In a retrospective analysis of 274 surviving war‑wounded soldiers from Afghanistan, ketamine administration was not independently associated with development of acute or post‑traumatic stress disorder; multivariable regression identified only acute stress disorder and total number of surgical procedures as independent predictors of PTSD. Although a higher proportion of PTSD cases had received ketamine on unadjusted analysis, injury severity and other covariables accounted for that association.
Summary The objective of this study was to explore whether ketamine prevents or exacerbates acute or post‐traumatic stress disorders in military trauma patients. We conducted a retrospective study of a database from the French Military Health Service, including all soldiers surviving a war injury in Afghanistan (2010–2012). The diagnosis of post‐traumatic stress disorder was made by a psychiatrist and patients were analysed according to the presence or absence of this condition. Analysis included the following covariables: age; sex; acute stress disorder; blast injury; associated fatality; brain injury; traumatic amputation; Glasgow coma scale; injury severity score; administered drugs; number of surgical procedures; physical, neurosensory or aesthetic sequelae; and the development chronic pain. Covariables related to post‐traumatic and acute stress disorders with a p ≤ 0.10 were included in a multivariable logistic regression model. The data from 450 soldiers were identified; 399 survived, of which 274 were analysed. Among these, 98 (36%) suffered from post‐traumatic stress disorder and 89 (32%) had received ketamine. Fifty‐four patients (55%) in the post‐traumatic stress disorder group received ketamine vs. 35 (20%) in the no PTSD group (p < 0.001). The 89 injured soldiers who received ketamine had a median ( IQR [range]) injury severity score of 5 (3–13 [1–26]) vs. 3 (2–4 [1–6] in the 185 patients who did not (p < 0.001). At multivariable analysis, only acute stress disorder and total number of surgical procedures were independently associated with the development of post‐traumatic stress disorder. In this retrospective study, ketamine administration was not a risk factor for the development of post‐traumatic stress disorder in the military trauma setting.
Papers cited by this study that are also in Blossom
Mcghee, L. L., Maani, C. V., Garza, T. H. et al. · Journal of Trauma Injury Infection and Critical Care (2008)
Berman, R. M., Cappiello, A., Anand, A. et al. · Biological Psychiatry (2000)
Ketamine is widely used for pre-hospital anaesthesia in combat settings, but its effects on subsequent psychological sequelae of trauma are unclear. Acute stress disorder (ASD) and post‑traumatic stress disorder (PTSD) have distinct diagnostic windows and symptom clusters under DSM‑5: ASD occurs within days to a month after trauma and is a risk factor for PTSD, which is typically diagnosed 3–6 months later. Previous clinical reports have been conflicting: some observational work suggested peri-operative ketamine might reduce PTSD in burned soldiers, other analyses found no effect, and a randomised trial has reported ketamine can treat chronic PTSD symptoms. Mechanistically, ketamine has both psychomimetic and rapid antidepressant effects, and preclinical studies propose multiple neurobiological pathways by which it could either protect against or exacerbate stress‑related disorders. Mion and colleagues set out to examine whether ketamine administration to combat‑injured soldiers was associated with the subsequent development of ASD or PTSD. The primary objective was to assess the association between ketamine given to the war wounded and later PTSD; the secondary objective was to evaluate the relationship between ketamine and the incidence of ASD.
This was a retrospective, observational cohort study using a database compiled from French Military Health Service (FMHS) sources covering injuries sustained between 1 January 2010 and 31 December 2012 during the French military intervention in Afghanistan. The sampling frame comprised all soldiers who survived injuries from firearms, explosives or military transport accidents. A 780‑item database created by one of the authors from cross‑checked FMHS records supplied the data used for analysis. Diagnoses of PTSD followed FMHS procedures: soldiers completed a PTSD checklist (PCL‑S) 3–6 months after returning from mission, underwent a medical interview, and a psychiatrist made the final diagnosis. The investigators selected a prespecified set of covariables from the database to explore factors associated with ASD and PTSD. These included ASD presence, blast injury, associated fatality among the unit, injury severity score (ISS), head injury (ranging from scalp lesion to brain trauma), initial Glasgow Coma Scale (GCS), traumatic amputation, administered drugs (ketamine pre‑hospital or at field hospital with doses where available; morphine and midazolam at battlefield and field hospital), number of surgical procedures from field hospital to metropolitan hospital, physical/neurosensory/aesthetic sequelae, and chronic pain. Patients were classified into PTSD and no‑PTSD groups based on psychiatrist diagnosis; ASD status was also recorded. Bivariate comparisons used Student’s t‑test or Mann–Whitney test for quantitative variables, Chi‑square for categorical variables, and Kruskal–Wallis for groupings by ISS. Tests were two‑tailed with p < 0.05 considered significant; Bonferroni correction was not applied because multivariable analysis was planned. Covariates with p ≤ 0.10 on univariate comparison were entered into a multivariable logistic regression to identify independent associations with PTSD. Two regression approaches were used: a full model including all covariates and a backward stepwise model removing non‑significant variables. To reduce collinearity, ketamine and morphine were considered as global administrations (battlefield plus field hospital) in the multivariate models. Statistical analyses were performed with StatEL software.
The initial database contained 450 patients; 51 died, leaving 399 survivors, of whom 274 were analysed (two women, 0.7%). Age was available for 257 patients with mean ages of 30 years in the PTSD group and 29 years in the no‑PTSD group. Overall, 89 patients (32%) received ketamine, 98 (36%) were diagnosed with PTSD, and 49 (18%) had ASD. Patients who received ketamine had higher injury severity: median ISS was 5 (IQR 3–13) in the 89 who received ketamine versus 3 (IQR 2–4) in the 185 who did not (p < 0.001). ISS correlated positively with the number of surgical procedures (p < 0.001). In bivariate comparisons, PTSD cases were more severely injured, had higher rates of head injury and amputation, more chronic pain and physical/neurosensory/aesthetic sequelae, and underwent a substantially greater number of surgical procedures (the number was described as nearly six‑times higher than in the no‑PTSD group). Medication exposure differed by group: soldiers in the PTSD group received roughly twice as much morphine, three times as much ketamine and six times as much midazolam as those without PTSD. Ketamine exposure was more common among PTSD cases: 54 of 98 patients with PTSD (55%) received ketamine versus 35 of 176 without PTSD (20%); this difference was statistically significant (p < 0.001). The pattern held for ketamine given on the battlefield (27/98 [28%] vs 14/176 [8%]; p < 0.001) and at the field hospital (44/98 [45%] vs 26/176 [15%]; p < 0.001); some patients received ketamine at both time points. Among those who received ketamine, 54 of 89 (61%) developed PTSD compared with 24% of the 185 patients who did not receive ketamine (p < 0.001). Ketamine doses were incompletely recorded: battlefield doses were available for 193 patients, field hospital doses for 63, and both for 50. For the subset with dose data, mean ketamine dose did not differ between PTSD and no‑PTSD patients (173 (SD 144) mg vs 209 (SD 152) mg; p = 0.42). In multivariable logistic regression including covariates with p ≤ 0.10, ketamine administration was not an independent predictor of PTSD. The two variables that remained independently associated with PTSD were presence of ASD and total number of surgical procedures. In the stepwise model ASD had an odds ratio (OR) 16.5 (95% CI 6.5–41.8; p < 0.001) and number of surgical procedures had OR 1.4 per procedure (95% CI 1.2–1.6; p < 0.001). ASD itself predicted subsequent PTSD with 42% sensitivity and 95% specificity. There was no independent association between ketamine and ASD development; among 50 patients with simultaneous battlefield and field hospital dose data there was no significant difference in mean ketamine dose between those who developed ASD and those who did not (226 (SD 171) mg vs 175 (SD 139) mg; p = 0.34).
Mion and colleagues interpret their findings to indicate that, in this retrospective cohort of 274 war‑wounded soldiers with 36% PTSD prevalence, ketamine administration was not independently associated with either ASD or subsequent PTSD after adjustment for injury‑related covariates. The authors note that unadjusted comparisons showed higher ketamine exposure among PTSD cases, but attribute this to confounding by injury severity: more severely injured soldiers received more ketamine, and injury severity and number of surgical procedures were linked to PTSD risk. The discussion places these results in the context of prior, conflicting literature. Some retrospective studies had suggested ketamine might elicit or be associated with stress symptoms, while other work—including a randomised trial—has found ketamine can reduce PTSD symptoms. The authors review plausible mechanisms by which ketamine could either exacerbate or protect against stress disorders, including NMDA‑receptor antagonism, AMPA receptor effects, modulation of brain‑derived neurotrophic factor, activation of mTOR pathways, and subsequent synaptogenesis and network reconfiguration in prefrontal and cingulate regions. Strengths highlighted include the relatively long post‑event observation period and psychiatrist‑based PTSD diagnosis rather than relying solely on screening checklists. The investigators acknowledge several limitations: the retrospective design, incomplete dose data for many patients, and possible insufficient power for the primary endpoint despite a cohort size comparable with prior studies. They also state a randomised, controlled trial in the war setting would be impractical. The authors conclude that while debate continues, their data suggest ketamine use in traumatic military care does not have a detrimental effect on the later development of acute or chronic stress disorders.
This was an observational, retrospective cohort study covering a 3-year period of the French military intervention in Afghanistan. All soldiers surviving an injury by firearms, explosive or in a military transport accident, were included. All characteristics of soldiers wounded between 1 January 2010 and 31 December 2012 were recorded at the different steps of the medical support. Data were retrospectively extracted from the database (780 items) created by one of the authors (CH) from cross-checked information sources of the French Military Health Service (FMHS): The diagnosis of PTSD was carefully conducted. The FMHS is particularly involved with the identification, prevention, treatment and reparation of the psychological trauma in wounded soldiers. Tracing soldiers' exposure to risky situations is considered essential. Circumstances are noted in the soldier's medical record and in a register of the combat unit (Registre des constatations). Cases are reported for epidemiological monitoring with F5 cards. Between three and six months after return from a war mission, the soldier completes a PTSD checklist (PCL-S) for first screening. A complete, personal clinical interview is then undertaken under the supervision of a military physician. A psychiatrist always makes the final diagnosis of PSTD. Analysis of the scientific literature allowed the selection of relevant covariables among the 780 basic items: 1 ASD; 2 explosion (blast); 3 associated fatality (at least one killed among the wounded); 4 injury severity score (ISS); 5 head injury (from scalp lesion to brain trauma); 6 initial Glasgow coma scale (GCS); 7 traumatic amputation; 8 administered drugs: 8.1 ketamine in the pre-hospital settingor at the field hospital with, whenever possible, the delivered dose (mg); 8.2 morphine at battlefield: subcutaneous (s.c.) or intravenous (i.v.) and at field hospital; 8.3 midazolam at battlefield and field hospital; 9 number of surgical procedures undergone from field hospital to metropolitan hospital; 10 physical, neurosensory or aesthetic sequelae; and 11 chronic pain. All the individuals in whom a psychiatrist had made a diagnosis of PTSD were included in the PTSD group, with all other patients analysed in the no PTSD group. For quantitative variables, groups were compared with Student t-test or Mann-Whitney test. Binary variables were compared by a Chi-square test. Groups analysed according to ISS were compared with a Kruskal-Wallis test. All tests were two-tailed with a p < 0.05 considered as significant. Because a multivariate analysis had been planned, the Bonferroni correction was not applied. All covariates for which comparison provided a p ≤ 0.10 were then included in a multivariable logistic regression analysis in order to detect which among them would be independently linked to the development of PTSD. No cut-off value was used and covariates were introduced either as numerical values or as dichotomic covariates. Significance was estimated with the Wald test. Two explanatory models were tested, one incorporating all covariates and a stepwise model in which non-significant variables were removed step by step from the model (backward approach). Statistical analyses were performed with the StatEL software (ad Science, Paris, France).
The database comprised of 450 patients. Fifty-one died and of the remaining 399 patients, 274 patients (only two women, 0.7%) were analysed. Age was known for 257 patients, mean (SD) 30years in the PTSD group and 29years in the no PTSD group. Overall, 89 (32%) patients received ketamine, 98 (36%) were diagnosed with PTSD and 49 with ASD (18%) (Table). The 89 injured soldiers who received ketamine had a median (IQR [range]) ISS of 5 (3-13 ) vs. 3 (2-4) in the 185 patients who did not (p < 0.001). There was a greater number of surgical procedures undertaken in those soldiers who were most severely injured (Fig.). Injury severity score and the number of surgical procedures were positively correlated (p < 0.001). Tablesummarises the comparison between PTSD and no PTSD groups for the 16 covariates. Age, GCS, blast injury and associated fatality were not significantly related to PTSD. Soldiers in the PTSD group were more severely wounded, had a higher incidence of head injury and/or amputation, and were more likely to develop chronic pain, physical, neurosensory or aesthetic sequelae. The number of the surgical procedures they underwent was nearly six-times higher than in the no PTSD group. Soldiers in the PTSD group received twice as much morphine, three times as much ketamine and six times as much midazolam than patients in the no PTSD group. Fifty-four patients (55%) in the PTSD group received ketamine, which was 2.8 times more frequent than in the no PTSD group (35 patients (20%); p < 0.001). This result was similar whether ketamine was administered on the battlefield (27 (28%) in the PTSD group vs. 14 (8%) in the no PTSD group; p < 0.001) or at the field hospital (44 (45%) in the PTSD group vs. 26 (15%) in the no PTSD group; p < 0.001). It should be noted that several patients received ketamine twice. Thus, among the war wounded who received ketamine, 54 out of 89 (61%) developed PTSD, whereas it was only 24% among the 185 patients who did not (p < 0.001). The doses of ketamine administered (Fig.) were recorded for 193 patients on the battlefield and 63 at the field hospital, but simultaneously in only 50 patients. There was no difference between the mean (SD) doses of ketamine administered to the 32 patients in the PTSD group and the 18 patients in the no PTSD group (173 (144) mg vs. 209 (152) mg, respectively; p = 0.42). Covariates linked to the development of PTSD (p < 0.1) were introduced into a logistic regression model. To avoid colinearity, only global administration of ketamine and morphine was considered (battlefield and field hospital). In the analysis with all significant covariates, the only independent associations with PTSD development were the number of surgical procedures and development of ASD (Table). Administration of ketamine was not independently related to PTSD development. In the stepwise model, the same two covariates remained independently associated with PTSD development: ASD odds ratio (OR) (95%CI) 16.5 (6.5-41.8) (p < 0.001); and number of surgical procedures OR (95%CI) 1.4 (1.2-1.6) (p < 0.001) (Table). Forty-one among 98 patients (42%) in the PTSD group experienced an ASD vs. only 8 (5%) among 176 in the no PTSD group (p < 0.001). Thus, an ASD predicted subsequent PTSD with 42% sensitivity and 95% specificity. Among the 50 patients who developed ASD for whom ketamine dose was simultaneously known for battlefield and field hospital, there was no difference between the mean (SD) ketamine doses given to the 11 ASD and 39 no ASD patients (226 (171) mg vs. 175 (139) mg, respectively (p = 0.34). The only ASDrelated covariate was the development of PTSD. There was no independent association between ketamine administration and ASD development (Table).
Medical practitioners in the French army commonly administer ketamine to injured patients on the battlefield, during surgical procedures and for postoperative analgesia. In this retrospective cohort study of 274 war-wounded soldiers with a prevalence of PTSD of 36%, ketamine administration was not independently associated with the occurrence of either ASD or subsequent PTSD. Regarding the effects of ketamine in relation to psychological reactions to stress, scarce clinical evidence leads to conflicting results. Ketamine administered during stressful events may aggravate, preventor have no effecton the subsequent development of PTSD. Since the first clinical experiments, the psychomimetic effect of ketamine has been The table includes ketamine and morphine administration on the battlefield and in the field hospital. When the model was tested with ketamine and morphine administered separately on battlefield and in the field hospital, neither ketamine nor morphine were independently linked to post-traumatic stress disorder occurrence. considered to be a problem. In 1999, involvement of hyperglutamatergic states in the context of traumatic stress had been suspected; N-methyl-D-aspartate receptor antagonists (NMDA-R), which stimulate corticolimbic glutamate release, could be responsible for the dissociative symptoms associated with PTSD. In a retrospective study of 56 trauma victims, Sch€ onenberg et al. suggested that ketamine could have elicited PTSD symptomsand later found a statistical link between ketamine and ASD, but without multivariate analysis. However, other studies have not validated these assumptions. Although ketamine may elicit dissociative symptoms, it also possesses an antidepressant activityand, contrary to other antidepressants, acts within only a few hours. It has thus been proposed as a fast antidepressant for treatment-resistant depression, particularly in patients at high-risk of suicide. Studies have demonstrated that besides its amnesic, anxiolyticand antidepressant properties, ketamine is able to alleviate the symptoms of PTSD in both animalsand humans. McGhee et al. found that ketamine may protect burned soldiers from PTSDalthough this was not confirmed in a later study involving a larger cohort. Subanaesthetic doses of ketamine have not been shown to worsen dissociative symptoms in subjects with PTSDand a randomised, controlled trial demonstrated that ketamine was superior to midazolam for reducing PTSD symptoms. Moreover, recent case reports showed that ketamine, or other NMDA-R antagonists, far from exacerbating stress disorders, dramatically reduced them, at least transiently. Experimentally, when given immediately after a psychological stressor, ketamine prevented synaptic transmission from prefrontal cortex to amygdala. Several mechanisms have been advocated; the most interesting among them may be NMDA-R inhibitionwith glutamatergic stimulation of a-amino-3hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors located upon GABAergic interneurons, up-regulation of brain-derived neurotrophic factoror intracellular targets such as activation of mammalian target of rapamycin pathway (m-TOR) and subsequent synaptogenesis. Thus, ketamine modulates neurotransmitter concentration, synaptic activity, cerebral plasticity and reconfiguration of cortical networks, especially in the anterior cingulate and prefrontal cortices. By restoring the correct balance between detrimental and protective inputs from the amygdala to the medial prefrontal cortex, 'prophylactic' Injury severity score and number of surgical procedures were included as numeric values for each patient. Drug administration has been considered only for initial forward step of management (on battlefield). ketamine may be able to elicit an actual 'resilience to stress'and prevent subsequent PTSD. Concerning the real-life setting, few series have been published. The main strength of our study is a long post-event observation period. The diagnosis was assessed at least 3 months after the traumatic event and consolidated with a 2-year observation period. Using these strict criteria, we found a PTSD incidence of 35.8%, which is 10% higher than in the study by McGee et al.. However, the diagnosis of PTSD in our study was made by a psychiatrist rather than a checklist and took place later in relation to the traumatic event; the diagnosis of PTSD is much more likely at 3-6 months following the traumatic event. The relationship between ketamine administration and trauma severity explains the confounding fact that patients in the PTSD group received three times more ketamine than those in the no PTSD group. Indeed, the multivariate analysis showed no significant difference in the incidence of PTSD between soldiers who received ketamine and those who did not. In addition, PTSD development was only related to ASD and total number of surgical procedures; these combine the composite effects of initial trauma severity (associated fatality, ISS, blast injury) and the sequelae of the injury, including the development of chronic pain. There are some limitations to our study. First, like other published studies, our study suffers from its retrospective design. A randomised, controlled trial to decide whether ketamine is able to prevent PTSD in wounded soldiers is obviously impractical in the context of war. Second, although our study consisted of a large cohort, it may not be adequately powered for the primary end-point. However, our cohort is similar in size to other published work; Sch€ onenberg et al. studied fewer patients (n = 56). The two studies by McGhee et al. included 147 and 289 patients, respectively, but included imprecise timing of PTSD diagnosis. Finally, we only obtained a limited number of precise ketamine doses, which may influence the effects of the drug. The use of ketamine use has increased over the last 15 years. Whether it has the potential to be able to prevent PTSD development when used for warrelated injuries will be debated until a randomised, controlled trial is available. For the moment, our study suggests that ketamine administration in the traumatic military setting does not have a detrimental effect in terms of the development of acute and chronic stress disorders.