Initial restrictive versus liberal oxygen for traumatic brain injury: A follow-up study of the TRAUMOX2 randomized trial

Initial restrictive versus liberal oxygen for traumatic brain injury: A follow-up study of the TRAUMOX2 randomized trial | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Initial restrictive versus liberal oxygen for traumatic brain injury: A follow-up study of the TRAUMOX2 randomized trial Josefine Bækgaard, Felicia Dinesen, Volkert Siersma, Oscar Rosenkrantz, and 18 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9423108/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Purpose: Hyperoxia in the initial phase of trauma has been associated with harm in patients with traumatic brain injury (TBI), but relevant large, randomized trials are lacking. This study aimed to determine whether TBI patients treated with a restrictive compared with a liberal oxygen strategy had better functional outcome 12 months after TBI. Methods: This preplanned follow-up study of the TRAUMOX2 randomized clinical trial included adult TBI patients from Denmark, the Netherlands, and Switzerland between December 2021, and September 2023. TBI was defined as an Abbreviated Injury Scale (AIS) head score of ≥2 or AIS-defined concussion. Participants were randomly assigned to a restrictive oxygen strategy (target arterial oxygen saturation of 94%) or a liberal oxygen strategy (12–15 L/min oxygen or a fraction of inspired oxygen of 0.6–1.0) for 8 hours. The primary outcome was a favorable outcome at 12 months after trauma, defined as a Glasgow Outcome Scale–Extended (GOSE) score of 5-8. Results: Among 674 eligible patients, 106 were excluded during follow-up, leaving 568 for analysis. The primary analysis revealed no significant difference between the restrictive and liberal oxygen group in favorable outcome at 12 months (65% vs 70%; risk difference, 4.6 percentage points; 95% CI, −2.6 to 11.7; P = .21). Adjusted analyses showed similar results (risk difference, 4.3 percentage points; 95% CI, −2.1 to 10.7; P = .18). Conclusions: In adult TBI patients, an early 8-hour restrictive oxygen strategy did not significantly improve functional outcome at 12 months compared to a liberal strategy. Trial registration: ClinicalTrials.gov Identifier: NCT05146700 Funding/Support: This trial was supported by Novo Nordisk Foundation grant No. NNF20OC0063985. Additionally, the trial was funded by the Joint Research Fund of Odense University Hospital and Rigshospitalet grant No. 136-A5566 and an individual grant to Dr Baekgaard from the Lundbeck Foundation. Intensive care oxygen trauma traumatic brain injury Figures Figure 1 Figure 2 Introduction Traumatic brain injury (TBI) contributes to worldwide death and disability more than any other traumatic insult,[1] and health economic consequences can be substantial.[2, 3] TBI generally results from direct or indirect biomechanical forces that disrupt normal cerebral function. Severity varies from mild injuries such as concussions to severe cases that may result in coma or death. Although the majority of TBIs are classified as mild, accurately distinguishing mild from more severe injuries in the prehospital setting can be clinically challenging. Nonetheless, as death from severe cases often occurs within the first hours after the insult, prehospital and early in-hospital physiological stabilization is critical, as observational studies have shown a synergistic and deleterious effect of hypoxia and hypotension.[4, 5] Thus, guidelines on TBI management recommend continuous monitoring and prevention of both in order to prevent secondary insults.[6–9] While hypoxia is thus generally discouraged, recommendations regarding upper limits of oxygen administration vary: some guidelines do not recommend a specific threshold for partial pressure of oxygen in arterial blood (PaO₂),[9] while others discourage PaO₂ levels above 100 mmHg unless guided by in-hospital neuromonitoring.[8] These discrepancies arise as some observational studies have suggested a worse functional outcomes with hyperoxia early after TBI, whereas a recent systematic review including 12 observational and one smaller randomized clinical trial (RCT) found indications of an improved outcome if only studies with a low-to-moderate risk of bias were analyzed.[10, 11] However, there are no definitive published randomized trials evaluating oxygen regimens in patients with head injuries.[12] The few randomized trials on the subject are small and with diverging results,[13–15] and thus there seems to be a notable lack of large RCTs randomizing TBI patients specifically to different levels of oxygenation. The findings of the RCT TRAUMOX2, evaluating restrictive vs liberal oxygen administration during the initial eight hours following trauma, have recently been reported.[16] The present follow-up study constitutes a preplanned follow-up study of patients with TBI within the TRAUMOX2 population. The hypothesis was that a restrictive oxygen strategy results in improved functional outcome 12 months following trauma compared with a liberal oxygen strategy. Methods This follow-up study of the TRAUMOX2 trial is reported in accordance with the Consolidated Standards of Reporting Trials (CONSORT) guidelines, including the Reporting of Patient-Reported Outcomes extension (CONSORT-PRO).[17] The trial protocol and the statistical analysis plan for the primary trial publication were uploaded to the TRAUMOX2 website (www.traumox2.org) before the completion of data collection and have been published in peer reviewed journals.[18, 19] The protocol and combined statistical analysis plan for this follow-up study was uploaded to the TRAUMOX2 website (www.traumox2.org) prior to completion of data collection and further uploaded to www.zenodo.com (https://doi.org/10.5281/zenodo.15837312). Study design and participants The TRAUMOX2 trial was an international, multicenter, open-label, parallel-group, superiority, randomized, controlled, clinical trial. The primary findings have previously been reported.[16] Participants were enrolled either in the prehospital setting or at trauma center admission. Inclusion criteria included age ≥18 years, blunt or penetrating trauma, direct transport to a participating trauma center with a full trauma team activation, and an anticipated hospital length of stay (LOS) of at least 24 hours by the including physician. Exclusion criteria were suspicion of carbon monoxide intoxication and cardiac arrest prior to randomization. Secondary exclusions were classified as participants excluded after randomization in the trauma resuscitation room after secondary survey due to no or minor injuries. Participants were randomized in a 1:1 ratio to receive either a restrictive or a liberal oxygen strategy for the first eight hours after randomization. Block sizes were variable with stratification based on the site of inclusion (specific prehospital base or trauma center) as well as endotracheal intubation at randomization. The randomization table was generated electronically by a statistician not affiliated with the trial and transferred to KLIFO A/S who produced sealed randomization envelopes. Participants in the restrictive oxygen group received the minimal amount of oxygen to maintain an oxygen saturation of 94%, meaning that only participants not requiring supplemental oxygen could exceed the SpO 2 level of 94%. In the liberal oxygen group, participants were allocated to receive 12–15L/min oxygen via mask or an initial FiO 2 of 1.0, which could be reduced to 0.6 after resuscitation. Further details can be found in the TRAUMOX2 primary publication.[16] The current study was a preplanned subgroup analysis of participants with TBI within the TRAUMOX2 population. The TRAUMOX2 study ended inclusion on September 12, 2023, with a total of 1,508 included participants. TBI was defined as a head injury classified with an Abbreviated Injury Scale (AIS) score ≥2 and/or a documented concussion (AIS codes 161000.1 or 161001.1), based on the AIS 2005 revision, 2008 update,[20] resulting in 674 eligible participants for this study. All the included participants who consented to participate at both 6 months and 12 months after trauma were contacted. Outcome measures The primary outcome was the proportion of TBI participants with a favorable recovery (GOSE 5-8) at 12 months after trauma in the two groups of oxygen strategies. The GOSE score contains a series of questions regarding sequelae of the trauma and the effect on everyday life. GOSE can be assessed through a standardized questionnaire completed by the patient or a caregiver, or by structured interview. Based on the answers to the questions, the patient will be given a score from 1-8 ranging from “Death” to “Upper good recovery”. The scale is often dichotomized by dividing the scores into unfavorable and favorable recovery. GOSE was dichotomized to reflect functional independence (GOSE≥5, being able to take care of oneself, grocery shop and transport oneself locally, but challenges with work and social life). GOSE scores were assessed in a telephone interview and reported reflecting disabilities due to both TBI and other injuries as has been done by others and as described in the manual.[21, 22] Secondary outcomes included: favorable recovery (GOSE 5-8) at 6 and 12 months after trauma, complete recovery (GOSE 8) at 6 and 12 months after trauma, and GOSE score reported as a continuous outcome at 6 and 12 months after trauma, 6- and 12-month mortality. Furthermore, the EQ-5D-5L index value at 6 and 12 months after trauma, and EQ VAS at 6 and 12 months after trauma were assessed. The EuroQol EQ-5D-5L questionnaire is a patient-reported outcome measure used to assess participants’ health-related quality of life. In the first section, participants rate five dimensions—mobility, self-care, usual activities, pain/discomfort, and anxiety/depression—and these responses are combined to generate the EQ-5D index score. In the second part, participants assess their current overall health on the EQ-5D VAS, scored from 0 to 100.[23] Finally, the primary outcome was analyzed according to the following chosen subgroups: Participants with AIS head≥5 (severe TBI), participants with AIS head≥3 (moderate to severe TBI), participants with Glasgow Coma Scale (GCS)≤8 (first registered GCS, prehospital or in trauma bay), and participants with GCS≤13 (first registered GCS, prehospital or in the trauma resuscitation room). Statistical analysis A response rate of approximately 80% was anticipated for both the 6- and 12-month follow-ups. An observational study of TBI patients found that 75% of the patients with moderate TBI and 52% of the patients with severe TBI had favorable recoveries (GOSE scores 4-8) at 12 months after trauma.[24] Our follow-up study included participants registered with AIS head ≥2 and/or concussion. A favorable recovery (GOSE 5–8) of 70% was estimated for the liberal oxygen group. With a power of 80% and a 5% significance level, we would be able to detect a clinically relevant difference of 10 percentage points between the two oxygen strategy groups. The primary outcome, favorable GOSE score at 12 months after trauma, was compared between the two groups using multivariable linear regression with binomial error term adjusted for the stratification variables. Results were presented as risks differences (RD) with corresponding 95% CI. Risk differences are presented as liberal minus restrictive. To increase power an additional analysis was conducted further adjusting for certain characteristics tentatively associated with the outcome: age, sex, injury severity scale (ISS), dominating injury type, and first recorded GCS score. Estimation was performed with generalized estimating equations, using the site of inclusion as a clustering variable, to adjust for excess association within the stratification. Inverse probability weighting was used to account for potential differential loss to follow-up by weighting observed outcomes by the inverse probability of remaining in the cohort through the follow-up time point. These probabilities were estimated using a logistic regression model that included treatment allocation, predominant injury type, supplemental oxygen modality, psychiatric comorbidity, and, among participants enrolled in Denmark, whether the participant had a permanent personal identification number or a temporary identification number (i.e., nonresidents). Secondary outcomes were analyzed using multivariable linear regression, with binomial error term for binary outcomes and normal error term for continuously valued outcomes, and with methods and adjustments similar to the primary outcome. Binary outcomes were reported as RD with 95% CIs, and continuous outcomes as mean differences (MD) with 95% CIs. Results Patient characteristics A total of 674 participants (44.7%) from the TRAUMOX2 trial met inclusion criteria for this follow-up study. Among these, 106 (15.7%) did not complete the 12-month follow-up assessment, primarily because they could not be reached. Consequently, the primary analysis included 302 participants in the restrictive oxygen group and 266 participants in the liberal oxygen group (Figure 1). Screening and follow-up of participants with TBI enrolled in the study. Participants were assigned to either a restrictive or a liberal oxygen strategy for the first 8 hours after randomization. Randomization occurred either in the prehospital setting or upon arrival in the trauma resuscitation room, and participants were followed for one year thereafter. Abbreviations: TBI, Traumatic Brain Injury a Arleth T, Baekgaard J, Siersma V, et al.: Early Restrictive vs Liberal Oxygen for Trauma Patients: The TRAUMOX2 Randomized Clinical Trial. JAMA 2025; 333:479–489 https://doi.org/10.1001/jama.2024.25786 b AIS codes 161000.1/161001.1 c Lost to follow-up represents participants where contact was unsuccessful e.g. due to invalid phone number or e-mail address recorded in their electronic health record Baseline characteristics were comparable between the two groups (Table 1). The median PaO 2 was 88 mmHg (IQR, 73-114) in the restrictive oxygen group and 258 mmHg (IQR, 138-399) in the liberal oxygen group. Primary outcome The primary analysis found no statistically significant difference between the restrictive and liberal oxygen strategies in the proportion of participants with a favorable recovery (GOSE 5-8) at 12 months (65% vs 70%, risk difference, 4.6 percentage points; 95% CI, −2.6 to 11.7; P = .21) (Table 2 and Figure 2). The adjusted analysis showed similar results (risk difference, 4.3 percentage points; 95% CI, −2.1 to 10.7; P = .18) (Table 2). Stacked Bar Charts of the Glasgow Outcome Scale – Extended (GOSE) at 12 months among participants with TBI. Participants were assigned to either a restrictive or a liberal oxygen strategy for the first 8 hours after randomization. Randomization occurred either in the prehospital setting or upon arrival in the trauma resuscitation room, and participants were followed for one year thereafter. Secondary outcomes In the restrictive group, 21% of participants had achieved complete recovery (GOSE=8) at 12 months versus 17% in the liberal group (risk difference, −4.3 percentage points; 95% CI, -10.5 to 1.9; P = .17). The analyses on the non-dichotomized GOSE score at 12 months showed no difference between the two oxygen groups (mean difference 0.14; 95% CI, -0.25 to 0.53; P = .48). At 12 months, 19% in the restrictive group and 16% in the liberal group had died (risk difference -2.3; 95% CI, -7.4 to 2.9; P = .39). The remaining secondary outcomes were not statistically significant (Table 2), nor were any of the subgroup analyses (Table 3). The per-protocol analyses also did not demonstrate significant differences either (Supplementary Table 1). Discussion This preplanned follow-up study of TBI patients from the TRAUMOX2 trial revealed no statistically significant difference between a restrictive and liberal oxygen strategy on favorable recovery (GOSE 5-8) at 12 months after trauma. None of the secondary functional outcomes showed significant differences either. None of our predefined subgroups revealed a difference either. To our knowledge, it is the first large, randomized study to compare different levels of oxygen to TBI related outcomes. Both the extent and duration of tissue hypoxia in patients with TBI have long been recognized as predictors of poor outcomes,[25] but whether the same holds true for hyperoxia remains under debate. Early after TBI, patterns of regional cerebral supply-dependency mismatch have been described, with inappropriately low cerebral blood flow relative to the local oxidative metabolism.[26] Hyperoxia has been associated with reductions in ICP, enhanced brain tissue oxygenation, lower lactate concentrations, and a lower lactate-to-pyruvate ratio[27, 28] and may thus confer neuroprotective effects, limiting secondary ischemic injury. Bearing these findings in mind, the present study focused on patient-centered outcomes, specifically neurological outcome and mortality in relation to oxygen levels. Apart from our study, only a few small, randomized trials on the topic exist. Taher et al. conducted a study involving 68 patients, in whom an FiO₂ of 80% administered for six hours post-injury resulted in more favorable long-term neurological outcomes compared with an FiO₂ of 50%, although there was an uncertain risk of bias across all domains.[14] A recent randomized trial by Liu et al., including 110 patients with mild TBI, found that patients receiving high-flow oxygen or hyperbaric oxygen treatment had better neurological outcomes than those receiving low-flow oxygen. However, the patients were evaluated using a low sensitivity cognitive test that is not typically applied to trauma populations, and the control group had Mini Mental State Examination (MMSE) scores of 16–20, indicating moderate dementia rather than mild TBI.[13] Thus, both trials demonstrated beneficial effects of increased levels of supplemental oxygen; however, the interventions in these studies differ from ours, making comparisons difficult. A systematic review on the topic published in 2021 also concluded that evidence on oxygenation strategies in TBI was extremely limited.[29] A very large ongoing trial, Mega-ROX Brains, plans to randomize between 7500 and 9500 patients with nonhypoxic ischemic encephalopathy acute brain injuries and conditions to a conservative and liberal ICU oxygen therapy regimen; the anticipated date of last data collection early 2026.[30] Meanwhile, observational evidence thus far has been mixed. Several studies have suggested that hyperoxia may be detrimental in TBI. A 2015 meta-analysis consisting of 24 studies found arterial hyperoxia during admission increased hospital mortality in critically ill patients, including those with TBI.[31] Of note, many of the studies reporting unfavorable outcomes, define severe hyperoxemia as a PaO₂ exceeding 200 mmHg or much more.[32–34] Consistent with our findings, recent retrospective studies reported no association between mild hyperoxemia (PaO₂ >100–120 mmHg) and unfavorable neurological outcome and/or mortality.[35–37] An additional observational study demonstrated a U-shaped relationship between early arterial oxygenation and long-term functional and cognitive outcomes, with PaO₂ thresholds of 150 and 200 mmHg, associated with the most favorable outcomes.[38] Moreover, a recent systematic review and meta-analysis including 13 studies, one randomized controlled trial (RCT) and two secondary analyses of RCTs evaluating other interventions, found no association between hyperoxia and mortality. For functional outcomes, pooled analyses suggested a potential benefit of hyperoxia when restricted to studies with low to moderate risk of bias. However, the overall certainty of evidence was rated as low for both outcomes according to GRADE, underscoring the need for the present study.[39] Importantly, observational studies of oxygen supply are susceptible to simultaneity, a statistical phenomenon implying that a treatment that is administered at higher doses in more severely ill patients will almost invariably be erroneously interpreted as being detrimental to the patients, even if the analysis is adjusted for severity of illness by, e.g., inverse probability of treatment weighting. In other words, some observational studies may have overestimated the risk of hyperoxia, because it was administered to the most desperately ill patients.[40, 41] In our trial, the median PaO₂ in the restrictive group was 88 mmHg, lower than that reported in many prior studies, suggesting that targeting even lower PaO₂ levels may be reasonable; that said, the point estimate for the primary outcome calls for caution as 65% achieved a favorable outcome in the restrictive group compared to 70% in the liberal group. Notably, the liberal group achieved a median PaO₂ of 258 mmHg, yet no evidence of harm was observed at this level. Strengths and Limitations This follow-up study has several notable strengths. First, its randomized, international, multicenter, and pragmatic design was intended to strengthen internal validity and external generalizability, supporting the interpretation of the findings in a representative trauma population. Second, the outcomes assessed in this study were pre-specified as follow-up outcomes of the TRAUMOX2 trial, reducing the risk of selective reporting. Further, this cohort achieved relatively high follow-up rates in a trauma setting. Finally, the analyses employed inverse probability weighting to mitigate potential attrition bias arising from differential follow-up. Several limitations must also be recognized. First, the choice of using GOSE as our primary outcome measure could be discussed. The score has been criticized for being a non-interval scale and dichotomizing the score into favorable and unfavorable outcome has been said to cause loss of information. The Functional Status Examination (FSE) has been suggested as an alternative and has shown to be related more strongly to severity of brain injury than the GOSE.[42] Nevertheless, to enhance comparability between studies, the GOSE score was reported, consistent with its use in most major TBI research to date.[43–45] Furthermore, the FSE requires more intensive training and clinical familiarity to achieve reliability. Second, neither participants, their relatives, nor the GOSE interviewer were blinded towards the oxygen treatment allocation, which could affect the evaluation of the participants’ functionality according to preexisting beliefs on oxygen treatment. Third, the definition of the TBI population in this study must also be discussed. In an attempt to include all relevant cases, one may risk including participants without TBI. Participants with AIS head=1 were not included to avoid cases with only minor scalp injuries. However, 14 participants in this follow-up study were marked with penetrating injury as the dominating injury type. The cases were manually double checked and all but two clearly suffered a TBI. The two doubtful cases were registered with “Cranial Nerve NSF” damage. The criteria for the TBI population for this follow-up study were defined in the publicly available protocol before last data collection and all 14 participants were included in the analyses. Lastly, the study was conducted in European countries, and while the results may be generalizable to similar populations, caution should be exercised when interpreting the findings in other settings. Definitions and classifications of TBI vary considerably between studies and are sometimes insufficiently reported, limiting comparability across the literature. Declarations Dr Baekgaard reported receiving grants from Novo Nordisk Foundation and Lundbeck Foundation during the conduct of the study. Dr Dinesen reported receiving grants from Danish Air Ambulance and Rigshospitalets Forskningspuljer during the conduct of the study. Dr van Lieshout reported receiving grants from Rigshospitalet during the conduct of the study. Dr Hautz reported receiving grants from Swiss National Science Foundation and the European Union; and personal fees from AO Foundation Zurich, Mundipharma Switzerland, and MDI Australia outside the submitted work. Dr Klimek reported receiving grants to organization for research laboratory Music as Medicine; honoraria from Paion as a member of the data safety board and Tijdstroom uitgeverij for serving as an editor of a Dutch textbook; and serving as a course director for Advanced Trauma Life Support outside the submitted work. Dr Steinmetz reported receiving grants from Novo Nordisk Foundation during the conduct of the study; and funding for professorship from Norwegian Air Ambulance Foundation outside the submitted work. No other disclosures were reported Conclusion In adults with traumatic brain injury, an early restrictive oxygen strategy did not significantly improve functional recovery at 12 months compared with a liberal oxygen strategy initiated in the prehospital setting or on trauma center admission for 8 hours. Declarations Acknowledgments We thank the participants and their relatives, investigators, clinical staff, collaborators, funding bodies, regulatory authorities, and all others who contributed to the TRAUMOX2 trial. Author Contributions: Drs Steinmetz, Dinesen, Arleth, Baekgaard and Siersma had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Concept and design: Baekgaard, Arleth, Dinesen, Siersma, and Steinmetz Acquisition, analysis, or interpretation of data: All authors. Drafting of the manuscript: Baekgaard, Rosenkrantz, Arleth, Dinesen, Siersma, and Steinmetz. Critical review of the manuscript for important intellectual content: All authors. Statistical analysis: Baekgaard, Dinesen, Siersma, and Steinmetz. Obtained funding: Baekgaard, Mikkelsen, Zwisler, Steinmetz. Funding/support: The trial was supported by Novo Nordisk Foundation grant No. NNF20OC0063985. Furthermore, the trial was funded by the Joint Research Fund of Odense University Hospital and Rigshospitalet grant No. 136-A5566 and an individual grant to Dr Baekgaard from the Lundbeck Foundation. Role of Funder/Sponsor: Novo Nordisk Foundation, the Joint Research Fund of Odense University Hospital and Rigshospitalet, and Lundbeck Foundation had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. Data Sharing statement: Please see supplemental 4 of the primary trial publication.[16] References Dewan MC, Rattani A, Gupta S, et al (2019) Estimating the global incidence of traumatic brain injury. J Neurosurg 130:1080–1097. https://doi.org/10.3171/2017.10.JNS17352 Gaudette É, Seabury SA, Temkin N, et al (2022) Employment and Economic Outcomes of Participants With Mild Traumatic Brain Injury in the TRACK-TBI Study. JAMA Netw Open 5:e2219444. https://doi.org/10.1001/jamanetworkopen.2022.19444 Malhotra AK, Jaffe RH, Shakil H, et al (2024) Unemployment and Personal Income Loss After Traumatic Brain Injury. JAMA Surg 159:1415–1422. https://doi.org/10.1001/jamasurg.2024.4285 Chesnut RM, Marshall LF, Klauber MR, et al (1993) The role of secondary brain injury in determining outcome from severe head injury. J Trauma 34:216–222. https://doi.org/10.1097/00005373-199302000-00006 Maiga AW, Lin H-HS, Wisniewski SR, et al (2025) Adverse Prehospital Events and Outcomes After Traumatic Brain Injury. JAMA Netw Open 8:e2457506. https://doi.org/10.1001/jamanetworkopen.2024.57506 Lulla A, Lumba-Brown A, Totten AM, et al (2023) Prehospital Guidelines for the Management of Traumatic Brain Injury - 3rd Edition. Prehosp Emerg Care 27:507–538. https://doi.org/10.1080/10903127.2023.2187905 Gaither JB, Spaite DW, Bobrow BJ, et al (2024) EMS Treatment Guidelines in Major Traumatic Brain Injury With Positive Pressure Ventilation. JAMA Surg 159:363–372. https://doi.org/10.1001/jamasurg.2023.7155 ACS TQP Best Practices Guidelines. In: ACS. https://www.facs.org/quality-programs/trauma/quality/best-practices-guidelines/. Accessed 15 Sep 2025 Overview | Head injury: assessment and early management | Guidance | NICE. https://www.nice.org.uk/guidance/ng232. Accessed 15 Sep 2025 Brenner M, Stein D, Hu P, et al (2012) Association between early hyperoxia and worse outcomes after traumatic brain injury. Arch Surg 147:1042–1046. https://doi.org/10.1001/archsurg.2012.1560 Rincon F, Kang J, Vibbert M, et al (2014) Significance of arterial hyperoxia and relationship with case fatality in traumatic brain injury: a multicentre cohort study. J Neurol Neurosurg Psychiatry 85:799–805. https://doi.org/10.1136/jnnp-2013-305505 Fornaciari A, Zangari R, Polato M, et al (2025) The impact of hyperoxemia on mortality and neurological outcomes in traumatic brain injury: a systematic review and meta-analysis. J Neurosurg Sci 69:481–492. https://doi.org/10.23736/S0390-5616.25.06480-X Liu Z, Wang X, Wu Z, et al (2023) HBOT has a better cognitive outcome than NBH for patients with mild traumatic brain injury: A randomized controlled clinical trial. Medicine (Baltimore) 102:e35215. https://doi.org/10.1097/MD.0000000000035215 Taher A, Pilehvari Z, Poorolajal J, Aghajanloo M (2016) Effects of Normobaric Hyperoxia in Traumatic Brain Injury: A Randomized Controlled Clinical Trial. Trauma Mon 21:e26772. https://doi.org/10.5812/traumamon.26772 Lång M, Skrifvars MB, Siironen J, et al (2018) A pilot study of hyperoxemia on neurological injury, inflammation and oxidative stress. Acta Anaesthesiol Scand 62:801–810. https://doi.org/10.1111/aas.13093 Arleth T, Baekgaard J, Siersma V, et al (2025) Early Restrictive vs Liberal Oxygen for Trauma Patients: The TRAUMOX2 Randomized Clinical Trial. JAMA 333:479–489. https://doi.org/10.1001/jama.2024.25786 Calvert M, Blazeby J, Altman DG, et al (2013) Reporting of Patient-Reported Outcomes in Randomized Trials: The CONSORT PRO Extension. JAMA 309:814–822. https://doi.org/10.1001/jama.2013.879 Arleth T, Baekgaard J, Siersma V, et al (2023) Comparing restrictive versus liberal oxygen strategies for trauma patients: The TRAUMOX2 trial-Statistical analysis plan. Acta Anaesthesiol Scand 67:829–838. https://doi.org/10.1111/aas.14230 Baekgaard J, Arleth T, Siersma V, et al (2022) Comparing restrictive versus liberal oxygen strategies for trauma patients - the TRAUMOX2 trial: protocol for a randomised clinical trial. BMJ Open 12:e064047. https://doi.org/10.1136/bmjopen-2022-064047 Ringdal KG, Hestnes M, Palmer CS (2012) Differences and discrepancies between 2005 and 2008 Abbreviated Injury Scale versions - time to standardise. Scand J Trauma Resusc Emerg Med 20:11. https://doi.org/10.1186/1757-7241-20-11 Wilson L, Boase K, Nelson LD, et al (2021) A Manual for the Glasgow Outcome Scale-Extended Interview. J Neurotrauma 38:2435–2446. https://doi.org/10.1089/neu.2020.7527 Nelson LD, Temkin NR, Barber J, et al (2023) Functional Recovery, Symptoms, and Quality of Life 1 to 5 Years After Traumatic Brain Injury. JAMA Netw Open 6:e233660. https://doi.org/10.1001/jamanetworkopen.2023.3660 User Guides. In: EuroQol. https://euroqol.org/information-and-support/documentation/user-guides/. Accessed 15 Dec 2025 McCrea MA, Giacino JT, Barber J, et al (2021) Functional Outcomes Over the First Year After Moderate to Severe Traumatic Brain Injury in the Prospective, Longitudinal TRACK-TBI Study. JAMA Neurol 78:982–992. https://doi.org/10.1001/jamaneurol.2021.2043 van den Brink WA, van Santbrink H, Steyerberg EW, et al (2000) Brain oxygen tension in severe head injury. Neurosurgery 46:868–876; discussion 876-878. https://doi.org/10.1097/00006123-200004000-00018 Abate MG, Trivedi M, Fryer TD, et al (2008) Early derangements in oxygen and glucose metabolism following head injury: the ischemic penumbra and pathophysiological heterogeneity. Neurocrit Care 9:319–325. https://doi.org/10.1007/s12028-008-9119-2 Tolias CM, Reinert M, Seiler R, et al (2004) Normobaric hyperoxia--induced improvement in cerebral metabolism and reduction in intracranial pressure in patients with severe head injury: a prospective historical cohort-matched study. J Neurosurg 101:435–444. https://doi.org/10.3171/jns.2004.101.3.0435 Nortje J, Coles JP, Timofeev I, et al (2008) Effect of hyperoxia on regional oxygenation and metabolism after severe traumatic brain injury: preliminary findings. Crit Care Med 36:273–281. https://doi.org/10.1097/01.CCM.0000292014.60835.15 Hansen TE, Christensen RE, Baekgaard J, et al (2022) Supplemental oxygen for traumatic brain injury: A systematic review. Acta Anaesthesiol Scand 66:307–316. https://doi.org/10.1111/aas.14019 Young PJ, Al-Fares A, Aryal D, et al (2023) Protocol and statistical analysis plan for the mega randomised registry trial comparing conservative vs. liberal oxygenation targets in adults with nonhypoxic ischaemic acute brain injuries and conditions in the intensive care unit (Mega-ROX Brains). Crit Care Resusc J Australas Acad Crit Care Med 25:53–59. https://doi.org/10.1016/j.ccrj.2023.04.011 Helmerhorst HJF, Roos-Blom M-J, van Westerloo DJ, de Jonge E (2015) Association Between Arterial Hyperoxia and Outcome in Subsets of Critical Illness: A Systematic Review, Meta-Analysis, and Meta-Regression of Cohort Studies. Crit Care Med 43:1508–1519. https://doi.org/10.1097/CCM.0000000000000998 Rincon F, Kang J, Vibbert M, et al (2014) Significance of arterial hyperoxia and relationship with case fatality in traumatic brain injury: a multicentre cohort study. J Neurol Neurosurg Psychiatry 85:799–805. https://doi.org/10.1136/jnnp-2013-305505 Brenner M, Stein D, Hu P, et al (2012) Association between early hyperoxia and worse outcomes after traumatic brain injury. Arch Surg Chic Ill 1960 147:1042–1046. https://doi.org/10.1001/archsurg.2012.1560 Davis DP, Meade W, Sise MJ, et al (2009) Both hypoxemia and extreme hyperoxemia may be detrimental in patients with severe traumatic brain injury. J Neurotrauma 26:2217–2223. https://doi.org/10.1089/neu.2009.0940 Lalla LT, Czorlich P, Fischer M, et al (2025) Dose-dependent association of hyperoxia and decreased favorable outcomes in mechanically ventilated patients with traumatic brain injury, a retrospective cohort study. Eur J Trauma Emerg Surg Off Publ Eur Trauma Soc 51:75. https://doi.org/10.1007/s00068-024-02730-5 Khan R, Alromaih S, Alshabanat H, et al (2021) The Impact of Hyperoxia Treatment on Neurological Outcomes and Mortality in Moderate to Severe Traumatic Brain Injured Patients. J Crit Care Med Univ Med Si Farm Din Targu-Mures 7:227–236. https://doi.org/10.2478/jccm-2021-0014 Raj R, Bendel S, Reinikainen M, et al (2013) Hyperoxemia and long-term outcome after traumatic brain injury. Crit Care Lond Engl 17:R177. https://doi.org/10.1186/cc12856 Alali AS, Temkin N, Vavilala MS, et al (2020) Matching early arterial oxygenation to long-term outcome in severe traumatic brain injury: target values. J Neurosurg 132:537–544. https://doi.org/10.3171/2018.10.JNS18964 Fornaciari A, Zangari R, Polato M, et al (2025) The impact of hyperoxemia on mortality and neurological outcomes in traumatic brain injury: a systematic review and meta-analysis. J Neurosurg Sci 69:481–492. https://doi.org/10.23736/S0390-5616.25.06480-X de Grooth H-J, Girbes ARJ, van der Ven F, et al (2020) Observational Research for Therapies Titrated to Effect and Associated With Severity of Illness: Misleading Results From Commonly Used Statistical Methods. Crit Care Med 48:1720–1728. https://doi.org/10.1097/CCM.0000000000004612 Leisman DE (2020) The Goldilocks Effect in the ICU-When the Data Speak, but Not the Truth. Crit Care Med 48:1887–1889. https://doi.org/10.1097/CCM.0000000000004669 Dikmen S, Machamer J, Manley GT, et al (2019) Functional Status Examination versus Glasgow Outcome Scale Extended as Outcome Measures in Traumatic Brain Injuries: How Do They Compare? J Neurotrauma 36:2423–2429. https://doi.org/10.1089/neu.2018.6198 Roberts I, Yates D, Sandercock P, et al (2004) Effect of intravenous corticosteroids on death within 14 days in 10008 adults with clinically significant head injury (MRC CRASH trial): randomised placebo-controlled trial. Lancet Lond Engl 364:1321–1328. https://doi.org/10.1016/S0140-6736(04)17188-2 Maas AIR, Menon DK, Steyerberg EW, et al (2015) Collaborative European NeuroTrauma Effectiveness Research in Traumatic Brain Injury (CENTER-TBI): a prospective longitudinal observational study. Neurosurgery 76:67–80. https://doi.org/10.1227/NEU.0000000000000575 McCrea MA, Giacino JT, Barber J, et al (2021) Functional Outcomes Over the First Year After Moderate to Severe Traumatic Brain Injury in the Prospective, Longitudinal TRACK-TBI Study. JAMA Neurol 78:982–992. https://doi.org/10.1001/jamaneurol.2021.2043 Tables Table 1. Patient Characteristics Characteristics Restrictive oxygen group (N=352) Liberal oxygen group (N=322) Age (years), median [IQR] 56 (34–68) (n=352) 54 (38–70) (n=322) Sex, n/total (%) Male 251/352 (71.3) 221/322 (68.6) Comorbidities prior to trauma, n/total (%) 176/352 (51.0) 154/322 (48.4) Cardiovascular disease 83/345 (24.1) 69/318 (21.7) Lung disease 29/345 (8.4) 37/318 (11.6) Other 128/345 (37.1) 115/318 (36.2) Psychiatric disorder 32/345 (9.3) 30/318 (9.4) Type of injury, n/total (%) Penetrating Shot by firearm of any caliber 3/346 (0.9) 1/322 (0.3) Stabbed 8/346 (2.3) 2/322 (0.6) Blunt Traffic accident 198/346 (57.2) 184/322 (57.1) Hit by blunt object 17/346 (4.9) 25/322 (7.7) Fall 114/346 (32.9) 100/322 (31.1) Blast/explosion 2/346 (0.6) 1/322 (0.3) Other 4/346 (1.2) 9/322 (2.8) Intubated at randomization, n/total (%) 134/352 (38.1) 122/322 (37.9) Site of inclusion, n/total (%) Inhospital 205/352 (58.2) 179/322 (55.6) Prehospital 147/352 (41.8) 143/322 (44.4) Country of inclusion, n/total (%) Denmark 280/352 (79.5) 267/322 (82.9) The Netherlands 45/352 (12.8) 33/322 (10.2) Switzerland 27/352 (7.7) 22/322 (6.8) Prehospital information Use of prehospital or in-hospital supplemental oxygen prior to randomization, n/total (%) 192/331 (58.0) 164/308 (53.2) Time with supplemental oxygen treatment prior to randomization (min), median [IQR] 30 (18–50) (n=152) 30 (18–53) (n=128) First vital signs , median [IQR] Systolic blood pressure (mmHg) 133 (113–155) (n=352) 135 (119–154) (n=320) Heart rate (beats/min) 83 (69–102) (n=351) 85 (72–101) (n=320) Respiratory rate (breaths/min) 20 (16–24) (n=350) 20 (16–24) (n=313) SpO 2 <90% 56/352 (15.9) 46/320 (14.4) Glasgow Coma Scale (GCS) score<9 111/352 (31.5) 112/322 (34.8) Injury Severity Scale (ISS) score 17 (11–27) (n=352) 17 (10–27) (n=322) Oxygen intervention characteristics PaO 2 (mmHg) a 88 (73–114) (n=305) 258 (138–399) (n=278) Hemoglobin (mmol/l) 7.8 (7.1–8.5) (n=300) 7.8 (6.8–8.7) (n=266) Type of supplemental oxygen, n/total (%) No supplemental oxygen 85/348 (24.4) 8/318 (2.5) Nasal cannula 67/348 (19.3) 16/318 (5.0) Non-rebreather mask 8/348 (2.3) 133/318 (41.8) Intubated (at any point during the intervention period) 188/348 (54.0) 161/318 (50.6) Characteristics of trauma patients with TBI who were allocated to receive either a restrictive or a liberal oxygen strategy for 8 hours after randomization. Participants were randomized either in the prehospital setting or upon admission to the trauma resuscitation room and were followed for the first year after randomization. Abbreviations: SpO 2 , arterial oxygen saturation measured by pulse oximetry; PaO 2 , partial pressure of oxygen in arterial blood. a Arterial blood gases were obtained at hour 1±30 minutes after randomization. Table 2. Primary and Secondary Outcomes Restrictive Oxygen Group, No./Total (%) Liberal oxygen group, No./Total (%) Risk difference, %, (95% CI) P value Adjusted risk difference, %, (95% CI) a P value Primary analysis GOSE favorable recovery (5–8) at 12 months 196/302 (65) 185/266 (70) 4.6 (-2.6 to 11.7) P =.21 4.3 (-2.1 to 10.7) P =.18 Secondary analyses GOSE favorable recovery (5–8) at 6 months 198/303 (65) 186/265 (70) 5.0 (-2.3 to 12.3) P =.18 4.9 (-1.5 to 11.4) P =.13 GOSE complete recovery (8) at 6 months 49/303 (16) 39/265 (15) -2.1 (-7.7 to 3.5) P =.46 -1.2 (-6.8 to 4.4) P =.67 GOSE complete recovery (8) at 12 months 64/302 (21) 45/266 (17) -4.3 (-10.5 to 1.9) P =.17 -4.4 (-10.6 to 1.8) P =.16 Death at 6 months 58/345 (17) 47/305 (15) -1.4 (-6.2 to 3.3) P =.56 -0.9 (-5.6 to 3.8) P =.70 Death at 12 months 66/343 (19) 50/304 (16) -2.3 (-7.4 to 2.9) P =.39 -2.2 (-7.2 to 2.8) P =.39 Restrictive Oxygen Group, Median [IQR] Liberal oxygen group, Median [IQR] Mean Difference, (95% CI) P value Adjusted mean difference (95% CI) a P value GOSE non-dichotomized at 6 months 5 [3–7] 5 [4–7] 0.10 (-0.27 to 0.46) P =.60 0.10 (-0.20 to 0.41) P =.50 GOSE non-dichotomized at 12 months 6 [3–7] 6 [4–7] 0.14 (-0.25 to 0.53) P =.48 0.12 (-0.21 to 0.46) P =.47 EQ-5D-5L b index value at 6 months after trauma 0.85 [0.58–0.95] 0.83 [0.58–0.95] -0.02 (-0.08 to 0.04) P =.51 -0.01 (-0.07 to 0.05) P =.72 EQ-5D-5L b index value at 12 months after trauma 0.88 [0.64–0.97] 0.83 [0.58–0.95] -0.03 (-0.09 to 0.02) P =.24 -0.03 (-0.08 to 0.03) P =.34 EQ VAS c at 6 months after trauma 75 [50–85] 70 [50–85] -0.82 (-5.0 to 3.3) P =.70 -0.44 (-4.4 to 3.5) P =.83 EQ VAS c at 12 months after trauma 75 [50–90] 70 [50–85] -2.3 (-6.5 to 1.9) P =.28 -2.0 (-6.0 to 2.1) P =.34 Functional outcomes, mortality, and health-related quality of life assessed in trauma patients assigned to either a restrictive or liberal oxygen strategy for the first 8 hours after randomization. Participants were randomized in the prehospital setting or upon arrival in the trauma resuscitation room and were followed for one year after randomization. Abbreviations: GOSE, Glasgow Outcome Scale – Extended a Adjusted for the stratification variables (site of inclusion and status of endotracheal intubation upon inclusion) and the following variables: age, sex, Injury Severity Scale (ISS) score, predominant type of injury, and first recorded Glasgow Coma Scale (GCS) score. b EQ-5D-5L is a generic instrument that reflects the health-related quality of life. The questionnaire contains five domains, and each domain has five levels of response: no problems, slight problems, moderate problems, severe problems, extreme problems/unable to. c EQ-5D VAS range from 0 to 100, from worst imaginable health to best imaginable health, and reflect the participant’s perceived health. Participants or a relative were interviewed by telephone to obtain the scores. Up to five attempts for contact were made within each timeframe. Table 3. Subgroup Analyses of GOSE Favorable Recovery (Scores 5–8) at 12 Months Restrictive oxygen group, No./Total (%) Liberal oxygen group, No./Total (%) Risk difference estimate (95% CI) P value Adjusted risk difference a (95% CI) P value Severe TBI (AIS head≥5) 13/52 (25) 19/52 (37) 12.6 (-4.7 to 30.0) P = .15 11.3 (-4.1 to 26.8) P =.15 Moderate to severe TBI (AIS head≥3) 117/203 (58) 98/165 (59) 2.7 (-7.1 to 12.4) P =.59 33.1 (-5.0 to 11.6) P =.43 First registered GCS≤8 b 46/102 (45) 46/90 (51) 7.0 (-7.1 to 21.1) P =.33 8.0 (-3.8 to 19.7) P =.18 First registered GCS≤13 b 106/182 (58) 90/155 (58) 0.6 (-9.5 to 10.8) P =.90 4.6 (-3.8 to 13.1) P =.28 Subgroup analyses on participants who had a favorable recovery at 12 months, defined as a GOSE score of 5–8. The subgroups included patients with severe TBI (AIS head≥ 5), patients with moderate to severe TBI (AIS head≥ 3), patients with a GCS≤ 8 (first recorded GCS in the prehospital setting or trauma bay), and patients with a GCS≤ 13 (first recorded GCS in the prehospital setting or trauma resuscitation room). Participants were assigned to either a restrictive or a liberal oxygen strategy for the first 8 hours after randomization. Randomization occurred either in the prehospital setting or upon arrival in the trauma resuscitation room, and participants were followed for one year thereafter. Abbreviations: GOSE, Glasgow Outcome Scale Extended; TBI, traumatic brain injury; AIS, abbreviated injury scale; GCS; Glasgow Coma Scale a Adjusted for the stratification variables (site of inclusion and status of endotracheal intubation upon inclusion) and the following variables: age, sex, Injury Severity Scale (ISS) score, predominant type of injury, and first recorded Glasgow Coma Scale (GCS) score. b The first registered GCS could be either in the prehospital phase or trauma bay Table 4. Glasgow Outcome Scale–Extended (GOSE) Response Characteristics Characteristics Restrictive oxygen group (N=352) Liberal oxygen group (N=322) Response rate at 6 months, n/total (%) Denmark Yes 197/352 (56.0) 186/322 (57.8) No 37/352 (10.5) 40/322 (12.4) Dead 46/352 (13.1) 37/322 (11.5) Withdrawn consent 0/352 (0.0) 4/322 (1.2) Lost to follow-up a 0/352 (0.0) 0/322 (0.0) The Netherlands Yes 29/352 (8.2) 21/322 (6.5) No 6/352 (1.7) 6/322 (1.9) Dead 7/352 (2.0) 5/322 (1.6) Withdrawn consent 2/352 (0.6) 0/322 (0.0) Lost to follow-up a 1/352 (0.3) 1/322 (0.3) Switzerland Yes 19/352 (5.4) 11/322 (3.4) No 3/352 (0.9) 6/322 (1.9) Dead 5/352 (1.4) 5/322 (1.6) Withdrawn consent 0/352 (0.0) 0/322 (0.0) Lost to follow-up a 0/352 (0.0) 0/322 (0.0) Response rate at 12 months, n/total (%) Denmark Yes 193/352 (54.8) 181/322 (56.2) No 34/352 (9.7) 43/322 (13.4) Dead 53/352 (15.1) 39/322 (12.1) Withdrawn consent 0/352 (0.0) 4/322 (1.2) Lost to follow-up a 0/352 (0.0) 0/322 (0.0) The Netherlands Yes 28/352 (8.0) 22/322 (6.8) No 6/352 (1.7) 5/322 (1.6) Dead 8/352 (2.3) 5/322 (1.6) Withdrawn consent 2/352 (0.6) 0/322 (0.0) Lost to follow-up a 1/352 (0.3) 1/322 (0.3) Switzerland Yes 15/352 (4.3) 13/322 (4.0) No 7/352 (2.0) 3/322 (0.9) Dead 5/352 (1.4) 6/322 (1.9) Withdrawn consent 0/352 (0.0) 0/322 (0.0) Lost to follow-up a 0/352 (0.0) 0/322 (0.0) No response at 6 months, reason, n/total no-response (%) No answer despite several attempts 28/46 (60.9) 25/52 (48.1) Missing contact information 7/46 (15.2) 13/52 (25.0) Did not wish to participate 9/46 (19.6) 10/52 (19.2) Other 2/46 (4.3) 4/52 (7.7) Unknown 0/46 (0.0) 0/52 (0.0) No response at 12 months, reason, n/total no-response (%) No answer despite several attempts 23/47 (48.9) 24/51 (47.1) Missing contact information 8/47 (17.0) 14/51 (27.5) Did not wish to participate 14/47 (29.8) 10/51 (19.6) Other 2/47 (4.3) 2/51 (3.9) Unknown 0/47 (0.0) 1/51 (2.0) Respondent at 6 months, n/total responders (%) Participant 217/245 (88.6) 197/218 (90.4) Next-of-kin/friend/caretaker 21/245 (8.6) 15/218 (6.9) Participant and next-of-kin/friend/caretaker 7/245 (2.9) 6/218 (2.8) Unknown 0/245 (0.0) 0/218 (0.0) Respondent at 12 months, n/total responders (%) Participant 205/236 (86.9) 192/217 (88.5) Next-of-kin/friend/caretaker 26/236 (11.0) 21/217 (9.7) Participant and next-of-kin/friend/caretaker 5/236 (2.1) 4/217 (1.8) Unknown 0/236 (0.0) 0/217 (0.0) Contact established at 6 months (days), median [IQR] 189 (185–200) (n=245) 190 (185–199) (n=218) Contact established at 12 months (days), median [IQR] 371 (367–378) (n=236) 371 (367–378) (n=217) Characteristics of participants assessed using the Glasgow Outcome Scale–Extended (GOSE). Participants were assigned to either a restrictive or a liberal oxygen strategy for the first 8 hours after randomization. Randomization occurred either in the prehospital setting or upon arrival in the trauma resuscitation room, and participants were followed for one year thereafter. a Lost to follow-up represents participants where contact was unsuccessful e.g. due to invalid phone number or e-mail address recorded in their electronic health record Supplementary Files CONSORT2025expandedchecklist.pdf eTable1.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9423108","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":624318608,"identity":"623eb320-41aa-4937-8018-1fa3e3434368","order_by":0,"name":"Josefine Bækgaard","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABCUlEQVRIiWNgGAWjYBACxgYgUWHAwABEzBCh4w1EaDmDouXMASKsOsOArOVGAn7VzO3tDx8cKLgjZ85++LDhlxq7fL6brxMfMNTYRON0WM8ZY4MDBs+MLXvSkpNljiVbzrydu9mA4VhaLi4vMc7IYZP+YHA4ccOBHOPDEmzMBga3c7dJMDYcxqMl/fmPAwaH6zecf//5sMS/egODm2e3/8CvJcGMAaglweBGDnPix7bDBgY3eLcx4NUC9IsEUIvhzhnPjI0Z+44bSJ7J3SyRgMcvhsAQ+3Dgz2F5c/7kx5I/vlUb8B0/u/HDhxob3FqQJZh5YKwEHMpBQB7FlT/wqBwFo2AUjIKRCwBQqmbNsHyUYwAAAABJRU5ErkJggg==","orcid":"https://orcid.org/0000-0002-5420-723X","institution":"Rigshospitalet","correspondingAuthor":true,"prefix":"","firstName":"Josefine","middleName":"","lastName":"Bækgaard","suffix":""},{"id":624318609,"identity":"b6232c37-dafb-4ade-a40b-8c51eaa42f9e","order_by":1,"name":"Felicia Dinesen","email":"","orcid":"","institution":"Rigshospitalet","correspondingAuthor":false,"prefix":"","firstName":"Felicia","middleName":"","lastName":"Dinesen","suffix":""},{"id":624318610,"identity":"fcd12d98-20f9-44ff-af14-d6c2b371f763","order_by":2,"name":"Volkert Siersma","email":"","orcid":"","institution":"University of Copenhagen Department of Public Health: Kobenhavns Universitet Institut for Folkesundhedsvidenskab","correspondingAuthor":false,"prefix":"","firstName":"Volkert","middleName":"","lastName":"Siersma","suffix":""},{"id":624318611,"identity":"115cc02d-3fa7-4f08-bcf9-15b598db31af","order_by":3,"name":"Oscar Rosenkrantz","email":"","orcid":"","institution":"Rigshospitalet","correspondingAuthor":false,"prefix":"","firstName":"Oscar","middleName":"","lastName":"Rosenkrantz","suffix":""},{"id":624318612,"identity":"569f1070-0c2e-4d28-be2a-f39bb5e82d60","order_by":4,"name":"Kirsten Moeller","email":"","orcid":"","institution":"Rigshospitalet","correspondingAuthor":false,"prefix":"","firstName":"Kirsten","middleName":"","lastName":"Moeller","suffix":""},{"id":624318613,"identity":"5ecc4dc9-13b1-49d2-a526-038733c1ebd7","order_by":5,"name":"Johan Heiberg","email":"","orcid":"","institution":"Rigshospitalet","correspondingAuthor":false,"prefix":"","firstName":"Johan","middleName":"","lastName":"Heiberg","suffix":""},{"id":624318614,"identity":"5bfb71df-17c9-4030-aad9-f4c9f95428b9","order_by":6,"name":"Ramona Astrand","email":"","orcid":"","institution":"Rigshospitalet","correspondingAuthor":false,"prefix":"","firstName":"Ramona","middleName":"","lastName":"Astrand","suffix":""},{"id":624318615,"identity":"37a68278-7c08-4c49-8d43-9131c4a0843b","order_by":7,"name":"Soeren Mikkelsen","email":"","orcid":"","institution":"Odense Universitetshospital","correspondingAuthor":false,"prefix":"","firstName":"Soeren","middleName":"","lastName":"Mikkelsen","suffix":""},{"id":624318616,"identity":"f281f8e8-70ee-444a-a29a-491689111c23","order_by":8,"name":"Stine Zwisler","email":"","orcid":"","institution":"Odense Universitetshospital","correspondingAuthor":false,"prefix":"","firstName":"Stine","middleName":"","lastName":"Zwisler","suffix":""},{"id":624318617,"identity":"719ebd5d-f9b5-4eaa-bc03-5ec03d8ff46f","order_by":9,"name":"Louise Breum","email":"","orcid":"","institution":"Odense Universitetshospital","correspondingAuthor":false,"prefix":"","firstName":"Louise","middleName":"","lastName":"Breum","suffix":""},{"id":624318618,"identity":"b2dd9005-05ec-4df8-8ef3-407e71f6e4c4","order_by":10,"name":"Mikkel Andersen","email":"","orcid":"","institution":"Aarhus University Hospital Skejby: Aarhus Universitetshospital","correspondingAuthor":false,"prefix":"","firstName":"Mikkel","middleName":"","lastName":"Andersen","suffix":""},{"id":624318619,"identity":"d57a8efd-195a-423f-9426-de695d31e5dc","order_by":11,"name":"Christian Fenger-Eriksen","email":"","orcid":"","institution":"Aarhus Universitetshospital Skejby: Aarhus Universitetshospital","correspondingAuthor":false,"prefix":"","firstName":"Christian","middleName":"","lastName":"Fenger-Eriksen","suffix":""},{"id":624318620,"identity":"9eb6077e-843b-4264-b02d-f0bc9e88cb00","order_by":12,"name":"Lars Henrik Østergaard","email":"","orcid":"","institution":"Aarhus University Hospital Skejby: Aarhus Universitetshospital","correspondingAuthor":false,"prefix":"","firstName":"Lars","middleName":"Henrik","lastName":"Østergaard","suffix":""},{"id":624318621,"identity":"ceac02a2-d7d5-4647-bf94-a1ea9ea1e1e5","order_by":13,"name":"Esther M.M. Van Lieshout","email":"","orcid":"","institution":"Erasmus University Rotterdam: Erasmus Universiteit Rotterdam","correspondingAuthor":false,"prefix":"","firstName":"Esther","middleName":"M.M. Van","lastName":"Lieshout","suffix":""},{"id":624318622,"identity":"f7a9f0c1-82cd-4dc2-9768-f6ff2e1d27b6","order_by":14,"name":"Mark G. Van Vledder","email":"","orcid":"","institution":"Erasmus University Rotterdam: Erasmus Universiteit Rotterdam","correspondingAuthor":false,"prefix":"","firstName":"Mark","middleName":"G. Van","lastName":"Vledder","suffix":""},{"id":624318623,"identity":"1291022a-cf98-4a81-ab2c-91b561c3dac8","order_by":15,"name":"Dennis den Hartog","email":"","orcid":"","institution":"Erasmus University Rotterdam: Erasmus Universiteit Rotterdam","correspondingAuthor":false,"prefix":"","firstName":"Dennis","middleName":"den","lastName":"Hartog","suffix":""},{"id":624318624,"identity":"adcf3c2c-4210-420d-aafe-331e0d9a23b8","order_by":16,"name":"Dominik A. Jakob","email":"","orcid":"","institution":"University of Bern: Universitat Bern","correspondingAuthor":false,"prefix":"","firstName":"Dominik","middleName":"A.","lastName":"Jakob","suffix":""},{"id":624318625,"identity":"a8539c4e-c4ee-4a10-b8c4-18da7a08d864","order_by":17,"name":"Matthias Haenggi","email":"","orcid":"","institution":"University of Zurich: Universitat Zurich","correspondingAuthor":false,"prefix":"","firstName":"Matthias","middleName":"","lastName":"Haenggi","suffix":""},{"id":624318626,"identity":"b477c709-96ac-449e-ac85-a85b40d7494f","order_by":18,"name":"Wolf E. Hautz","email":"","orcid":"","institution":"University of Bern: Universitat Bern","correspondingAuthor":false,"prefix":"","firstName":"Wolf","middleName":"E.","lastName":"Hautz","suffix":""},{"id":624318627,"identity":"07df190b-e818-4bb0-a72b-1d9ff2cf007e","order_by":19,"name":"Markus Klimek","email":"","orcid":"","institution":"Erasmus University Rotterdam: Erasmus Universiteit Rotterdam","correspondingAuthor":false,"prefix":"","firstName":"Markus","middleName":"","lastName":"Klimek","suffix":""},{"id":624318628,"identity":"fe4b1eb3-864a-4d55-bb26-85c20811348b","order_by":20,"name":"Tobias Arleth","email":"","orcid":"","institution":"Rigshospitalet","correspondingAuthor":false,"prefix":"","firstName":"Tobias","middleName":"","lastName":"Arleth","suffix":""},{"id":624318629,"identity":"8e085aca-a533-4737-b7df-1ce49b50bfae","order_by":21,"name":"Jacob Steinmetz","email":"","orcid":"","institution":"Rigshospitalet","correspondingAuthor":false,"prefix":"","firstName":"Jacob","middleName":"","lastName":"Steinmetz","suffix":""}],"badges":[],"createdAt":"2026-04-15 07:18:02","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9423108/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9423108/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":107833644,"identity":"57abbc18-5c83-4797-b417-6fe41e19b637","added_by":"auto","created_at":"2026-04-26 15:40:09","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":59479,"visible":true,"origin":"","legend":"\u003cp\u003eScreening and Follow-Up of TBI Patients in the TRAUMOX2 Trial\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-9423108/v1/d2d541a59a3f1e7743d48acd.png"},{"id":107833657,"identity":"c314156e-2afb-4d15-a5a2-5ddfdffc97cb","added_by":"auto","created_at":"2026-04-26 15:40:14","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":88513,"visible":true,"origin":"","legend":"\u003cp\u003eStacked Bar Charts of the Glasgow Outcome Scale – Extended (GOSE) at 12 months.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-9423108/v1/e2950bbe5b7c3e7dd006f360.png"},{"id":109204665,"identity":"24c1538d-15be-4443-ba58-83d93ad391f6","added_by":"auto","created_at":"2026-05-13 15:01:44","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":495406,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9423108/v1/9cc060fa-ea92-4aa8-b509-aee7f63eab44.pdf"},{"id":107833694,"identity":"259b20be-d4da-4291-abfb-f2ee9af04636","added_by":"auto","created_at":"2026-04-26 15:40:24","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":460061,"visible":true,"origin":"","legend":"","description":"","filename":"CONSORT2025expandedchecklist.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9423108/v1/4ce8411edfab3cfcdcdce7d8.pdf"},{"id":107833656,"identity":"9c2102ce-df70-453c-a033-99e68ebe2abe","added_by":"auto","created_at":"2026-04-26 15:40:13","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":18689,"visible":true,"origin":"","legend":"","description":"","filename":"eTable1.docx","url":"https://assets-eu.researchsquare.com/files/rs-9423108/v1/f3abde29035a01c2d658290f.docx"}],"financialInterests":"","formattedTitle":"Initial restrictive versus liberal oxygen for traumatic brain injury: \n\nA follow-up study of the TRAUMOX2 randomized trial","fulltext":[{"header":"Introduction","content":"\u003cp\u003eTraumatic brain injury (TBI) contributes to worldwide death and disability more than any other traumatic insult,[1]\u0026nbsp;and health economic consequences can be substantial.[2, 3] TBI generally results from direct or indirect biomechanical forces that disrupt normal cerebral function. Severity varies from mild injuries such as concussions to severe cases that may result in coma or death. Although the majority of TBIs are classified as mild, accurately distinguishing mild from more severe injuries in the prehospital setting can be clinically challenging. Nonetheless, as death from severe cases often occurs within the first hours after the insult, prehospital and early in-hospital physiological stabilization is critical, as observational studies have shown a synergistic and deleterious effect of hypoxia and hypotension.[4, 5] Thus, guidelines on TBI management recommend continuous monitoring and prevention of both in order to prevent secondary insults.[6\u0026ndash;9]\u003c/p\u003e\n\u003cp\u003eWhile hypoxia is thus generally discouraged, recommendations regarding upper limits of oxygen administration vary: some guidelines do not recommend a specific threshold for partial pressure of oxygen in arterial blood (PaO₂),[9] while others discourage PaO₂\u0026nbsp;levels above 100 mmHg unless guided by in-hospital neuromonitoring.[8]\u0026nbsp;These discrepancies arise as some observational studies have suggested a worse functional outcomes with hyperoxia early after TBI, whereas a recent systematic review including 12 observational and one smaller randomized clinical trial (RCT) found indications of an improved outcome if only studies with a low-to-moderate risk of bias were analyzed.[10, 11]\u0026nbsp;However, there are no definitive published randomized trials evaluating oxygen regimens in patients with head injuries.[12]\u0026nbsp;The few randomized trials on the subject are small and with diverging results,[13\u0026ndash;15]\u0026nbsp;and thus there seems to be a notable lack of large RCTs randomizing TBI patients specifically to different levels of oxygenation. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe findings of the RCT TRAUMOX2, evaluating restrictive vs liberal oxygen administration during the initial eight hours following trauma, have recently been reported.[16] The present follow-up study constitutes a preplanned follow-up study of patients with TBI within the TRAUMOX2 population. The hypothesis was that a restrictive oxygen strategy results in improved functional outcome 12 months following trauma compared with a liberal oxygen strategy.\u003c/p\u003e"},{"header":"Methods ","content":"\u003cp\u003eThis follow-up study of the TRAUMOX2 trial is reported in accordance with the Consolidated Standards of Reporting Trials (CONSORT) guidelines, including the Reporting of Patient-Reported Outcomes extension (CONSORT-PRO).[17] The trial protocol and the statistical analysis plan for the primary trial publication were uploaded to the TRAUMOX2 website (www.traumox2.org) before the completion of data collection and have been published in peer reviewed journals.[18, 19] The protocol and combined statistical analysis plan for this follow-up study was uploaded to the TRAUMOX2 website (www.traumox2.org) prior to completion of data collection and further uploaded to www.zenodo.com (https://doi.org/10.5281/zenodo.15837312).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStudy design and participants\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe TRAUMOX2 trial was an international, multicenter, open-label, parallel-group, superiority, randomized, controlled, clinical trial. The primary findings have previously been reported.[16] Participants were enrolled either in the prehospital setting or at trauma center admission. Inclusion criteria included age \u0026ge;18 years, blunt or penetrating trauma, direct transport to a participating trauma center with a full trauma team activation, and an anticipated hospital length of stay (LOS) of at least 24 hours by the including physician. Exclusion criteria were suspicion of carbon monoxide intoxication and cardiac arrest prior to randomization. Secondary exclusions were classified as participants excluded after randomization in the trauma resuscitation room after secondary survey due to no or minor injuries.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eParticipants were randomized in a 1:1 ratio to receive either a restrictive or a liberal oxygen strategy for the first eight hours after randomization. Block sizes were variable with stratification based on the site of inclusion (specific prehospital base or trauma center) as well as endotracheal intubation at randomization.\u0026nbsp;The randomization table was generated electronically by a statistician not affiliated with the trial and transferred to KLIFO A/S who produced sealed randomization envelopes.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;Participants in the restrictive oxygen group received the minimal amount of oxygen to maintain an oxygen saturation of 94%, meaning that only participants not requiring supplemental oxygen could exceed the SpO\u003csub\u003e2\u003c/sub\u003e level of 94%. In the liberal oxygen group, participants were allocated to receive 12\u0026ndash;15L/min oxygen via mask or an initial FiO\u003csub\u003e2\u003c/sub\u003e of 1.0, which could be reduced to 0.6 after resuscitation. Further details can be found in the TRAUMOX2 primary publication.[16]\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe current study was a preplanned subgroup analysis of participants with TBI within the TRAUMOX2 population. The TRAUMOX2 study ended inclusion on September 12, 2023, with a total of 1,508 included participants. TBI was defined as a head injury classified with an Abbreviated Injury Scale (AIS) score \u0026ge;2 and/or a documented concussion (AIS codes 161000.1 or 161001.1), based on the AIS 2005 revision, 2008 update,[20] resulting in 674 eligible participants for this study. All the included participants who consented to participate at both 6 months and 12 months after trauma were contacted.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eOutcome measures\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe primary outcome was the proportion of TBI participants with a favorable recovery (GOSE 5-8) at 12 months after trauma in the two groups of oxygen strategies. The GOSE score contains a series of questions regarding sequelae of the trauma and the effect on everyday life. GOSE can be assessed through a standardized questionnaire completed by the patient or a caregiver, or by structured interview. Based on the answers to the questions, the patient will be given a score from 1-8 ranging from \u0026ldquo;Death\u0026rdquo; to \u0026ldquo;Upper good recovery\u0026rdquo;. The scale is often dichotomized by dividing the scores into unfavorable and favorable recovery. GOSE was dichotomized to reflect functional independence (GOSE\u0026ge;5, being able to take care of oneself, grocery shop and transport oneself locally, but challenges with work and social life). GOSE scores were assessed in a telephone interview and reported reflecting disabilities due to both TBI and other injuries as has been done by others and as described in the manual.[21, 22]\u003c/p\u003e\n\u003cp\u003eSecondary outcomes included: favorable recovery (GOSE 5-8) at 6 and 12 months after trauma, complete recovery (GOSE 8) at 6 and 12 months after trauma, and GOSE score reported as a continuous outcome at 6 and 12 months after trauma, 6- and 12-month mortality.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFurthermore, the EQ-5D-5L index value at 6 and 12 months after trauma, and EQ VAS at 6 and 12 months after trauma were assessed. The EuroQol EQ-5D-5L questionnaire is a patient-reported outcome measure used to assess participants\u0026rsquo; health-related quality of life. In the first section, participants rate five dimensions\u0026mdash;mobility, self-care, usual activities, pain/discomfort, and anxiety/depression\u0026mdash;and these responses are combined to generate the EQ-5D index score. In the second part, participants assess their current overall health on the EQ-5D VAS, scored from 0 to 100.[23]\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFinally, the primary outcome was analyzed according to the following chosen subgroups: Participants with AIS head\u0026ge;5 (severe TBI), participants with AIS head\u0026ge;3 (moderate to severe TBI), participants with Glasgow Coma Scale (GCS)\u0026le;8 (first registered GCS, prehospital or in trauma bay), and participants with GCS\u0026le;13 (first registered GCS, prehospital or in the trauma resuscitation room).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStatistical analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA response rate of approximately 80% was anticipated for both the 6- and 12-month follow-ups. An observational study of TBI patients found that 75% of the patients with moderate TBI and 52% of the patients with severe TBI had favorable recoveries (GOSE scores 4-8) at 12 months after trauma.[24] Our follow-up study included participants registered with AIS head \u0026ge;2 and/or concussion. A favorable recovery (GOSE 5\u0026ndash;8) of 70% was estimated for the liberal oxygen group. With a power of 80% and a 5% significance level, we would be able to detect a clinically relevant difference of 10 percentage points between the two oxygen strategy groups.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe primary outcome, favorable GOSE score at 12 months after trauma, was compared between the two groups using multivariable linear regression with binomial error term adjusted for the stratification variables. Results were presented as risks differences (RD) with corresponding 95% CI. Risk differences are presented as liberal minus restrictive.\u003c/p\u003e\n\u003cp\u003eTo increase power an additional analysis was conducted further adjusting for certain characteristics tentatively associated with the outcome: age, sex, injury severity scale (ISS), dominating injury type, and first recorded GCS score.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eEstimation was performed with generalized estimating equations, using the site of inclusion as a clustering variable, to adjust for excess association within the stratification. Inverse probability weighting was used to account for potential differential loss to follow-up by weighting observed outcomes by the inverse probability of remaining in the cohort through the follow-up time point. These probabilities were estimated using a logistic regression model that included treatment allocation, predominant injury type, supplemental oxygen modality, psychiatric comorbidity, and, among participants enrolled in Denmark, whether the participant had a permanent personal identification number or a temporary identification number (i.e., nonresidents).\u003c/p\u003e\n\u003cp\u003eSecondary outcomes were analyzed using multivariable linear regression, with binomial error term for binary outcomes and normal error term for continuously valued outcomes, and with methods and adjustments similar to the primary outcome. Binary outcomes were reported as RD with 95% CIs, and continuous outcomes as mean differences (MD) with 95% CIs.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003ePatient characteristics\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA total of 674 participants (44.7%) from the TRAUMOX2 trial met inclusion criteria for this follow-up study. Among these, 106 (15.7%) did not complete the 12-month follow-up assessment, primarily because they could not be reached. Consequently, the primary analysis included 302 participants in the restrictive oxygen group and 266 participants in the liberal oxygen group (Figure 1).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eScreening and follow-up of participants with TBI enrolled in the study. Participants were assigned to either a restrictive or a liberal oxygen strategy for the first 8 hours after randomization. Randomization occurred either in the prehospital setting or upon arrival in the trauma resuscitation room, and participants were followed for one year thereafter.\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eAbbreviations: TBI, Traumatic Brain Injury\u003c/p\u003e\n\u003cp\u003e\u003csup\u003ea\u003c/sup\u003eArleth T, Baekgaard J, Siersma V, et al.: Early Restrictive vs Liberal Oxygen for Trauma Patients: The TRAUMOX2 Randomized Clinical Trial. JAMA 2025; 333:479\u0026ndash;489 https://doi.org/10.1001/jama.2024.25786\u003cbr\u003e\u003csup\u003eb\u003c/sup\u003eAIS codes 161000.1/161001.1\u003cbr\u003e\u003csup\u003ec\u003c/sup\u003eLost to follow-up represents participants where contact was unsuccessful e.g. due to invalid phone number or e-mail address recorded in their electronic health record\u003c/p\u003e\n\u003cp\u003eBaseline characteristics were comparable between the two groups (Table 1).\u003c/p\u003e\n\u003cp\u003eThe median PaO\u003csub\u003e2\u003c/sub\u003e was 88 mmHg (IQR, 73-114) in the restrictive oxygen group and 258 mmHg (IQR, 138-399) in the liberal oxygen group.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePrimary outcome\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe primary analysis found no statistically significant difference between the restrictive and liberal oxygen strategies in the proportion of participants with a favorable recovery (GOSE 5-8) at 12 months (65% vs 70%, risk difference, 4.6 percentage points; 95% CI, \u0026minus;2.6 to 11.7; P = .21) (Table 2 and Figure 2). The adjusted analysis showed similar results (risk difference, 4.3 percentage points; 95% CI, \u0026minus;2.1 to 10.7; P = .18) (Table 2).\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eStacked Bar Charts of the Glasgow Outcome Scale \u0026ndash; Extended (GOSE) at 12 months among participants with TBI. Participants were assigned to either a restrictive or a liberal oxygen strategy for the first 8 hours after randomization. Randomization occurred either in the prehospital setting or upon arrival in the trauma resuscitation room, and participants were followed for one year thereafter.\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSecondary outcomes\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn the restrictive group, 21% of participants had achieved complete recovery (GOSE=8) at 12 months versus 17% in the liberal group (risk difference, \u0026minus;4.3 percentage points; 95% CI, -10.5 to 1.9; P = .17). The analyses on the non-dichotomized GOSE score at 12 months showed no difference between the two oxygen groups (mean difference 0.14; 95% CI, -0.25 to 0.53; P = .48). At 12 months, 19% in the restrictive group and 16% in the liberal group had died (risk difference -2.3; 95% CI, -7.4 to 2.9; P = .39).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe remaining secondary outcomes were not statistically significant (Table 2), nor were any of the subgroup analyses (Table 3). The per-protocol analyses also did not demonstrate significant differences either (Supplementary Table 1).\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eThis preplanned follow-up study of TBI patients from the TRAUMOX2 trial revealed no statistically significant difference between a restrictive and liberal oxygen strategy on favorable recovery (GOSE 5-8) at 12 months after trauma. None of the secondary functional outcomes showed significant differences either. None of our predefined subgroups revealed a difference either.\u003c/p\u003e\n\u003cp\u003eTo our knowledge, it is the first large, randomized study to compare different levels of oxygen to TBI related outcomes.\u003c/p\u003e\n\u003cp\u003eBoth the extent and duration of tissue hypoxia in patients with TBI have long been recognized as predictors of poor outcomes,[25] but whether the same holds true for hyperoxia remains under debate. Early after TBI, patterns of regional cerebral supply-dependency mismatch have been described, with inappropriately low cerebral blood flow relative to the local oxidative metabolism.[26] Hyperoxia has been associated with reductions in ICP, enhanced brain tissue oxygenation, lower lactate concentrations, and a lower lactate-to-pyruvate ratio[27, 28] and may thus confer neuroprotective effects, limiting secondary ischemic injury.\u003cbr\u003e\u0026nbsp;Bearing these findings in mind, the present study focused on patient-centered outcomes, specifically neurological outcome and mortality in relation to oxygen levels. Apart from our study, only a few small, randomized trials on the topic exist. Taher et al. conducted a study involving 68 patients, in whom an FiO₂\u0026nbsp;of 80% administered for six hours post-injury resulted in more favorable long-term neurological outcomes compared with an FiO₂\u0026nbsp;of 50%, although there was an uncertain risk of bias across all domains.[14]\u0026nbsp;A recent randomized trial by Liu et al., including 110 patients with mild TBI, found that patients receiving high-flow oxygen or hyperbaric oxygen treatment had better neurological outcomes than those receiving low-flow oxygen. However, the patients were evaluated using a low sensitivity cognitive test that is not typically applied to trauma populations, and the control group had Mini Mental State Examination (MMSE) scores of 16–20, indicating moderate dementia rather than mild TBI.[13]\u0026nbsp;Thus, both trials demonstrated beneficial effects of increased levels of supplemental oxygen; however, the interventions in these studies differ from ours, making comparisons difficult. A systematic review on the topic published in 2021 also concluded that evidence on oxygenation strategies in TBI was extremely limited.[29]\u0026nbsp;A very large ongoing trial, Mega-ROX Brains, plans to randomize between 7500 and 9500 patients with nonhypoxic ischemic encephalopathy acute brain injuries and conditions to a conservative and liberal ICU oxygen therapy regimen; the anticipated date of last data collection early 2026.[30]\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eMeanwhile, observational evidence thus far has been mixed. Several studies have suggested that hyperoxia may be detrimental in TBI. A 2015 meta-analysis consisting of 24 studies found arterial hyperoxia during admission increased hospital mortality in critically ill patients, including those with TBI.[31] Of note, many of the studies reporting unfavorable outcomes, define severe hyperoxemia as a PaO₂ exceeding 200 mmHg or much more.[32–34] Consistent with our findings, recent retrospective studies reported no association between mild hyperoxemia (PaO₂\u0026nbsp;\u0026gt;100–120 mmHg) and unfavorable neurological outcome and/or mortality.[35–37]\u0026nbsp;An additional observational study demonstrated a U-shaped relationship between early arterial oxygenation and long-term functional and cognitive outcomes, with PaO₂\u0026nbsp;thresholds of 150 and 200 mmHg, associated with the most favorable outcomes.[38]\u0026nbsp;Moreover, a recent systematic review and meta-analysis including 13 studies, one randomized controlled trial (RCT) and two secondary analyses of RCTs evaluating other interventions, found no association between hyperoxia and mortality. For functional outcomes, pooled analyses suggested a potential benefit of hyperoxia when restricted to studies with low to moderate risk of bias. However, the overall certainty of evidence was rated as low for both outcomes according to GRADE, underscoring the need for the present study.[39]\u003c/p\u003e\n\u003cp\u003eImportantly, observational studies of oxygen supply are susceptible to simultaneity, a statistical phenomenon implying that a treatment that is administered at higher doses in more severely ill patients will almost invariably be erroneously interpreted as being detrimental to the patients, even if the analysis is adjusted for severity of illness by, e.g., inverse probability of treatment weighting. In other words, some observational studies may have overestimated the risk of hyperoxia, because it was administered to the most desperately ill patients.[40, 41]\u003c/p\u003e\n\u003cp\u003eIn our trial, the median PaO₂\u0026nbsp;in the restrictive group was 88 mmHg, lower than that reported in many prior studies, suggesting that targeting even lower PaO₂\u0026nbsp;levels may be reasonable; that said, the point estimate for the primary outcome calls for caution as 65% achieved a favorable outcome in the restrictive group compared to 70% in the liberal group. Notably, the liberal group achieved a median PaO₂\u0026nbsp;of 258 mmHg, yet no evidence of harm was observed at this level.\u003c/p\u003e\n\u003ch2\u003e\u003cstrong\u003eStrengths and Limitations\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eThis follow-up study has several notable strengths. First, its randomized, international, multicenter, and pragmatic design was intended to strengthen internal validity and external generalizability, supporting the interpretation of the findings in a representative trauma population.\u0026nbsp;\u003cbr\u003e\u0026nbsp;Second, the outcomes assessed in this study were pre-specified as follow-up outcomes of the TRAUMOX2 trial, reducing the risk of selective reporting. Further, this cohort achieved relatively high follow-up rates in a trauma setting. Finally, the analyses employed inverse probability weighting to mitigate potential attrition bias arising from differential follow-up.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eSeveral limitations must also be recognized. First, the choice of using GOSE as our primary outcome measure could be discussed. The score has been criticized for being a non-interval scale and dichotomizing the score into favorable and unfavorable outcome has been said to cause loss of information. The Functional Status Examination (FSE) has been suggested as an alternative and has shown to be related more strongly to severity of brain injury than the GOSE.[42] Nevertheless, to enhance comparability between studies, the GOSE score was reported, consistent with its use in most major TBI research to date.[43–45] Furthermore, the FSE requires more intensive training and clinical familiarity to achieve reliability. Second, neither participants, their relatives, nor the GOSE interviewer were blinded towards the oxygen treatment allocation, which could affect the evaluation of the participants’ functionality according to preexisting beliefs on oxygen treatment. Third, the definition of the TBI population in this study must also be discussed. In an attempt to include all relevant cases, one may risk including participants without TBI. Participants with AIS head=1 were not included to avoid cases with only minor scalp injuries. However, 14 participants in this follow-up study were marked with penetrating injury as the dominating injury type. The cases were manually double checked and all but two clearly suffered a TBI. The two doubtful cases were registered with “Cranial Nerve NSF” damage. The criteria for the TBI population for this follow-up study were defined in the publicly available protocol before last data collection and all 14 participants were included in the analyses.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eLastly, the study was conducted in European countries, and while the results may be generalizable to similar populations, caution should be exercised when interpreting the findings in other settings. Definitions and classifications of TBI vary considerably between studies and are sometimes insufficiently reported, limiting comparability across the literature.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003eDr Baekgaard reported receiving grants from Novo Nordisk Foundation and Lundbeck Foundation during the conduct of the study. Dr Dinesen reported receiving grants from Danish Air Ambulance and Rigshospitalets Forskningspuljer during the conduct of the study. Dr van Lieshout reported receiving grants from Rigshospitalet during the conduct of the study. Dr Hautz reported receiving grants from Swiss National Science Foundation and the European Union; and personal fees from AO Foundation Zurich, Mundipharma Switzerland, and MDI Australia outside the submitted work. Dr Klimek reported receiving grants to organization for research laboratory Music as Medicine; honoraria from Paion as a member of the data safety board and Tijdstroom uitgeverij for serving as an editor of a Dutch textbook; and serving as a course director for Advanced Trauma Life Support outside the submitted work. Dr Steinmetz reported receiving grants from Novo Nordisk Foundation during the conduct of the study; and funding for professorship from Norwegian Air Ambulance Foundation outside the submitted work. No other disclosures were reported\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eIn adults with traumatic brain injury, an early restrictive oxygen strategy did not significantly improve functional recovery at 12 months compared with a liberal oxygen strategy initiated in the prehospital setting or on trauma center admission for 8 hours.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003eAcknowledgments\u003c/p\u003e\n\u003cp\u003eWe thank the participants and their relatives, investigators, clinical staff, collaborators, funding bodies, regulatory authorities, and all others who contributed to the TRAUMOX2 trial.\u003c/p\u003e\n\u003cp\u003eAuthor Contributions:\u003c/p\u003e\n\u003cp\u003eDrs Steinmetz, Dinesen, Arleth, Baekgaard and Siersma had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eConcept and design:\u003c/em\u003e Baekgaard, Arleth, Dinesen, Siersma, and Steinmetz\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eAcquisition, analysis, or interpretation of data:\u003c/em\u003e All authors.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eDrafting of the manuscript:\u003c/em\u003e Baekgaard, Rosenkrantz, Arleth, Dinesen, Siersma, and Steinmetz.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eCritical review of the manuscript for important intellectual content:\u003c/em\u003e All authors.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eStatistical analysis:\u003c/em\u003e Baekgaard, Dinesen, Siersma, and Steinmetz.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eObtained funding:\u003c/em\u003e Baekgaard, Mikkelsen, Zwisler, Steinmetz.\u003c/p\u003e\n\u003cp\u003eFunding/support: The trial was supported by Novo Nordisk Foundation grant No. NNF20OC0063985. Furthermore, the trial was funded by the Joint Research Fund of Odense University Hospital and Rigshospitalet grant No. 136-A5566 and an individual grant to Dr Baekgaard from the Lundbeck Foundation.\u003c/p\u003e\n\u003cp\u003eRole of Funder/Sponsor: Novo Nordisk Foundation, the Joint Research Fund of Odense University Hospital and Rigshospitalet, and Lundbeck Foundation had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.\u003c/p\u003e\n\u003cp\u003eData Sharing statement: Please see supplemental 4 of the primary trial publication.[16] \u0026nbsp;\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eDewan MC, Rattani A, Gupta S, et al (2019) Estimating the global incidence of traumatic brain injury. J Neurosurg 130:1080\u0026ndash;1097. https://doi.org/10.3171/2017.10.JNS17352\u003c/li\u003e\n\u003cli\u003eGaudette \u0026Eacute;, Seabury SA, Temkin N, et al (2022) Employment and Economic Outcomes of Participants With Mild Traumatic Brain Injury in the TRACK-TBI Study. JAMA Netw Open 5:e2219444. https://doi.org/10.1001/jamanetworkopen.2022.19444\u003c/li\u003e\n\u003cli\u003eMalhotra AK, Jaffe RH, Shakil H, et al (2024) Unemployment and Personal Income Loss After Traumatic Brain Injury. JAMA Surg 159:1415\u0026ndash;1422. https://doi.org/10.1001/jamasurg.2024.4285\u003c/li\u003e\n\u003cli\u003eChesnut RM, Marshall LF, Klauber MR, et al (1993) The role of secondary brain injury in determining outcome from severe head injury. J Trauma 34:216\u0026ndash;222. https://doi.org/10.1097/00005373-199302000-00006\u003c/li\u003e\n\u003cli\u003eMaiga AW, Lin H-HS, Wisniewski SR, et al (2025) Adverse Prehospital Events and Outcomes After Traumatic Brain Injury. JAMA Netw Open 8:e2457506. https://doi.org/10.1001/jamanetworkopen.2024.57506\u003c/li\u003e\n\u003cli\u003eLulla A, Lumba-Brown A, Totten AM, et al (2023) Prehospital Guidelines for the Management of Traumatic Brain Injury - 3rd Edition. Prehosp Emerg Care 27:507\u0026ndash;538. https://doi.org/10.1080/10903127.2023.2187905\u003c/li\u003e\n\u003cli\u003eGaither JB, Spaite DW, Bobrow BJ, et al (2024) EMS Treatment Guidelines in Major Traumatic Brain Injury With Positive Pressure Ventilation. JAMA Surg 159:363\u0026ndash;372. https://doi.org/10.1001/jamasurg.2023.7155\u003c/li\u003e\n\u003cli\u003eACS TQP Best Practices Guidelines. In: ACS. https://www.facs.org/quality-programs/trauma/quality/best-practices-guidelines/. Accessed 15 Sep 2025\u003c/li\u003e\n\u003cli\u003eOverview | Head injury: assessment and early management | Guidance | NICE. https://www.nice.org.uk/guidance/ng232. Accessed 15 Sep 2025\u003c/li\u003e\n\u003cli\u003eBrenner M, Stein D, Hu P, et al (2012) Association between early hyperoxia and worse outcomes after traumatic brain injury. Arch Surg 147:1042\u0026ndash;1046. https://doi.org/10.1001/archsurg.2012.1560\u003c/li\u003e\n\u003cli\u003eRincon F, Kang J, Vibbert M, et al (2014) Significance of arterial hyperoxia and relationship with case fatality in traumatic brain injury: a multicentre cohort study. J Neurol Neurosurg Psychiatry 85:799\u0026ndash;805. https://doi.org/10.1136/jnnp-2013-305505\u003c/li\u003e\n\u003cli\u003eFornaciari A, Zangari R, Polato M, et al (2025) The impact of hyperoxemia on mortality and neurological outcomes in traumatic brain injury: a systematic review and meta-analysis. J Neurosurg Sci 69:481\u0026ndash;492. https://doi.org/10.23736/S0390-5616.25.06480-X\u003c/li\u003e\n\u003cli\u003eLiu Z, Wang X, Wu Z, et al (2023) HBOT has a better cognitive outcome than NBH for patients with mild traumatic brain injury: A randomized controlled clinical trial. Medicine (Baltimore) 102:e35215. https://doi.org/10.1097/MD.0000000000035215\u003c/li\u003e\n\u003cli\u003eTaher A, Pilehvari Z, Poorolajal J, Aghajanloo M (2016) Effects of Normobaric Hyperoxia in Traumatic Brain Injury: A Randomized Controlled Clinical Trial. Trauma Mon 21:e26772. https://doi.org/10.5812/traumamon.26772\u003c/li\u003e\n\u003cli\u003eL\u0026aring;ng M, Skrifvars MB, Siironen J, et al (2018) A pilot study of hyperoxemia on neurological injury, inflammation and oxidative stress. Acta Anaesthesiol Scand 62:801\u0026ndash;810. https://doi.org/10.1111/aas.13093\u003c/li\u003e\n\u003cli\u003eArleth T, Baekgaard J, Siersma V, et al (2025) Early Restrictive vs Liberal Oxygen for Trauma Patients: The TRAUMOX2 Randomized Clinical Trial. JAMA 333:479\u0026ndash;489. https://doi.org/10.1001/jama.2024.25786\u003c/li\u003e\n\u003cli\u003eCalvert M, Blazeby J, Altman DG, et al (2013) Reporting of Patient-Reported Outcomes in Randomized Trials: The CONSORT PRO Extension. JAMA 309:814\u0026ndash;822. https://doi.org/10.1001/jama.2013.879\u003c/li\u003e\n\u003cli\u003eArleth T, Baekgaard J, Siersma V, et al (2023) Comparing restrictive versus liberal oxygen strategies for trauma patients: The TRAUMOX2 trial-Statistical analysis plan. Acta Anaesthesiol Scand 67:829\u0026ndash;838. https://doi.org/10.1111/aas.14230\u003c/li\u003e\n\u003cli\u003eBaekgaard J, Arleth T, Siersma V, et al (2022) Comparing restrictive versus liberal oxygen strategies for trauma patients - the TRAUMOX2 trial: protocol for a randomised clinical trial. BMJ Open 12:e064047. https://doi.org/10.1136/bmjopen-2022-064047\u003c/li\u003e\n\u003cli\u003eRingdal KG, Hestnes M, Palmer CS (2012) Differences and discrepancies between 2005 and 2008 Abbreviated Injury Scale versions - time to standardise. Scand J Trauma Resusc Emerg Med 20:11. https://doi.org/10.1186/1757-7241-20-11\u003c/li\u003e\n\u003cli\u003eWilson L, Boase K, Nelson LD, et al (2021) A Manual for the Glasgow Outcome Scale-Extended Interview. J Neurotrauma 38:2435\u0026ndash;2446. https://doi.org/10.1089/neu.2020.7527\u003c/li\u003e\n\u003cli\u003eNelson LD, Temkin NR, Barber J, et al (2023) Functional Recovery, Symptoms, and Quality of Life 1 to 5 Years After Traumatic Brain Injury. JAMA Netw Open 6:e233660. https://doi.org/10.1001/jamanetworkopen.2023.3660\u003c/li\u003e\n\u003cli\u003eUser Guides. In: EuroQol. https://euroqol.org/information-and-support/documentation/user-guides/. Accessed 15 Dec 2025\u003c/li\u003e\n\u003cli\u003eMcCrea MA, Giacino JT, Barber J, et al (2021) Functional Outcomes Over the First Year After Moderate to Severe Traumatic Brain Injury in the Prospective, Longitudinal TRACK-TBI Study. JAMA Neurol 78:982\u0026ndash;992. https://doi.org/10.1001/jamaneurol.2021.2043\u003c/li\u003e\n\u003cli\u003evan den Brink WA, van Santbrink H, Steyerberg EW, et al (2000) Brain oxygen tension in severe head injury. Neurosurgery 46:868\u0026ndash;876; discussion 876-878. https://doi.org/10.1097/00006123-200004000-00018\u003c/li\u003e\n\u003cli\u003eAbate MG, Trivedi M, Fryer TD, et al (2008) Early derangements in oxygen and glucose metabolism following head injury: the ischemic penumbra and pathophysiological heterogeneity. Neurocrit Care 9:319\u0026ndash;325. https://doi.org/10.1007/s12028-008-9119-2\u003c/li\u003e\n\u003cli\u003eTolias CM, Reinert M, Seiler R, et al (2004) Normobaric hyperoxia--induced improvement in cerebral metabolism and reduction in intracranial pressure in patients with severe head injury: a prospective historical cohort-matched study. J Neurosurg 101:435\u0026ndash;444. https://doi.org/10.3171/jns.2004.101.3.0435\u003c/li\u003e\n\u003cli\u003eNortje J, Coles JP, Timofeev I, et al (2008) Effect of hyperoxia on regional oxygenation and metabolism after severe traumatic brain injury: preliminary findings. Crit Care Med 36:273\u0026ndash;281. https://doi.org/10.1097/01.CCM.0000292014.60835.15\u003c/li\u003e\n\u003cli\u003eHansen TE, Christensen RE, Baekgaard J, et al (2022) Supplemental oxygen for traumatic brain injury: A systematic review. Acta Anaesthesiol Scand 66:307\u0026ndash;316. https://doi.org/10.1111/aas.14019\u003c/li\u003e\n\u003cli\u003eYoung PJ, Al-Fares A, Aryal D, et al (2023) Protocol and statistical analysis plan for the mega randomised registry trial comparing conservative vs. liberal oxygenation targets in adults with nonhypoxic ischaemic acute brain injuries and conditions in the intensive care unit (Mega-ROX Brains). Crit Care Resusc J Australas Acad Crit Care Med 25:53\u0026ndash;59. https://doi.org/10.1016/j.ccrj.2023.04.011\u003c/li\u003e\n\u003cli\u003eHelmerhorst HJF, Roos-Blom M-J, van Westerloo DJ, de Jonge E (2015) Association Between Arterial Hyperoxia and Outcome in Subsets of Critical Illness: A Systematic Review, Meta-Analysis, and Meta-Regression of Cohort Studies. Crit Care Med 43:1508\u0026ndash;1519. https://doi.org/10.1097/CCM.0000000000000998\u003c/li\u003e\n\u003cli\u003eRincon F, Kang J, Vibbert M, et al (2014) Significance of arterial hyperoxia and relationship with case fatality in traumatic brain injury: a multicentre cohort study. J Neurol Neurosurg Psychiatry 85:799\u0026ndash;805. https://doi.org/10.1136/jnnp-2013-305505\u003c/li\u003e\n\u003cli\u003eBrenner M, Stein D, Hu P, et al (2012) Association between early hyperoxia and worse outcomes after traumatic brain injury. Arch Surg Chic Ill 1960 147:1042\u0026ndash;1046. https://doi.org/10.1001/archsurg.2012.1560\u003c/li\u003e\n\u003cli\u003eDavis DP, Meade W, Sise MJ, et al (2009) Both hypoxemia and extreme hyperoxemia may be detrimental in patients with severe traumatic brain injury. J Neurotrauma 26:2217\u0026ndash;2223. https://doi.org/10.1089/neu.2009.0940\u003c/li\u003e\n\u003cli\u003eLalla LT, Czorlich P, Fischer M, et al (2025) Dose-dependent association of hyperoxia and decreased favorable outcomes in mechanically ventilated patients with traumatic brain injury, a retrospective cohort study. Eur J Trauma Emerg Surg Off Publ Eur Trauma Soc 51:75. https://doi.org/10.1007/s00068-024-02730-5\u003c/li\u003e\n\u003cli\u003eKhan R, Alromaih S, Alshabanat H, et al (2021) The Impact of Hyperoxia Treatment on Neurological Outcomes and Mortality in Moderate to Severe Traumatic Brain Injured Patients. J Crit Care Med Univ Med Si Farm Din Targu-Mures 7:227\u0026ndash;236. https://doi.org/10.2478/jccm-2021-0014\u003c/li\u003e\n\u003cli\u003eRaj R, Bendel S, Reinikainen M, et al (2013) Hyperoxemia and long-term outcome after traumatic brain injury. Crit Care Lond Engl 17:R177. https://doi.org/10.1186/cc12856\u003c/li\u003e\n\u003cli\u003eAlali AS, Temkin N, Vavilala MS, et al (2020) Matching early arterial oxygenation to long-term outcome in severe traumatic brain injury: target values. J Neurosurg 132:537\u0026ndash;544. https://doi.org/10.3171/2018.10.JNS18964\u003c/li\u003e\n\u003cli\u003eFornaciari A, Zangari R, Polato M, et al (2025) The impact of hyperoxemia on mortality and neurological outcomes in traumatic brain injury: a systematic review and meta-analysis. J Neurosurg Sci 69:481\u0026ndash;492. https://doi.org/10.23736/S0390-5616.25.06480-X\u003c/li\u003e\n\u003cli\u003ede Grooth H-J, Girbes ARJ, van der Ven F, et al (2020) Observational Research for Therapies Titrated to Effect and Associated With Severity of Illness: Misleading Results From Commonly Used Statistical Methods. Crit Care Med 48:1720\u0026ndash;1728. https://doi.org/10.1097/CCM.0000000000004612\u003c/li\u003e\n\u003cli\u003eLeisman DE (2020) The Goldilocks Effect in the ICU-When the Data Speak, but Not the Truth. Crit Care Med 48:1887\u0026ndash;1889. https://doi.org/10.1097/CCM.0000000000004669\u003c/li\u003e\n\u003cli\u003eDikmen S, Machamer J, Manley GT, et al (2019) Functional Status Examination versus Glasgow Outcome Scale Extended as Outcome Measures in Traumatic Brain Injuries: How Do They Compare? J Neurotrauma 36:2423\u0026ndash;2429. https://doi.org/10.1089/neu.2018.6198\u003c/li\u003e\n\u003cli\u003eRoberts I, Yates D, Sandercock P, et al (2004) Effect of intravenous corticosteroids on death within 14 days in 10008 adults with clinically significant head injury (MRC CRASH trial): randomised placebo-controlled trial. Lancet Lond Engl 364:1321\u0026ndash;1328. https://doi.org/10.1016/S0140-6736(04)17188-2\u003c/li\u003e\n\u003cli\u003eMaas AIR, Menon DK, Steyerberg EW, et al (2015) Collaborative European NeuroTrauma Effectiveness Research in Traumatic Brain Injury (CENTER-TBI): a prospective longitudinal observational study. Neurosurgery 76:67\u0026ndash;80. https://doi.org/10.1227/NEU.0000000000000575\u003c/li\u003e\n\u003cli\u003eMcCrea MA, Giacino JT, Barber J, et al (2021) Functional Outcomes Over the First Year After Moderate to Severe Traumatic Brain Injury in the Prospective, Longitudinal TRACK-TBI Study. JAMA Neurol 78:982\u0026ndash;992. https://doi.org/10.1001/jamaneurol.2021.2043\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Tables","content":"\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" colspan=\"3\"\u003e\n \u003cp\u003e\u003cstrong\u003eTable 1. Patient Characteristics\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e\u003cstrong\u003eCharacteristics\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e\u003cstrong\u003eRestrictive oxygen group\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e(N=352)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e\u003cstrong\u003eLiberal oxygen group\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e(N=322)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eAge (years), \u003cem\u003emedian [IQR]\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e56 (34\u0026ndash;68)\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e(n=352)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e54 (38\u0026ndash;70)\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e(n=322)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eSex, \u003cem\u003en/total (%)\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp;Male\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e251/352 (71.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e221/322 (68.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eComorbidities prior to trauma, \u003cem\u003en/total (%)\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e176/352 (51.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e154/322 (48.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp;Cardiovascular disease\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e83/345 (24.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e69/318 (21.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp;Lung disease\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e29/345 (8.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e37/318 (11.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp;Other\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e128/345 (37.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e115/318 (36.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; Psychiatric disorder\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e32/345 (9.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e30/318 (9.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eType of injury, \u003cem\u003en/total (%)\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003ePenetrating\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp;Shot by firearm of any caliber\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e3/346 (0.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e1/322 (0.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp;Stabbed\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e8/346 (2.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e2/322 (0.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eBlunt\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp;Traffic accident\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e198/346 (57.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e184/322 (57.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp;Hit by blunt object\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e17/346 (4.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e25/322 (7.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp;Fall\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e114/346 (32.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e100/322 (31.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp;Blast/explosion\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e2/346 (0.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e1/322 (0.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp;Other\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e4/346 (1.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e9/322 (2.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eIntubated at randomization, \u003cem\u003en/total (%)\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e134/352 (38.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e122/322 (37.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eSite of inclusion, \u003cem\u003en/total (%)\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eInhospital\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e205/352 (58.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e179/322 (55.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003ePrehospital\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e147/352 (41.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e143/322\u003c/p\u003e\n \u003cp\u003e(44.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eCountry of inclusion, \u003cem\u003en/total (%)\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eDenmark\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e280/352 (79.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e267/322 (82.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eThe Netherlands\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e45/352 (12.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e33/322 (10.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eSwitzerland\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e27/352 (7.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e22/322 (6.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e\u003cstrong\u003ePrehospital information\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eUse of prehospital or in-hospital supplemental oxygen prior to randomization, \u003cem\u003en/total (%)\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e192/331 (58.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e164/308 (53.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eTime with supplemental oxygen treatment prior to randomization (min), \u003cem\u003emedian [IQR]\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e30 (18\u0026ndash;50)\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e(n=152)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e30 (18\u0026ndash;53)\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e(n=128)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e\u003cstrong\u003eFirst vital signs\u003c/strong\u003e, \u003cem\u003emedian [IQR]\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eSystolic blood pressure (mmHg)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e133 (113\u0026ndash;155) (n=352)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e135 (119\u0026ndash;154) (n=320)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eHeart rate (beats/min)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e83 (69\u0026ndash;102)\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e(n=351)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e85 (72\u0026ndash;101)\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e(n=320)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eRespiratory rate (breaths/min)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e20 (16\u0026ndash;24)\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e(n=350)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e20 (16\u0026ndash;24)\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e(n=313)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eSpO\u003csub\u003e2\u003c/sub\u003e\u0026lt;90%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e56/352 (15.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e46/320 (14.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eGlasgow Coma Scale (GCS) score\u0026lt;9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e111/352 (31.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e112/322 (34.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eInjury Severity Scale (ISS) score\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e17 (11\u0026ndash;27)\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e(n=352)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e17 (10\u0026ndash;27)\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e(n=322)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e\u003cstrong\u003eOxygen intervention characteristics\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003ePaO\u003csub\u003e2\u003c/sub\u003e (mmHg)\u003csup\u003ea\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e88 (73\u0026ndash;114) (n=305)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e258 (138\u0026ndash;399) (n=278)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eHemoglobin (mmol/l)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e7.8 (7.1\u0026ndash;8.5)\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e(n=300)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e7.8 (6.8\u0026ndash;8.7)\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e(n=266)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eType of supplemental oxygen, \u003cem\u003en/total (%)\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp;No supplemental oxygen\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e85/348 (24.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e8/318 (2.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp;Nasal cannula\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e67/348 (19.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e16/318 (5.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp;Non-rebreather mask\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e8/348 (2.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e133/318 (41.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp;Intubated (at any point during the intervention period)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e188/348 (54.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e161/318 (50.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cem\u003eCharacteristics of trauma patients with TBI who were allocated to receive either a restrictive or a liberal oxygen strategy for 8 hours after randomization. Participants were randomized either in the prehospital setting or upon admission to the trauma resuscitation room and were followed for the first year after randomization.\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eAbbreviations: SpO\u003csub\u003e2\u003c/sub\u003e, arterial oxygen saturation measured by pulse oximetry; PaO\u003csub\u003e2\u003c/sub\u003e, partial pressure of oxygen in arterial blood.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\u003csup\u003ea\u003c/sup\u003e\u003c/em\u003eArterial blood gases were obtained at hour 1\u0026plusmn;30 minutes after randomization.\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"100%\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"7\"\u003e\n \u003cp\u003e\u003cstrong\u003eTable 2. Primary and Secondary Outcomes\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eRestrictive Oxygen Group, No./Total (%)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eLiberal\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003eoxygen group, No./Total (%)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eRisk difference, %, (95% CI)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;value\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eAdjusted risk difference, %,\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e(95% CI)\u003cem\u003e\u003csup\u003e\u0026nbsp;a\u003c/sup\u003e\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;value\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"7\"\u003e\n \u003cp\u003e\u003cstrong\u003ePrimary analysis\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eGOSE favorable recovery (5\u0026ndash;8) at 12 months\u003c/p\u003e\u0026nbsp;\u0026nbsp;\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e196/302 (65)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e185/266 (70)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e4.6 (-2.6 to 11.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eP =.21\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e4.3 (-2.1 to 10.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eP =.18\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"7\"\u003e\n \u003cp\u003e\u003cstrong\u003eSecondary analyses\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eGOSE favorable recovery (5\u0026ndash;8) at 6 months\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e198/303 (65)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e186/265 (70)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e5.0 (-2.3 to 12.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eP =.18\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e4.9 (-1.5 to 11.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eP =.13\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eGOSE complete recovery (8) at 6 months\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e49/303 (16)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e39/265 (15)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-2.1 (-7.7 to 3.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eP =.46\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-1.2 (-6.8 to 4.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eP =.67\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eGOSE complete recovery (8) at 12 months\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e64/302 (21)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e45/266 (17)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-4.3 (-10.5 to 1.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eP =.17\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-4.4 (-10.6 to 1.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eP =.16\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eDeath at 6 months\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e58/345 (17)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e47/305 (15)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-1.4 (-6.2 to 3.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eP =.56\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.9 (-5.6 to 3.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eP =.70\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eDeath at 12 months\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e66/343 (19)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e50/304 (16)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-2.3 (-7.4 to 2.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eP =.39\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-2.2 (-7.2 to 2.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eP =.39\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eRestrictive Oxygen Group, Median [IQR]\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eLiberal oxygen group, Median [IQR]\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eMean Difference, (95% CI)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;value\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eAdjusted mean difference (95% CI)\u003cem\u003e\u003csup\u003e\u0026nbsp;a\u003c/sup\u003e\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;value\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eGOSE non-dichotomized at 6 months\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e5 [3\u0026ndash;7]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e5 [4\u0026ndash;7]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.10 (-0.27 to 0.46)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eP =.60\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.10 (-0.20 to 0.41)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eP =.50\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eGOSE non-dichotomized at 12 months\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e6 [3\u0026ndash;7]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e6 [4\u0026ndash;7]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.14 (-0.25 to 0.53)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eP =.48\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.12 (-0.21 to 0.46)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eP =.47\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eEQ-5D-5L\u003cem\u003e\u003csup\u003eb\u003c/sup\u003e\u003c/em\u003e index value at 6 months after trauma\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.85 [0.58\u0026ndash;0.95]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.83 [0.58\u0026ndash;0.95]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.02 (-0.08 to 0.04)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eP =.51\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.01 (-0.07 to 0.05)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eP =.72\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eEQ-5D-5L\u003cem\u003e\u003csup\u003eb\u003c/sup\u003e\u003c/em\u003e index value at 12 months after trauma\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.88 [0.64\u0026ndash;0.97]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.83 [0.58\u0026ndash;0.95]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.03 (-0.09 to 0.02)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eP =.24\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.03 (-0.08 to 0.03)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eP =.34\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eEQ VAS\u003cem\u003e\u003csup\u003ec\u003c/sup\u003e\u003c/em\u003e at 6 months after trauma\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e75 [50\u0026ndash;85]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e70 [50\u0026ndash;85]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.82 (-5.0 to 3.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eP =.70\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-0.44 (-4.4 to 3.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eP =.83\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eEQ VAS\u003cem\u003e\u003csup\u003ec\u003c/sup\u003e\u003c/em\u003e at 12 months after trauma\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e75 [50\u0026ndash;90]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e70 [50\u0026ndash;85]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-2.3 (-6.5 to 1.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eP =.28\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-2.0 (-6.0 to 2.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eP =.34\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cem\u003eFunctional outcomes, mortality, and health-related quality of life assessed in trauma patients assigned to either a restrictive or liberal oxygen strategy for the first 8 hours after randomization. Participants were randomized in the prehospital setting or upon arrival in the trauma resuscitation room and were followed for one year after randomization.\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eAbbreviations: GOSE, Glasgow Outcome Scale \u0026ndash; Extended\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\u003csup\u003ea\u003c/sup\u003e\u003c/em\u003eAdjusted for the stratification variables (site of inclusion and status of endotracheal intubation upon inclusion) and the following variables: age, sex, Injury Severity Scale (ISS) score, predominant type of injury, and first recorded Glasgow Coma Scale (GCS) score.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\u003csup\u003eb\u003c/sup\u003e\u003c/em\u003e EQ-5D-5L is a generic instrument that reflects the health-related quality of life. The questionnaire contains five domains, and each domain has five levels of response: no problems, slight problems, moderate problems, severe problems, extreme problems/unable to.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\u003csup\u003ec\u003c/sup\u003e\u003c/em\u003e EQ-5D VAS range from 0 to 100, from worst imaginable health to best imaginable health, and reflect the participant\u0026rsquo;s perceived health. Participants or a relative were interviewed by telephone to obtain the scores. Up to five attempts for contact were made within each timeframe.\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"100%\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"7\"\u003e\n \u003cp\u003e\u003cstrong\u003eTable 3. Subgroup Analyses of GOSE Favorable Recovery (Scores 5\u0026ndash;8) at 12 Months\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eRestrictive oxygen group,\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003eNo./Total (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eLiberal\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003eoxygen group,\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003eNo./Total (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eRisk difference estimate\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e(95% CI)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;value\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eAdjusted risk difference\u003cem\u003e\u003csup\u003ea\u003c/sup\u003e\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e(95% CI)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;value\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eSevere TBI\u0026nbsp;\u003cbr\u003e\u0026nbsp;(AIS head\u0026ge;5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e13/52 (25)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e19/52 (37)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e12.6 (-4.7 to 30.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eP = .15\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e11.3 (-4.1 to 26.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eP =.15\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eModerate to severe TBI (AIS head\u0026ge;3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e117/203 (58)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e98/165 (59)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2.7 (-7.1 to 12.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eP =.59\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e33.1 (-5.0 to 11.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eP =.43\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eFirst registered GCS\u0026le;8\u003csup\u003eb\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e46/102 (45)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e46/90 (51)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e7.0 (-7.1 to 21.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eP =.33\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e8.0 (-3.8 to 19.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eP =.18\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eFirst registered GCS\u0026le;13\u003csup\u003eb\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e106/182 (58)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e90/155 (58)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.6 (-9.5 to 10.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eP =.90\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e4.6 (-3.8 to 13.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eP =.28\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cem\u003eSubgroup analyses on participants who had a favorable recovery at 12 months, defined as a GOSE score of 5\u0026ndash;8. The subgroups included patients with severe TBI (AIS head\u0026ge; 5), patients with moderate to severe TBI (AIS head\u0026ge; 3), patients with a GCS\u0026le; 8 (first recorded GCS in the prehospital setting or trauma bay), and patients with a GCS\u0026le; 13 (first recorded GCS in the prehospital setting or trauma resuscitation room). Participants were assigned to either a restrictive or a liberal oxygen strategy for the first 8 hours after randomization. Randomization occurred either in the prehospital setting or upon arrival in the trauma resuscitation room, and participants were followed for one year thereafter.\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eAbbreviations: GOSE, Glasgow Outcome Scale Extended; TBI, traumatic brain injury; AIS, abbreviated injury scale; GCS; Glasgow Coma Scale\u003c/p\u003e\n\u003cp\u003e\u003csup\u003ea\u003c/sup\u003eAdjusted for the stratification variables (site of inclusion and status of endotracheal intubation upon inclusion) and the following variables: age, sex, Injury Severity Scale (ISS) score, predominant type of injury, and first recorded Glasgow Coma Scale (GCS) score.\u003c/p\u003e\n\u003cp\u003e\u003csup\u003eb\u003c/sup\u003eThe first registered GCS could be either in the prehospital phase or trauma bay\u003cbr clear=\"all\"\u003e\u0026nbsp;\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" colspan=\"3\"\u003e\n \u003cp\u003e\u003cstrong\u003eTable 4. Glasgow Outcome Scale\u0026ndash;Extended (GOSE) Response Characteristics\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e\u003cstrong\u003eCharacteristics\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e\u003cstrong\u003eRestrictive oxygen group\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e(N=352)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e\u003cstrong\u003eLiberal oxygen group\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e(N=322)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eResponse rate at 6 months, \u003cem\u003en/total (%)\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eDenmark\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e197/352 (56.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e186/322 (57.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e37/352 (10.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e40/322 (12.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eDead\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e46/352 (13.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e37/322 (11.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eWithdrawn consent\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e0/352 (0.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e4/322 (1.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eLost to follow-up\u003cem\u003e\u003csup\u003e\u0026nbsp;a\u003c/sup\u003e\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e0/352 (0.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e0/322 (0.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eThe Netherlands\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e29/352 (8.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e21/322 (6.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e6/352 (1.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e6/322 (1.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eDead\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e7/352 (2.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e5/322 (1.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eWithdrawn consent\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e2/352 (0.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e0/322 (0.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eLost to follow-up\u003cem\u003e\u003csup\u003e\u0026nbsp;a\u003c/sup\u003e\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e1/352 (0.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e1/322 (0.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eSwitzerland\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e19/352 (5.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e11/322 (3.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e3/352 (0.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e6/322 (1.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eDead\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e5/352 (1.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e5/322 (1.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eWithdrawn consent\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e0/352 (0.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e0/322 (0.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eLost to follow-up\u003cem\u003e\u003csup\u003e\u0026nbsp;a\u003c/sup\u003e\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e0/352 (0.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e0/322 (0.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eResponse rate at 12 months, \u003cem\u003en/total (%)\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eDenmark\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e193/352 (54.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e181/322 (56.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e34/352 (9.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e43/322 (13.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eDead\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e53/352 (15.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e39/322 (12.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eWithdrawn consent\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e0/352 (0.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e4/322 (1.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eLost to follow-up\u003cem\u003e\u003csup\u003e\u0026nbsp;a\u003c/sup\u003e\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e0/352 (0.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e0/322 (0.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eThe Netherlands\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e28/352 (8.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e22/322 (6.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e6/352 (1.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e5/322 (1.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eDead\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e8/352 (2.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e5/322 (1.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eWithdrawn consent\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e2/352 (0.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e0/322 (0.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eLost to follow-up\u003cem\u003e\u003csup\u003e\u0026nbsp;a\u003c/sup\u003e\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e1/352 (0.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e1/322 (0.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eSwitzerland\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e15/352 (4.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e13/322 (4.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e7/352 (2.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e3/322 (0.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eDead\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e5/352 (1.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e6/322 (1.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eWithdrawn consent\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e0/352 (0.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e0/322 (0.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eLost to follow-up\u003cem\u003e\u003csup\u003e\u0026nbsp;a\u003c/sup\u003e\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e0/352 (0.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e0/322 (0.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eNo response at 6 months, reason, \u003cem\u003en/total no-response (%)\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eNo answer despite several attempts\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e28/46 (60.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e25/52 (48.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eMissing contact information\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e7/46 (15.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e13/52 (25.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eDid not wish to participate\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e9/46 (19.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e10/52 (19.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eOther\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e2/46 (4.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e4/52 (7.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eUnknown\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e0/46 (0.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e0/52 (0.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eNo response at 12 months, reason, \u003cem\u003en/total no-response (%)\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eNo answer despite several attempts\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e23/47 (48.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e24/51 (47.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eMissing contact information\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e8/47 (17.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e14/51 (27.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eDid not wish to participate\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e14/47 (29.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e10/51 (19.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eOther\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e2/47 (4.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e2/51 (3.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eUnknown\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e0/47 (0.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e1/51 (2.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eRespondent at 6 months,\u003cem\u003e\u0026nbsp;n/total responders (%)\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eParticipant\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e217/245 (88.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e197/218 (90.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eNext-of-kin/friend/caretaker\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e21/245 (8.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e15/218 (6.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eParticipant and next-of-kin/friend/caretaker\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e7/245 (2.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e6/218 (2.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eUnknown\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e0/245 (0.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e0/218 (0.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eRespondent at 12 months,\u003cem\u003e\u0026nbsp;n/total responders (%)\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eParticipant\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e205/236 (86.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e192/217 (88.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eNext-of-kin/friend/caretaker\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e26/236 (11.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e21/217 (9.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eParticipant and next-of-kin/friend/caretaker\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e5/236 (2.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e4/217 (1.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eUnknown\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e0/236 (0.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e0/217 (0.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eContact established at 6 months (days), \u003cem\u003emedian [IQR]\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e189 (185\u0026ndash;200) (n=245)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e190 (185\u0026ndash;199) (n=218)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003eContact established at 12 months (days), \u003cem\u003emedian [IQR]\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e371 (367\u0026ndash;378) (n=236)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\"\u003e\n \u003cp\u003e371 (367\u0026ndash;378) (n=217)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cem\u003eCharacteristics of participants assessed using the Glasgow Outcome Scale\u0026ndash;Extended (GOSE). Participants were assigned to either a restrictive or a liberal oxygen strategy for the first 8 hours after randomization. Randomization occurred either in the prehospital setting or upon arrival in the trauma resuscitation room, and participants were followed for one year thereafter.\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\u003csup\u003ea\u003c/sup\u003e\u003c/em\u003eLost to follow-up represents participants where contact was unsuccessful e.g. due to invalid phone number or e-mail address recorded in their electronic health record\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Intensive care, oxygen, trauma, traumatic brain injury","lastPublishedDoi":"10.21203/rs.3.rs-9423108/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9423108/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003ePurpose: \u003c/strong\u003eHyperoxia in the initial phase of trauma has been associated with harm in patients with traumatic brain injury (TBI), but relevant large, randomized trials are lacking. This study aimed to determine whether TBI patients treated with a restrictive compared with a liberal oxygen strategy had better functional outcome 12 months after TBI.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethods: \u003c/strong\u003eThis preplanned follow-up study of the TRAUMOX2 randomized clinical trial included adult TBI patients from Denmark, the Netherlands, and Switzerland between December 2021, and September 2023. TBI was defined as an Abbreviated Injury Scale (AIS) head score of ≥2 or AIS-defined concussion.\u003cbr\u003e\nParticipants were randomly assigned to a restrictive oxygen strategy (target arterial oxygen saturation of 94%) or a liberal oxygen strategy (12–15 L/min oxygen or a fraction of inspired oxygen of 0.6–1.0) for 8 hours. The primary outcome was a favorable outcome at 12 months after trauma, defined as a Glasgow Outcome Scale–Extended (GOSE) score of 5-8.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults:\u003c/strong\u003e Among 674 eligible patients, 106 were excluded during follow-up, leaving 568 for analysis. The primary analysis revealed no significant difference between the restrictive and liberal oxygen group in favorable outcome at 12 months (65% vs 70%; risk difference, 4.6 percentage points; 95% CI, −2.6 to 11.7; P = .21). Adjusted analyses showed similar results (risk difference, 4.3 percentage points; 95% CI, −2.1 to 10.7; P = .18).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusions: \u003c/strong\u003eIn adult TBI patients, an early 8-hour restrictive oxygen strategy did not significantly improve functional outcome at 12 months compared to a liberal strategy.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTrial registration:\u003c/strong\u003e ClinicalTrials.gov Identifier: NCT05146700\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding/Support:\u003c/strong\u003e This trial was supported by Novo Nordisk Foundation grant No. NNF20OC0063985. Additionally, the trial was funded by the Joint Research Fund of Odense University Hospital and Rigshospitalet grant No. 136-A5566 and an individual grant to Dr Baekgaard from the Lundbeck Foundation.\u003c/p\u003e","manuscriptTitle":"Initial restrictive versus liberal oxygen for traumatic brain injury: \nA follow-up study of the TRAUMOX2 randomized trial","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-26 15:38:31","doi":"10.21203/rs.3.rs-9423108/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"19aea7e3-e408-4d7e-adda-940dceb25509","owner":[],"postedDate":"April 26th, 2026","published":true,"recentEditorialEvents":[{"type":"decision","content":"Reject after review","date":"2026-05-09T14:37:21+00:00","index":"","fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-05-09T18:39:37+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-26 15:38:31","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9423108","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9423108","identity":"rs-9423108","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}