Data S1: Breakdown of TBI related ICD‐codes diagnosed during hospital stay. ICD, International Classification of Diseases.

Data S2: The unadjusted mortality, postintubation vitals and vitals at hospital handover of patients intubated after traumatic brain injury. Categorical data are presented as n (%) and continuous data as mean (SD) or as median (25th–75th percentiles) EtCO₂.

Laamanen J., Ljungqvist H., and Nurmi J., “Comparison of Esketamine and Propofol for Prehospital Emergency Anaesthesia in Patients With Traumatic Brain Injury—A Retrospective Observational Study,” Acta Anaesthesiologica Scandinavica 69, no. 10 (2025): e70131, 10.1111/aas.70131.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.