Accepted on 9 Feb 2020
Submitted on 26 Jan 2020
The Global Burden of Disease study estimates that 27 million new cases of traumatic brain injury occur worldwide every year [1]. According to data from the US National Trauma Data Bank, almost 4% of these will require neurosurgical intervention [2]. Timely and adequate emergency surgery for TBI has long been known to dramatically improve outcome [3], [4], [5], [6]. As such, multiple global health organisations recognise surgical procedures for neurotrauma, specifically the evacuation of traumatic intracranial haematomas and the elevation of depressed skull fractures, as essential to be available on an emergency basis to all worldwide [7], [8]. However, although most injuries occur in LMICs [9], access to neurosurgical care in many LMICs is extremely limited – in sub-Saharan Africa, the average percentage of a country’s population within 2 h of a neurosurgical service is estimated to be 25% [10].
For patients with neurosurgical pathology who do reach a centre that can provide the procedure they require, there is a paucity of data on their outcomes and the standard of management they receive [11]. To complicate matters, current best practise guidelines for the management of TBI typically advocate the use of expensive technologies (such as invasive intracranial pressure monitoring) which are often unavailable in low-resource settings [12], [13] despite the fact that this is where the majority of serious injuries occur globally [9]. As such, context-specific research is urgently required to determine best practice for delivering effective surgical care for traumatic neurological injuries globally. The first step in this is to describe current practice and outcomes across varied settings and neurosurgical care is an extremely important aspect of the management of TBI patients.
We propose to conduct an international, multi-centre, prospective cohort study of outcomes following emergency surgery for traumatic brain injury. We believe such a project would provide valuable information to advance the care of patients with traumatic brain injuries globally and be feasible with the proposed methodology.
The primary aim of the study is to compare 14-day mortality (or in-hospital mortality if discharge occurs before 14 days) following emergency surgery for traumatic brain injury (TBI) across HDI settings.
To describe and compare the following across HDI settings following emergency surgery for TBI:
This is a multicentre, international, prospective, observational cohort study of all patients receiving emergency surgery for TBI in any consecutive 30-day period. Local study teams may select any 30-day period from the 1st of November 2018 and the 31st of December 2019 to start their study. Patients operated on who meet the inclusion criteria between 00:01 on day 0 and 23:59 on day 30 of the selected study period will be included. The final study period will therefore begin on 31st December 2019 00:01 and end on 30th January 23:59. Teams must follow patients up for 14-days from the date of surgery or until they are discharged or die – whichever of these comes first. Final day of data collection for the primary outcome measure is therefore February 13th. Teams have 30 days after the end of their selected study period to complete the data validation exercise (see below) – all patient level data should therefore be submitted to the central study team by 29th February.
Any hospital or clinic worldwide performing emergency surgery for traumatic brain injury is eligible to participate. In the majority of institutions, emergency surgery for TBI is provided by neurosurgeons – however, centres in which emergency surgery for TBI is provided by general surgeons, trauma surgeons, general medical doctors or even non-physician clinicians are also eligible to participate.
Table 1 summarises the patient inclusion and exclusion criteria.
Table 1
Patient inclusion and exclusion criteria.
The primary outcome measure will be 14-day mortality (or in-hospital mortality if discharge occurs before 14 days).
A data set will be collected on all adult or paediatric patients receiving emergency surgery for TBI in the inclusion period (see Table 2, Table 3, Table 4, Table 5). The case report form is split into 3 sub-sections – injury/admission data, operative data and outcome data. The data fields were based on existing data sets and refined through iterative consultation with an interdisciplinary consortium of clinicians caring for TBI across a range of HDI settings.
Table 2
Injury/admission data.
Table 3
CT data (if answered ‘Yes’ to ‘Was a CT head performed before the operation?’ in Table 2).
| Date/time of first available CT head for review | … | |
| Midline shift in mm? | 0–5 mm, 5–10 mm, >10 mm | |
| Basal cisterns | Open, Compressed/obliterated, Unknown | |
| Traumatic subarachnoid haemorrhage? | Yes, No | |
| Is there a depressed skull fracture? | Yes, No If yes, are any of the fragments depressed > 1 cm relative to the rest of the skull on CT? Yes, No If yes, is there any evidence of pneumocephalus on CT? Yes, No If yes, is there evidence of a compound depressed skull fracture on examination of the patient? Yes, No |
|
| Left | Right | |
| Supratentorial extradural haematoma? | No, 0–10 mm, 10–20 mm, >20 mm, Unknown | No, 0–10 mm, 10–20 mm, >20 mm, Unknown |
| Supratentorial subdural haematoma? | No, 0–10 mm, 10–20 mm, >20 mm, Unknown | No, 0–10 mm, 10–20 mm, >20 mm, Unknown |
| Supratentorial traumatic parenchymal lesion | No, Small, Large - >50 cm3 volume, Unknown | No, Small, Large - >50 cm3 volume, Unknown |
| Is there a traumatic posterior fossa haemorrhage? | Yes, No | |
Table 4
Operative data.
Table 5
Outcomes data.
| Was the patient admitted to intensive care after the operation at any point during the 14-day follow up period? | Yes, No If yes, date of admission to ICU … If yes, was the patient discharged from ICU during the 14-day follow up period? Yes, No If yes, date of discharge from ICU … |
| Surgical site infection | Yes, No If yes, was it a superficial/deep incisional infection (i.e. scalp wound) or organ/space infection (i.e. bone flap osteitis, meningitis and/or empyema/abscess)? Superficial or deep incisional infection, Organ/space infection If yes, how was it diagnosed? Tick all that apply. Symptoms, Signs on examination, Blood tests, Imaging, Wound swab microscopy, Wound swab Gram stain, Wound swab culture, CSF microscopy, CSF Gram stain, CSF culture, Surgical diagnosis |
| Did the patient return to the operating theatre for cranial surgery during the current admission? | Yes, No If yes, what was the re-operation? Re-evacuation of ipsilateral haematoma, Infection - wound washout, Infection - removal of bone flap, Craniectomy, Cranioplasty, Evacuation of contralateral haematoma, Other neurosurgical procedure |
| Did the patient survive to the end of the follow up period (14 days post-operatively or until they were discharged from hospital, whichever came first)? | Yes, No Was the patient still admitted to hospital on the 14th day post-operatively? Yes, No |
| Date of death or discharge from your institution (as applicable) | … |
| Discharge destination (if applicable) | Transfer to another hospital, Transfer to rehabilitation unit, Usual place of residence, Absconded, Other (state) |
| Glasgow Coma Scale at discharge or at the end of the follow up period if not discharged at 14 days post-operatively (if applicable) | Total and Eyes, Verbal (‘T’ if intubated/tracheostomy), Motor (option for ‘Unknown’) |
Finally, each local study lead will complete a provider profiling questionnaire on resources available for the pre-hospital, emergency department, surgical, intensive care and rehabilitation management of patients with acute brain injuries at their institution. This questionnaire has been reviewed and endorsed by the WFNS Neurotraumatology Committee and British Neurotrauma Group.
Collected data will be stored exclusively on a secure bespoke web-based system within the Outcome Registry Intervention and Operation Network (ORION) (https://orioncloud.org/). The platform enables secure data collection, validation and storage in a standard (SQL) format and is compliant with NHS security standards (including the NHS Information Governance toolkit). All patient data will be transmitted and held anonymously. An internal pilot has been conducted using the platform to assess feasibility prior to roll out of the full study. Prior to inputting data into the live platform, members of local study teams will be invited to practise entering ‘dummy’ data into an identical test platform with the aid of a comprehensive software guide.
The protocol (including case report form and site questionnaire) has been translated, with the aid of global collaborators, into Arabic, Bengali, Chinese (traditional and simplified), French, Italian, Portuguese, Brazilian Portuguese, Spanish, Swahili and Urdu and to facilitate recruitment and participation from sites across the world. A comprehensive data dictionary is also available online.
Data validation will occur by two mechanisms. Firstly, web-based forms in the online data capture platform will contain fixed options at the point of data entry to maximise the likelihood of accurate and complete data capture from the outset. Secondly, a local data validator independent of the local study team will be appointed at every participating site. After the study period has ended, each data validator will be asked to retrospectively identify all patients who received emergency surgery for TBI in their centre’s study period (independent of the rest of the study team) to assess case ascertainment and collect data on 2 variables, namely the date and nature of the operation.
As there is a paucity of data on emergency surgery for TBI globally, it was not possible to perform an accurate sample size calculation. However, we expect between 100 and 200 centres to participate and expect the majority of these centres to collect data on 5–10 patients in a given 30-day study period. This will result in a total of 500–2000 patients which we believe will be sufficient to analyse the study endpoints.
For the patient level data, an a priori statistical analysis plan has been developed with a statistician. The mean and standard deviation will be reported for normally distributed variables and the median and interquartile range for not normally distributed variables. Participating centres will be stratified based on their country into groups based on their Human Development Index (HDI) rank. The Human Development Index is calculated for each country based on its life expectancy at birth, years of schooling and gross national income (GNI) per capita (http://hdr.undp.org/en/composite/HDI). 14 day mortality (or survival to discharge, whichever comes first) will be reported for each HDI group and compared using the Pearson chi squared test. Univariable logistic regression analyses will be conducted between covariates and the primary outcome measure. Based on the results, covariates with a p value <0.20 will be included in the multivariable model. The final multivariable logistic model will be determined using stepwise backward elimination. Pre-specified sub-group analyses will also be conducted for patients with severe TBI (GCS 3 to 8), and moderate and severe TBI (GCS 3 to 12). Individual surgeons and hospitals will not be identifiable from analyses (unless explicit consent to do so is given). Statistical analyses will be conducted using R (www.r-project.org). A p value of less than 0.05 will be considered statistically significant. The study and statistical analysis plan was pre-registered on Open Science Framework prior to commencement of data analysis: osf.io/ryjbs.
According to the United Kingdom National Health Service (NHS) Health Research Authority tool, this study is considered a clinical audit rather than research and, as such, does not require approval by a Research Ethics Committee. This has been confirmed formally by the South East Scotland NHS Research Ethics Committee in Edinburgh, Scotland as well as the NHS Research and Development Office at Cambridge University Hospitals. Local teams are expected to seek approval from the appropriate department (research or audit) at their institution prior to commencing the study. Where these departments do not exist, teams should follow the local procedure for obtaining approval for a study of this nature. Written confirmation of local approval and a signed data access and use agreement must be submitted electronically prior to local study teams being granted access to the online data capture platform and commencing data collection. The study has been registered on ClinicalTrials.gov (NCT04212754) and the Clinical Trials Registry – India (CTRI/2019/02/017479).
The study will be reported following the Strengthening the Reporting of Observational Studies in Epidemiology statement guidelines and checklists [14]. The results will be submitted for open access publication in a peer reviewed journal. Fully named authors on the byline of publications resulting from this study will be those that satisfy the International Committee of Medical Journal Editors (ICMJE) guidelines for authorship. In addition, ‘on behalf of the Global Neurotrauma Outcomes Study Collaborators’ will be listed on the byline. The local study lead, additional members of the local study team and independent data validator at each participating site will be listed as PubMed citable collaborator status authors on all publications resulting from this study.
After publication of the primary results, the pooled dataset will be available to all members of the GNOS collaboration for secondary analysis, after judgement and approval of each proposed analysis by the central study team.
We present a study protocol for a multi-centre, international prospective cohort study to ascertain outcomes following emergency surgery for traumatic brain injury globally. We believe the planned 30-day data collection period for each site coupled with 14-day follow up for all patients will be feasible for clinicians even in low-resource settings where their time is at a premium. The collaborative methodology we propose to use in this study has already been employed successfully to obtain outcome data in national neurosurgical studies [15] as well as in international projects in surgery and intensive care [16], [17], [18]. Moreover, we have prospectively planned a data validation strategy to assess case ascertainment.
There is a scarcity of data on outcomes following neurosurgery globally in the literature. The African Surgical Outcomes Study (ASOS), a 7-day prospective observational cohort study of outcomes following all surgery (including neurosurgery, which made up 2.2% of the 11,393 procedures included) in a convenience sample of 25 African countries, found that mortality and postoperative complications were significantly higher in their population relative to high income countries [19]. Similarly, the GlobalSurg study found that mortality following abdominal surgery is significantly higher in countries with a low HDI [17]. Finally, the Corticosteroid Randomisation After Head Injury (CRASH) trial found that mortality following severe, but not mild or moderate TBI, was higher in LMICs than high-income countries [20]. However, no detailed data on the differences in the participating sites, epidemiological characteristics of the patients or processes of care between the 2 groups was presented in this post-hoc subgroup analysis.
The proposed study has several important limitations. Firstly, there is a considerable risk of sampling bias due to our convenience sampling methodology which may limit our ability to extrapolate our findings. We plan to partially address this limitation by having each site fill out a detailed provider profile questionnaire (including questions such as whether each participating hospital is urban or rural, private or government funded and average annual total neurosurgical case volume) and reporting the findings to contextualise the patient level data. Moreover, we will not be collecting a true measure of functional outcome in this study. Studies in traumatic brain injury patients often utilise a rudimentary assessment of functional outcome, such as the Extended Glasgow Outcome Score (GOS-E). However, we believe this would be too onerous for all low resource centres to reliably collect in this study – in addition to the extra work involved in interviewing patients, it would likely be necessary for many centres to translate and possibly culturally validate the GOS-E questionnaire (or similar outcome assessment tool) prior to study commencement. As such, we have opted to include basic surrogate measures of functional outcome such as Glasgow Coma Scale at discharge and discharge destination.
The Global Neurotrauma Outcomes Study seeks to describe the differences in the epidemiological characteristics, management and outcome of emergency surgery for TBI across different Human Development Index settings. The results of GNOS could help identify practices that may best explain differences in outcomes, and will help inform the design of future local, national and international quality improvement programmes and clinical trials for this patient population globally. Moreover, we intend for this study and the international collaboration of clinicians involved in it to inform the development and facilitate the rollout of an international, prospective, hospital-based traumatic brain injury registry, supported by the World Federation of Neurosurgical Societies.
Confirmation of audit/service evaluation status by the South East Scotland Research Ethics Service in Edinburgh.
DC conceived the idea for the study, wrote the study protocol, designed the case report form and site questionnaire and revised the study design/data capture tools following feedback. PH, DC, AJ and AK are the Principal Investigators of the study. DC, AJ, AHB, TB, HB, KB, TF, DKG, MJ, TK, TL, VM, AR, HS, MT, MyT, MaT, RT, IDB, FS, DM, AK, PH contributed to the funding application. AJ, OIA, AOA, AHB, TB, HB, KB, RFM, TF, DKG, MJ, TK, TL, VM, KP, AR, HS, KS, MT, AT, MyT, MaT, RT, SV, IDB, FS, DM, AK, PH, NB, IC, CI, GR, BS contributed input and revisions to the study design and study protocol. SV, DC and RFM created the statistical analysis plan. AB, AI, YS and KB were involved with the revision of the case report form and data capture platform, including assessing their feasibility. DC, PH, AJ and AK made critical revisions to the manuscript. All authors contributed to the manuscript and agreed to submission.
Senior Advisory Panel
Amro Al-Habib, Ignatius Esene, Walt Johnson, Basant Misra, Mahmood Qureshi, Miguel Arraez, Shelly Timmons, Adrian Gelb, Nasser El-Ghandour, Wellingson Paiva.
Dissemination, Communications and Social Media Group
Alex Alamri, Bipin Charausia, Daniel Fountain, Dylan Griswold, Samer Hoz, Aliyu Ndajiwo, Frederico Nicolosi, Irene Panero Pérez, Faith Robertson, Sara Venturini, Laura Lippa, Soubhi Jabal, Gaousul Azam, Chaymae Benyaiche, Hira Burhan, Soham Bandyopadhyay, Leandro Oliveira, Nida Fatima, Davi J. Fontoura Solla, Boniface Kivevele, Sujit Gnanakumar
This research was commissioned by the National Institute for Health Research (NIHR) Global Health Research Group on Neurotrauma (16/137/105) using UK aid from the UK Government. Peter Hutchinson is supported by a Research Professorship from the National Institute for Health Research (NIHR), the NIHR Cambridge Biomedical Research Centre, a European Union Seventh Framework Program grant (CENTER-TBI; grant no. 602150) and the Royal College of Surgeons of England. The funder has not been involved in the drafting of this protocol or preparation of the manuscript.
The study has been registered on ClinicalTrials.gov (NCT04212754) and the Clinical Trials Registry – India (CTRI/2019/02/017479).
David Clark and Peter Hutchinson are the guarantors for the study.
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
GBD collaborators (2019). Global, regional, and national burden of traumatic brain injury and spinal cord injury, 1990-2016: a systematic analysis for the Global Burden of Disease Study 2016. Lancet Neurol. 18(1): 56–87. [published Online First: 2018/11/26]. [PubMed]
Esposito, T.J., Reed, R.L. and Gamelli, R.L. (2005). Neurosurgical coverage: essential, desired, or irrelevant for good patient care and trauma center status. Ann. Surg. 242(3): 364–370. discussion 70–4. [PubMed]
Seelig, J.M., Becker, D.P. and Miller, J.D. (1981). Traumatic acute subdural hematoma: major mortality reduction in comatose patients treated within four hours. N. Engl. J. Med. 304(25): 1511–1518. [PubMed]
Haselsberger, K., Pucher, R. and Auer, L.M. (1988). Prognosis after acute subdural or epidural haemorrhage. Acta Neurochir. (Wien) 90(3–4): 111–116. [PubMed]
Cohen, J.E., Montero, A. and Israel, Z.H. (1996). Prognosis and clinical relevance of anisocoria-craniotomy latency for epidural hematoma in comatose patients. J. Trauma 41(1): 120–122. [PubMed]
Sakas, D.E., Bullock, M.R. and Teasdale, G.M. (1995). One-year outcome following craniotomy for traumatic hematoma in patients with fixed dilated pupils. J. Neurosurg. 82(6): 961–965. [PubMed]
Botman, M., Meester, R.J. and Voorhoeve, R. (2015). The Amsterdam declaration on essential surgical care. World J. Surg. 39(6): 1335–1340. [PubMed]
Mock, C.N., Donkor, P. and Gawande, A. (2015). Essential surgery: key messages from Disease Control Priorities, 3rd edition. Lancet 385(9983): 2209–2219. [published Online First: 2015/02/05]. [PubMed]
WHO. Cause-specific mortality: estimates for 2015 http://www.who.int/healthinfo/global_burden_disease/estimates/en/index1.html: World Health Organisation; 2015 [accessed 24th of July 2017.
Punchak, M., Mukhopadhyay, S. and Sachdev, S. (2018). Neurosurgical care: availability and access in low-income and middle-income countries. World Neurosurg. 112: e240–e254. [published Online First: 2018/01/08]. [PubMed]
Hauptman, J.S., Chow, D.S. and Martin, N.A. (2011). Research productivity in neurosurgery: trends in globalization, scientific focus, and funding. J. Neurosurg. 115(6): 1262–1272. [published Online First: 2011/08/26]. [PubMed]
Patel, A., Vieira, M.M. and Abraham, J. (2016). Quality of the development of traumatic brain injury clinical practice guidelines: a systematic review. PLoS One 11(9)e0161554 [published Online First: 2016/09/01].
Foundation BT, Surgeons AAoN and Surgeons CoN (2007). Guidelines for the management of severe traumatic brain injury. J. Neurotrauma 24(Suppl 1): S1–106.
von Elm, E., Altman, D.G. and Egger, M. (2014). The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement: guidelines for reporting observational studies. Int. J. Surg. 12(12): 1495–1499. [published Online First: 2014/07/18]. [PubMed]
Brennan, P.M., Kolias, A.G. and Joannides, A.J. (2017). The management and outcome for patients with chronic subdural hematoma: a prospective, multicenter, observational cohort study in the United Kingdom. J. Neurosurg., : 1–8. [published Online First: 2017/03/17].
Collaborative, G. (2018). Surgical site infection after gastrointestinal surgery in high-income, middle-income, and low-income countries: a prospective, international, multicentre cohort study. Lancet Infect. Dis., [published Online First: 2018/02/13].
Collaborative, G. (2016). Mortality of emergency abdominal surgery in high-, middle- and low-income countries. Br. J. Surg. 103(8): 971–988. [PubMed]
Vincent, J.L., Marshall, J.C. and Namendys-Silva, S.A. (2014). Assessment of the worldwide burden of critical illness: the intensive care over nations (ICON) audit. LancetRespir. Med. 2(5): 380–386. [published Online First: 2014/04/14].
Biccard, B.M., Madiba, T.E. and Kluyts, H.L. (2018). Perioperative patient outcomes in the African Surgical Outcomes Study: a 7-day prospective observational cohort study. Lancet 391(10130): 1589–1598. [published Online First: 2018/01/03]. [PubMed]
De Silva, M.J., Roberts, I. and Perel, P. (2009). Patient outcome after traumatic brain injury in high-, middle- and low-income countries: analysis of data on 8927 patients in 46 countries. Int. J. Epidemiol. 38(2): 452–458. [published Online First: 2008/09/09]. [PubMed]
This project has received support from a NIHR Global Health Research Group grant. The funder has not been involved in the drafting of this protocol or review of this manuscript.