European Society of Emergency Radiology: guideline on radiological polytrauma imaging and service (short version)

Abstract

Background

Although some national recommendations for the role of radiology in a polytrauma service exist, there are no European guidelines to date. Additionally, for many interdisciplinary guidelines, radiology tends to be under-represented. These factors motivated the European Society of Emergency Radiology (ESER) to develop radiologically-centred polytrauma guidelines.

Results

Evidence-based decisions were made on 68 individual aspects of polytrauma imaging at two ESER consensus conferences. For severely injured patients, whole-body CT (WBCT) has been shown to significantly reduce mortality when compared to targeted, selective CT. However, this advantage must be balanced against the radiation risk of performing more WBCTs, especially in less severely injured patients. For this reason, we recommend a second lower dose WBCT protocol as an alternative in certain clinical scenarios. The ESER Guideline on Radiological Polytrauma Imaging and Service is published in two versions: a full version (download from the ESER homepage, https://www.eser-society.org) and a short version also covering all recommendations (this article).

Conclusions

Once a patient has been accurately classified as polytrauma, each institution should be able to choose from at least two WBCT protocols. One protocol should be optimised regarding time and precision, and is already used by most institutions (variant A). The second protocol should be dose reduced and used for clinically stable and oriented patients who nonetheless require a CT because the history suggests possible serious injury (variant B). Reading, interpretation and communication of the report should be structured clinically following the ABCDE format, i.e. diagnose first what kills first.

Key points

• If indicated, whole-body-CT (WBCT) saves lives in severely injured patients.

• However, WBCT radiation dose risk versus benefit depends on severity of injury.

• Two WBCT protocols should be established (A: time/precision optimised, B: dose reduced).

• Protocol A should be used for clinically unstable patients/life-threatening conditions.

• For all other patients, protocol B should be selected.

Background

The European Society of Emergency Radiology (ESER) is an apolitical, non-profit organisation, exclusively and directly dedicated to promoting and coordinating the scientific, philanthropic, intellectual and professional activities of Emergency Radiology. The Society’s mission at all times is to serve the health care needs of the general public through the support of science, teaching, research and the quality of service in the field of Emergency Radiology [1]. One particular aim of ESER is to advance and improve the radiological aspects of emergent patient care and to advance the quality of diagnosis and treatment of acutely ill or injured patients using imaging.

Emergency Radiology encompasses medical and surgical subspecialties including polytrauma services. Concerning the latter, past and present ESER board members had taken part in several interdisciplinary guideline processes. However, the ESER board has observed the lack of dedicated separate independent radiological recommendations for a radiological polytrauma service. The ESER has therefore created such recommendations with the hope that this will start to bring corresponding diverse national and international radiological societies together in order to refine the statements, gain visibility for national societies and in particular, strengthen the role of radiology in upcoming interdisciplinary polytrauma guideline development.

As ESER also wants to be a promotor of future scientific work, we hope to give advice on specific questions as well as for a more general principal direction. To update this guideline, ESER will refine the statements at appropriate time intervals, (currently estimated to be two years).

The Guideline on Radiological Polytrauma Imaging and Service is published simultaneously in a comprehensive short version (this article) and a full version (download of the full version from the ESER homepage [1]). This causes text overlap between the two versions. We mention this to avoid a potential conflict with respect to self-plagiarism.

Methods

The ESER Board instructed the former ESER President (S.W.) to divide the entire field of radiological polytrauma care into individual sections. S.W. assigned parts of the project to J.H. as part of his doctoral thesis at the Ludwig-Maximilian-University, Munich, Germany. Each section was processed and prepared according to a fixed schedule: determination of key issue(s), literature research, selection of literature, classification of literature, rating of literature, determining a level of evidence, suggesting a grade of recommendation, suggesting a statement for each key issue as a basis for the consensus conferences.

Key issues

Each section was related to at least one key issue/question. The consensus conference members had to discuss and vote on one (or more) answer(s) to each key question, but were also allowed to delete or change answers.

Literature research

For each key issue, a literature search was conducted with subjectively fitting keywords from the MeSH terms (Medical Subject Headings, [2]) including subjectively fitting synonymous keywords. The MeSH term search itself was performed using NCBI (National Center for Biotechnology Information) [2]. These keywords were used for searching through several databases: MEDLINE (via PubMed [3]), Cochrane Library [4] and Embase (via Ovid [5]). These databases were accessed via the Database Information System (DBIS) [6] of the University Library of the LMU Munich, where full text access was available for almost all journals. If there was only access to the abstract but not to the full text, the literature was excluded. Depending on the key issue the rate of such exclusion ranged between two and ten percent. Search terms and their connection were adapted to the individual databases. For guidelines, the databases of the NICE (National Institute for Health and Care Excellence [7]) and AWMF (Association of the Scientific Medical Societies in Germany [8]) were scanned. The AWMF database in German was included because S.W. and J.H. were able to understand it and, if necessary, translate it for the consensus conferences. The NICE search for guidelines was performed with the additional filter ‘Secondary Evidence’. The literature found was selected in a fixed order. The first step was to evaluate the relevance by title, then by abstract and, if necessary, by keyword search in the full text. Any literature not excluded in this first step was then subject to a more detailed second step examination of the inclusion and exclusion criteria on the basis of the full text. For more (very detailed) information about the literature search please refer to the guideline in full text [1].

Inclusion/exclusion criteria for literature selection

The literature was selected on the basis of a catalogue of inclusion criteria. The search key words were determined together by S.W. and J.H. according to ‘sections’ and related ‘key issue(s)’. Data were excluded if at least one of the inclusion criteria listed below was not fulfilled.

• 1 Publication period from January, 1 2010 to February, 15 2019

• 2 Study population: n ≥ 50, adults (an age limit was not applied as children develop at different rates and therefore, from a radiological point of view, there may be a smooth transition to the body of an adult.)

• 3 Language of publication: English or German (German because S.W. and J.H. were able to understand it and, if necessary, translate it for the consensus conferences)

• 4 Full text accessible, free of charge via the university portal used

• 5 Clinical relevance of the literature included with regard to the key issue (subjective evaluation)

• 6 Additional criteria for guidelines:

  • It is published as a guideline, i.e. using the word ‘guideline’ in the title

  • The guideline is described as being current or no updated version is available

• 7 Additional criteria for studies:

  • Allowed study types: meta-analyses, systematic reviews, randomised controlled trials, cohort studies, case–control studies, cross-sectional studies, before—after studies

  • Outcome: p-value < 0.05 and/or confidence interval (CI) > 95%

Classification, rating, and evidence level of studies

Classification In case of studies, the algorithm according to Hartling et al. [9] was used to classify the study type for included publications (e.g. prospective cohort study, case–control study, randomised controlled trial).

Rating Systematic reviews and meta-analyses were rated using AMSTAR 2 [10], randomised controlled trials using the Cochrane method [11], and cohort or case control studies using the Newcastle—Ottawa Scale (NOS) [11]. The Cochrane method and NOS have been performed according to the description in the Manual of Cochrane Germany and Association of Scientific Medical Societies in Germany (AWMF) [11] (e.g. were the items of PICO (population, intervention, comparator group and outcome) applied for a randomised controlled study? Were the sources of funding of the study published? [10]).

Evidence Level For every study, the level of evidence of was assigned using the scheme of the Oxford Centre for Evidence—Based Medicine in the 2011 version [12] (e.g. level 1 corresponds to systematic review of randomised trials or level 3 corresponds to a cohort study [12]).

Grade of recommendation (GoR)

Using the evidence levels according to the AWMF Guidance Manual [13] one out of three possible Grades of Recommendation (GoR) was issued on each answer to a key question: A = ‘should/ should not’ with the meaning of ‘certainly should/ should not’; B = ‘ought/ ought not’ with the meaning of ‘probably/preferably should/should not’; 0 = ‘may be considered’ or ‘consider’. The GoR was based on the evidence level of the included studies: evidence level 1 led to a recommendation level ‘A’ (strong recommendation), evidence level 2 led to a recommendation level ‘B’ (recommendation) and evidence levels 3, 4, 5 lead to a recommendation ‘0’ (open recommendation). For statements from guidelines, the specific degree of recommendation was adopted from the guideline. Following the AWMF principle [14] and in case the used scale allowed us to do so, the consensus conference was able to increase or decrease the GoR by one degree of recommendation [14].

Good clinical practice points (GPP)

If there was insufficient evidence in the literature included, the degree of recommendation—GPP (Good Clinical Practice Points, [15]-p. 27) was used. In contrast to GoR, GPP is based purely on the consensus of the experts. The GPP degrees of recommendation are identical to those for GoR: ‘A’ = strong recommendation, ‘B’ = recommendation and ‘0′ = open recommendation. For differentiation purposes, the degree of recommendation was therefore marked with GPP instead of GoR.

Consensus development at the conferences

For each key issue, S.W. and J.H. proposed a statement with a corresponding level of recommendation (GoR or GPP) to the consensus conference. This served as a basis for the discussion during the consensus conference. If present in person, each member of the last and the current ESER Board (from 2017 until now) had exactly one equal vote. As the consensus conferences were during congress meetings, not all consensus members were able to present for the entire consensus conferences. According to our constitution, attendance by 2/3 of members was considered quorate for each vote on each statement. All members reviewed each statement during the publication process. The suggested statements and grades of recommendation as well as the corresponding literature were sent to the participants in advance by email.

The procedure for each key question is as follows (see also Table 1): For each section, the suggested statement(s) and GoR/GPP were presented by S.W. and a discussion was opened with the possibility of further questions, amendments, additions and objections. This stopped when there seemed to be a majority on the wording and the final suggested or amended statement was then voted on (with % consensus recorded). This was followed by a second vote on the GoR/GPP for this statement (again, the % consensus was recorded). If necessary, a new proposal for GoR/GPP was formed by discussion considering the rules for GoR/GPP (AWMF principle as described earlier) until at least simple majority was reached. Each voting was performed anonymously by holding a laser pointer within a given area, which was interpreted as ‘yes’, outside this area as ‘no’, and missing pointer signals as abstention (did not occur). All voting results were recorded (Tables 2, 3, 4, 5).

Table 1 Section 1: Polytrauma classification*

Table 2 Section 2: Structural points, key issue 1: CT location

Table 3 Section 2: Structural points, key issue 2: CT type

Table 4 Section 2: Structural points, key issue 3: Diagnostic Environment and Communication

Table 5 Section 2: Structural points, key issue 4: Quality Management

An agreement of voting was achieved by a consensus strength of more than 50% of the present votes.

The consensus strength was graduated according to the AWMF rules ([14]-p.40) as follows: Strong (strong agreement): > 95% of votes; Normal (normal agreement): > 75–95%; Weak (majority agreement): > 50–75% and None (no agreement): < 50% (Tables 6, 7, 8, 9, 10).

Table 6 Section 3: Extended Focused Assessment with Sonography for Trauma (eFAST)

Table 7 Section 4: Conventional Radiography

Table 8 Section 5: Whole Body CT – Positioning, key issue 1: patient orientation

Table 9 Section 5: Whole Body CT – Positioning, key issue 2: Arm position

Table 10 Section 6: Whole Body CT – Protocol, key issue 1: CT scout

Results

The results of the consensus conferences are presented here and structured into ten sections. Each section may be subdivided into several key issues that were presented as tables in the following. Each table also holds a collection of ‘key literature’ that corresponds to the literature included. The tables also include a path through the literature classification as well as the evidence levels of the included literature (Tables 11, 12, 13, 14, 15).

Table 11 Section 6: Whole Body CT – Protocol, key issue 2: Cranial CT

Table 12 Section 6: Whole Body CT – Protocol, key issue 3: Cervical Neck/Spine

Table 13 Section 6: Whole Body CT – Protocol, key issue 4: contrast phase

Table 14 Section 6: Whole Body CT – Protocol, key issue 5: Injection of Contrast Media

Table 15 Section 7: Whole Body CT – Special protocols, key issue 1: CT—urography

Discussion and conclusions

For a detailed literature discussion of more than 50 print pages we have to refer to the guideline in full length (access via ESER homepage [1]).

As a relatively young society, ESER overcame challenges during the guideline development, consensus and publication process. The members of the consensus group were distributed throughout diverse nations, making the necessary distribution of information and communication time consuming. Financial limitations restrict the whole group from coming together face to face to only once or twice a year, during the European congresses of radiology in Vienna and the ESER congress meetings. This huge project required two sittings of the consensus group to adequately provide time for discussion, this, in addition to restrictions of the SARS-Cov-2 situation and cancellation of the European Congress of Radiology (ECR) 2020 caused unexpected delay in manuscript production (Tables 16, 17, 18, 19).

Table 16 Section 7: Whole Body CT – Special protocols, key issue 2: CT—angiography

Table 17 Section 8: Whole Body CT – Reading/ Reporting

Table 18 Section 9: Interventional Radiology

Table 19 Section 10: Summary: A proposal for two WBCT—Protocols in the Trauma Care

Limitations

• Only one person was involved in suggesting key issues (S.W.)

• Literature search was limited to two persons (S.W., J.H.)

• Literature preparation (exclusion, inclusion, grading) was also limited to S.W., J.H.

• The preparation of the consensus conference(s) including suggested statements and respective grading of them was limited to S.W., J.H.

• Literature inclusion was limited to free full access via the Ludwig-Maximilians University of Munich, Germany. However, this quote was about 95 percent in mean and always above 90 percent.

• German was the only non-English language that could be included in the literature search (because S.W. and J.H. were able to understand and translate it for the consensus conference members).

• The guideline does not cover special topics like paediatric patients or interventional radiology; these are an aspiration for future editions

Conclusions

By developing this guideline, the ESER aimed to redress the lack of dedicated separate independent radiological recommendations for radiological polytrauma service. ESER recommends that a patient should first be assessed as ‘polytrauma’, who will therefore receive whole-body CT (WBCT) or ‘non-polytrauma’ (assess patient as a ‘normal’ emergency patient in the Emergency Trauma Room: do not automatically perform a WBCT). For a polytrauma service, the CT distance to the Emergency Trauma Room should not exceed 50 m—the closer, the better [15, 19, 20]. The CT used should offer 64 rows and modern technology (cardiac capability is welcome but not essential) [15, 24, 26,27,28,29]. Radiology departments as part of Trauma centres should optimise communication and drive quality assurance/ management [15, 31, 32, 34, 150, 151]. eFAST should be part of the primary survey and Radiography should be immediately available [15, 37,38,39,40,41, 43, 46]. ESER prefers to position patients ‘feet first’. In case of stable patients and if possible, arms should be elevated for dose reduction (only if this is done prior to the body scout) [15, 54, 56, 57]. CT scouts preferably should present the whole patient (but may consist of different parts) [37, 63], may replace chest radiographs [15, 51] and sometimes also provide justification for deviating from the standard protocol by choosing different contrast phases or extend scanning to other suspicious body regions. The unenhanced cranial CT scan certainly should be done first [19]. At least when using the ‘dose’ protocol (WBCT variant B), the cranial CT scan should only cover the brain. For unstable patients, the midface/neck/cervical spine should be scanned together with the chest using arterial contrast including the arteries of the skull base [70] (for stable patients a separate low-dose scan with or without contrast enhancement may be an alternative before lifting the arms). A split bolus protocol should probably be used with a dose-optimised protocol [74, 101, 108]. Otherwise, ESER recommends overlapping scans of the neck/chest/upper abdomen in arterial phase and the abdomen/pelvis in portal-venous phase [26, 98, 99]. For specific questions related to the urogenital, interventional, (cardio)vascular or paediatric specialties, ESER recommends using existing guidelines from the respective (sub)societies [117, 118, 129,130,131]. First images should be available, read and communicated as fast as possible using the ABCDE approach [19, 66, 143]. In the second step ‘perfect’ images should be calculated (in both soft and enhancing kernels) and be interpreted (and archived) at least in the three standard planes, respectively [19, 65, 66, 144]. Interpretation should occur three times (immediately using first images, immediately reassessed using the final images and reassessed again by a different radiologist within 24 h) [141, 144, 146].

ESER endorses abandoning a ‘one-size-fits-all-concept’ ([63]-p.1142). Instead, ESER recommends introducing a double-track whole-body tomography protocol concept with a ‘Dose Protocol’ and a ‘Time/Precision Protocol’. Obviously, the choice between the two variants should be based on the individual clinical presentation and vital parameters of the polytrauma patient. The ‘Dose protocol’ should be designed in such a way that the patient is exposed to the lowest possible radiation exposure despite sufficient image quality in order to ensure a reliable diagnosis of injuries (often young and stable patients with dramatic injury history and a Glasgow Coma Scale = 15). A dose far below 20 mSv should be aimed for. A good potential ‘Dose protocol’ may consist of an unenhanced head scan, low dose CT of the midface/ neck/ cervical spine (with or without contrast enhancement), elevation of the arms, scout of the trunk, and a single pass scan of chest/ abdomen and pelvis using a split bolus injection protocol with a resulting arterial/venous mixed contrast of all vessels and organs. In (few) cases where a ‘Dose protocol’ scan leaves potentially important findings unclear, another CT scan should be performed accordingly.

In contrast, the ‘Time/Precision protocol’ is optimised for very fast, very high diagnostic accuracy and will more or less correspond to the institutional protocol used so far. The key advantage is the more sensitive detection of active bleeding [15, 19, 107, 108]. The assignment of the polytrauma patient to one of the two protocols is shown in Fig. 1.

Fig. 1

Decision guidance for polytrauma CT imaging. First, a potential polytrauma patient should be re-evaluated in the Emergency Trauma Room whether the criteria for a classification as polytrauma (Table 1) is given. If so, and in the case of a severe clinical presentation with life-threatening injuries and/or haemodynamic instability, the polytrauma ‘Time/Precision protocol’ (whole-body CT (WBCT) variant A) is applied. If the patient is also classed as polytrauma but does not fulfil criteria for MDCT protocol variant A, the ‘Dose protocol’ (WBCT variant B) may be used. Otherwise, the patient should receive imaging like other emergency patients

The ESER hopes that this guideline motivates diverse national and international radiological societies to come together in order to refine the statements over time. The ESER acknowledges that these guidelines do not focus on the radiological polytrauma service for children and Interventional Radiology. Rather the ESER invites the corresponding national and international radiological (sub)societies to contribute in the future. Where the guidelines do overlap with other radiological communities on topics such as Musculoskeletal, Abdominal & Urogenital imaging, the ESER anticipates arriving at a consensus in the future.

ESER sees this as way to gain visibility for national societies in the field and in particular to strengthen the role of Radiology in upcoming interdisciplinary polytrauma guideline processes. As ESER is active in the whole field of emergency radiology, we also aim to expand the guideline to non-traumatic Emergency Imaging in upcoming versions.