Bij de eerste opvang van patiënten met licht THL is een CT-scan hersenen de eerste keuze van beeldvormende diagnostiek. Het doel is om traumatische intracraniële afwijkingen te detecteren om te beoordelen of er een indicatie bestaat voor neurochirurgisch ingrijpen en/of het staken van stollingsbeïnvloedende medicatie.

In een onderzoek op SEH-afdelingen in Nederland tussen 2015-2017 werd bij 8,4% van de patiënten afwijkingen op de CT-scan gezien, bij 1,6% een afwijking met potentieel neurochirurgisch ingrijpen (zoals bijvoorbeeld een epiduraal hematoom of een groot subduraal hematoom) (van den Brand, 2022).

Er zijn diverse klinische beslisregels beschikbaar om te besluiten of een CT-scan hersenen is geïndiceerd. Deze beslisregels zijn gebaseerd op uiteenlopende risicofactoren die zijn geassocieerd met het optreden van traumatische intracraniële afwijkingen die leiden tot medisch beleid (opname, observatie, medicatie, poliklinische follow-up) of een indicatie vormen tot neurochirurgisch ingrijpen. Het is belangrijk om te beseffen dat deze beslisregels gebruikt worden om te bepalen of een CT-scan hersenen bij een patiënt op de SEH geïndiceerd is en dat dit niet hetzelfde is als verwijscriteria van de huisarts (zie module Verwijscriteria). Na introductie van de voorgaande richtlijn licht THL (2010), gebaseerd op de CHIP-beslisregels, is het gebruik van CT-scans hersenen fors toegenomen. Een reden hiervoor zou enerzijds de hoeveelheid geformuleerde risicofactoren en anderzijds de striktere adherentie aan de richtlijn na het invoeren daarvan kunnen zijn. Onder andere de leeftijdsgrens van 40 jaar als minor criterium werd als een knelpunt gezien.

Voorts werden na de introductie van de vorige richtlijn licht THL aanvullende factoren vastgesteld die een rol spelen in de besluitvorming om wel/geen CT-scan hersenen uit te voeren, zoals triviaal hoofdletsel (2019) en stollingsbeïnvloedende medicatie (2017) die nu zullen worden opgenomen in de huidige module.

Conclusions

1. Neurosurgery

Canadian CT Head Rule (CCHR); 2. New Orleans Criteria (NOC)

NICE Head Injury Guideline Recommendations; 4. Scandinavian guideline; 5. The National Emergency X-Radiography Utilisation Study II (NEXUS II); 6. CT in head injury patients (CHIP)

2. Intracranial injury

Canadian CT Head Rule (CCHR)

New Orleans Criteria (NOC)

NICE Head Injury Guideline Recommendations; 4. Scandinavian guideline; 5. The National Emergency X-Radiography Utilisation Study II (NEXUS II); 6. CT in head injury patients (CHIP)

Conclusions

Conclusions Antiplatelet medication and anticoagulation as a criterium for CT scanning

Outcome Need for neurosurgical intervention (critical)

DOACS

Antiplatelet therapy

Outcome Intracranial injury (critical)

Antiplatelet therapy

Description of studies

1. Clinical decision rules and additional criteria for CT scanning

1. Diagnostic performance studies
   Harnan (2011) conducted a systematic review on clinical decision rules for adults with mTBI for determining intracranial injuries or necessity of neurosurgical intervention. In total 22 articles were included in the systematic review, representing 19 studies, reporting 25 decision rules. To answer the clinical question for this module, only the following validated clinical prediction rules for CT assessment were extracted: The Canadian CT Head Rule (CCHR), New Orleans criteria (NOC), National Institute for Health and Care Excellence (NICE), Scandinavian guideline, National Emergency X-Radiography Utilisation Study II (NEXUS-II) and CT in Head Injury Patients (CHIP).
   
   Pandor (2011) conducted a systematic review on the diagnostic accuracy of clinical decision rules and individual characteristics for predicting intracranial injury, including need for neurosurgery in adults and children with mTBI. In total, 93 papers were included in the systematic review. In the analyses, the diagnostic accuracy of individual clinical characteristics, biomarkers and clinical decision rules were assessed. To answer the clinical question for this module, only the validated clinical prediction rules for CT assessment were extracted: CCHR, NOC, NICE, Scandinavian guideline, NEXUS-II and CHIP.
   
   Foks (2018) performed a prospective multicenter cohort study on the diagnostic performance of the NOC, CHIP, CCHR and NICE guidelines in the Netherlands (n=4557).
   
   Mower (2017) performed a prospective observational study on the validation of the NEXUS Head CT instrument and compared this with the Canadian Head CT decision rule (n=11770).
   
   Papa (2012) performed a prospective observational cohort study on the diagnostic performance of the CCHR and NOC in the United States (n=656).
   
   Van den Brand (2022) performed a secondary analysis of the data from Foks (2018) on the diagnostic performance of the CHIP (n=4557).
   
   
2. Diagnostic impact studies
   Stiell (2010) performed a matched-pair cluster-design trial on the impact of the Canadian Ct Head Rule. Patients were included when corresponding to the following criteria: 1) blunt trauma to the head resulting in witnessed loss of consciousness; 2) amnesia or witnessed disorientation; 3) an initial Glasgow Coma Scale score of 13 or greater; and 4) occurrence of the injury within the previous 24 hours. Patients were excluded if 1) younger than 16 years of age; 2) penetrating skull injury; 3) focal neurologic deficit; 4) suffered a seizure before arrival at the emergency department; 5) bleeding disorder or used warfarin; or 6) returned for reassessment of the same head injury. The Canadian CT Head Rule was implemented at six intervention sites (n=1049), while in the six control sites (n=876) no specific intervention were introduced to alter the CT-scan-ordering behavior. The following outcomes were assessed: referral for CT scan of the head, cases of brain injury, number of deaths from head injury and return visits to emergency department within 30 days.
   
   
3. Association studies
   In total, 19 observational studies assessed the association between individual clinical (baseline) variables and presence of intracranial injury or the need for neurosurgical intervention to determine whether CT scanning is needed.

2. Antiplatelet medication and anticoagulation as a criterium for CT scanning
   Cheng (2022) performed a systematic review and meta-analysis about the impact of preinjury use of antiplatelet drugs on outcomes of traumatic brain injury. A systematic literatures search was performed in Pubmed, Embase and Google Scholar from inception up to May 15^th, 2021. Cohort studies, in patients sustaining TBI; comparing preinjury use of an antiplatelet drug to no preinjury use of antiplatelet or anticoagulant drugs; on mortality, functional outcome, need for surgical intervention or length of stay as outcomes, were included (n=20). The studies assessing need for surgical intervention as an outcome were extracted for this module (Joseph, 2014; Cull, 2015; Grandhi, 2015; Sumiyoshi, 2017; Tollefsen, 2018; Suehiro, 2019; Robinson, 2021; Wettervik, 2021). It is important to note that these studies did not (only) include mild TBI patients, but (also) moderate and severe TBI patients.
   
   Fiorelli (2020) performed a systematic review and meta-analysis about the risk of intracranial hemorrhage after mild traumatic brain injury in patients on antiplatelet therapy. A systematic literature search was performed on MEDLINE and EMBASE from inception up to April 2020. Prospective and retrospective studies 1) comparing mild TBI patients on antiplatelet medication to those without antiplatelet medication nor antithrombotic therapy; and 2) assessing intracranial hemorrhage as outcome, were included. (n=9). All studies were extracted for this module (Probst, 2020; O’Brien, 2020; Gonzalez, 2020; Galliazoo, Uccella, 2018; Nishijma, 2018; Hamden, 2014; Riccardi, 2013; Spektor, 2003).
   
   Santing (2022) performed a systematic review and meta-analysis about the relationship between pre-injury direct oral anticoagulants (DOAC) use and traumatic hemorrhagic complications in elderly mTBI patients. A systematic literature search was performed in PubMed and Embase for retrospective and prospective observational cohort studies and case-control studies. Studies were included if 1) the relationship between (any type of) DOAC and traumatic intracranial hemorrhage after mTBI was assessed; 2) patients aged ≥60 were included; 3) mTBI was defined as a GCS score of 13-15. Studies were excluded when 1) there was no control group; 2) studies involved also non-TBI patients; only composite outcomes were reported. A total of 16 studies concerning 3671 elderly mTBI patients were included in the systematic review. For this module, only the studies comparing DOAC use, and no antithrombotic therapy use were extracted (Beynon, 2014; Galliazzo, 2019; Parra, 2013; Prexi, 2018). Effects were evaluated by assessing the outcome neurosurgical intervention.

Results

1. Clinical decision rules and additional criteria for CT scanning

The inclusion criteria for the selected clinical prediction rules were shown in table 1.

Table 1. Inclusion- and exclusion criteria per clinical decision rule

Abbreviations: GCS: Glasgow Coma Scale; LOC: loss of consciousness; CT: computed tomography.

Figure 3. Sensitivity and specificity of the original CCHR for intracranial injury.

Adapted

Figure 4. Sensitivity and specificity of the adapted CCHR for intracranial injury.

2. New Orleans criteria (NOC)

3. Neurosurgery

For neurosurgery, sensitivity of the NOC ranged between 0.99 and 1.00 (95% CI range was 0.03 – 1.00). Specificity ranged between 0.04 and 0.31 (95% CI range was 0.02 and 0.32). Results are shown in Figure 5.

Figure 5. Sensitivity and specificity of the NOC for neurosurgery.

4. Intracranial injury

For intracranial injury, sensitivity of the NOC ranged between 0.95 and 1.00 (95% CI range was 0.69 -1.00). Specificity ranged between 0.03 and 0.33 (95% CI range was 0.02 – 0.34). Results are shown in Figure 6.

Figure 6. Sensitivity and specificity of the NOC for intracranial injury.

3. NICE Head Injury Guideline Recommendations

5. Neurosurgery

For neurosurgery, sensitivity for the NICE ranged between 0.89 and 0.94 (95% CI range was 0.65 – 0.93). Specificity ranged between 0.29 and 0.67 (95% CI range was 0.28 – 0.68). Results are shown in Figure 7.

Lenient criteria

Figure 7. Sensitivity and specificity of the NICE for neurosurgery.

6. Intracranial injury

For intracranial injury, sensitivity for the NICE ranged between 0.72 and 0.99 (95% CI range was 0.68 – 1.00). Specificity ranged between 0.31 and 0.70 (95% CI range was 0.30 – 0.71). Results are shown in Figure 8.

Lenient criteria

Figure 8. Sensitivity and specificity of the NICE for intracranial injury.

4. Scandinavian guideline

7. Neurosurgery

For neurosurgery, sensitivity of the Scandinavian guideline ranged between 0.94 and 0.99 (95% CI range was 0.71 – 1.00). Specificity ranged between 0.20 and 0.50 (95% CI range was 0.19 – 0.51). Results are shown in Figure 9.

Lenient criteria

Figure 9. Sensitivity and specificity of the Scandinavian lenient and strict criteria for neurosurgery.

8. Intracranial injury

For intracranial injury, sensitivity of the Scandinavian guideline ranged between 0.93 and 0.98 (95% CI range was 0.90 – 0.97). Specificity ranged between 0.21 and 0.60 (95% CI range was 0.20 – 0.63). Results are shown in Figure 10.

Lenient criteria

Figure 10. Sensitivity and specificity of the Scandinavian lenient and strict criteria for intracranial injury.

5. The National Emergency X-Radiography Utilisation Study II (NEXUS II)

9. Neurosurgery

For neurosurgery, sensitivity of the NEXUS-II was 1.00 (95% CI 0.97 – 1.00). Specificity was 0.44 (95% CI 0.43 – 0.45). Results are show in Figure 11.

Figure 11. Sensitivity and specificity of NEXUS II rule for neurosurgery.

10. Intracranial injury

For intracranial injury, sensitivity of the NEXUS-II ranged between 0.98 and 0.98 (95% CI range was 0.95 – 0.99). Specificity ranged between 0.14 and 0.47 (95% CI range was 0.13 and 0.48). Results are shown in Figure 12.

Figure 12. Sensitivity and specificity of NEXUS II rule for intracranial injury.

6. CT in head injury patients (CHIP)

11. Neurosurgery

For neurosurgery, sensitivity of the CHIP simple decision rule ranged between 0.94 and 1.00 (95% CI range was 0.73 – 1.00). Specificity ranged between 0.20 and 0.23 (95% CI 0.19 – 0.25). Results are shown in Figure 13.

Simple decision rule

Figure 13. Sensitivity and specificity of CHIP simple decision rule for neurosurgery.

For neurosurgery, sensitivity of the CHIP detailed decision rule ranged between 0.93 and 1.00 (95% CI was 0.80 – 1.00). Specificity ranged between 0.27and 0.30 (95% range was 0.26 and 0.32). Results are shown in Figure 14.

Detailed decision rule

Figure 14. Sensitivity and specificity of CHIP detailed decision rule for neurosurgery.

12. Intracranial injury

For intracranial injury, sensitivity of the CHIP simple decision rule ranged between 0.94 and 0.96 (95% CI 0.91 – 0.98). Specificity ranged between 0.22 and 0.25 (95% CI range was 0.20 and 0.27). Results are shown in Figure 15.

Simple decision rule

Figure 15. Sensitivity and specificity of CHIP simple decision rule for intracranial injury.

For intracranial injury, sensitivity of the CHIP detailed decision rule ranged between 0.93 and 0.94 (95% CI 0.91 – 0.97). Specificity ranged between 0.27 and 0.32 (95% CI range was 0.26 and 0.34). Results are shown in Figure 16.

Detailed decision rule

Figure 16. Sensitivity and specificity of CHIP detailed decision rule for intracranial injury.

Level of evidence of the literature

1. Neurosurgery

The level of evidence in the literature started at high because it was based on diagnostic test accuracy studies. For the CCHR and the NOC, the level of evidence was downgraded by two levels due to verification bias and methodological shortcomings (-2, risk of bias). The final level is low. For the NICE, the Scandinavian guideline, NEXUS-II and the CHIP, the level of evidence was downgraded by three levels due to a short follow-up, verification bias and methodological shortcomings (<30 days in most studies) (-2, risk of bias) and low number of included studies (-1, imprecision). The final level is very low.

2. Intracranial injury

The level of evidence in the literature started at high because it was based on diagnostic test accuracy studies. For the CCHR and the NOC, the level of evidence was downgraded due to methodological shortcoming (-1, risk of bias). The final level is moderate. For the NICE, the Scandinavian guideline, NEXUS-II and the CHIP, the level of evidence was downgraded by two levels due to methodological shortcomings (-1, risk of bias) and low number of included studies (-1, imprecision). The final level is low.

Conclusions

1. Neurosurgery

Canadian CT Head Rule (CCHR); 2. New Orleans Criteria (NOC)

NICE Head Injury Guideline Recommendations; 4. Scandinavian guideline; 5. The National Emergency X-Radiography Utilisation Study II (NEXUS II); 6. CT in head injury patients (CHIP)

2. Intracranial injury

Canadian CT Head Rule (CCHR)

New Orleans Criteria (NOC)

NICE Head Injury Guideline Recommendations; 4. Scandinavian guideline; 5. The National Emergency X-Radiography Utilisation Study II (NEXUS II); 6. CT in head injury patients (CHIP)

b. Diagnostic impact studies

The trial from Stiell (2010) assessed the impact of applying the Canadian CT Head Rule compared to not applying this rule in mTBI patients by assessing the following clinical outcomes:

1. Clinically important brain injury (epidural hematoma, subdural hematoma, intracerebral hematoma, cerebellar hematoma, diffuse cerebral edema, cerebral contusion, subarachnoid hemorrhage, intraventricular hemorrhage, pneumocephalus, depressed skull fracture), requiring hospital admission and neurosurgical follow-up;
2. Clinically unimportant brain injury;
3. Neurosurgical intervention (craniotomy, elevation of skull fracture, isolated intubation for head injury, isolated intracranial pressure monitoring);
4. Death from brain injury;
5. Admitted to hospital.

Results are shown in Table 3.

Table 3. Impact of the Canadian CT Head rule on patients’ outcomes (extracted from Stiell, 2010)

The impact of the New Orleans Criteria, the National Institute for Health and Care Excellence, the Scandinavian guideline, the National Emergency X-Radiography Utilisation Study II and the CT in Head Injury Patients was not assessed.

The level of evidence in the literature

The level of evidence regarding all outcome measures started at high because they were based on a randomized controlled trial. The level of evidence was downgraded by three levels due to lack of randomization, low physician compliance with the requisition form and unbalanced baseline characteristics (-2, risk of bias) and low number of events (-1, imprecision). The final level is very low.

Conclusions

c. Association studies

From the 19 observational studies, the following variables were significantly associated with intracranial injury or neurosurgical intervention in addition to the criteria that were included in the clinical decision rules. Table 4 shows an overview of the included studies assessing clinical variables that are significantly associated with intracranial injury:

1. Alcohol/drugs abuse;
2. Head laceration or bruise;
3. New abnormalities found on neurologic examination;
4. Chronic kidney disease;
5. Duration since injury is 6 hours or less;
6. Anisocoria.

There were no clinical variables that were significantly associated with need for neurosurgical intervention (apart from the variables included in clinical decision rules).

Table 4. Additional criteria to clinical decision rules for CT scanning

Abbreviations: LOC, loss of consciousness; INR, internal normalized ratio; PTT, partial thromboplastin time; ISS, injury severity score; AIS, abbreviated injury score; MVC, motor vehicle crash; ICH, intracranial haemorrhage; TAO, thromboangiitis obliterans

Level of evidence of the literature

The level of evidence was not assessed because it is a descriptive overview of the literature. No GRADE assessment could be performed since no studies were included with at least internal validation of the prediction models.

Antiplatelet medication and anticoagulation as a criterium for CT scanning

Outcome Need for neurosurgical intervention (critical)

The systematic review of Cheng (2022) compared the need for neurosurgical intervention between patients with mild, moderate and severe TBI on antiplatelet medication and patients with TBI without antiplatelet medication in eight studies (n=2437). Data resulted in a RR of 1.09 (95% CI 0.87 to 1.37) favoring the intervention group receiving antiplatelet therapy (figure 17). This effect was considered clinically relevant (although non-significant).

Figure 17. Effects of antiplatelet therapy (AP) on the need to neurosurgical intervention compared to controls.

The systematic review of Santing (2022) assessed the need for neurosurgical intervention in elderly mTBI patients using direct oral anticoagulants (DOACS) versus elderly mTBI patients using no antithrombotic therapy in four studies (n=224). Data resulted in a RR of 0.84 (95% CI 0.25 to 2.88) favoring the intervention group receiving DOACS (figure 18). This effect was considered clinically relevant (although non-significant).

Figure 18. Effects of DOACS on the need to neurosurgical intervention compared to controls.

Outcome Intracranial injury (critical)

The systematic review of Fiorelli (2020) compared incidence of intracranial injury between patients with TBI on antiplatelet medication and patients with TBI without antiplatelet medication in nine studies (n=13791). Data resulted in a RR of 1.51 (95% CI 1.21 to 1.88) favoring the control group (figure 19). This significant difference was considered clinically relevant.

In addition, Fiorelli (2020) performed a subgroup analysis for the patients taking aspirin alone and for patients on dual antiplatelet therapy. In the aspirin subgroup, data resulted in a RR 1.27 (95% CI 1.00 – 1.61) favoring the control group. In the dual antiplatelet therapy subgroup, data resulted in an RR of 3.21 (95% 2.15 – 4.76) favoring the control group. These (borderline) significant differences were considered clinically relevant. Since the review only showed the pooled effects of these subgroup analyses, no forest plot was created.

Figure 19. Effects of antiplatelet therapy on the incidence of intracranial injury compared to controls.

The level of evidence in the literature

For both outcome measures, the level of evidence started at high because it was based on two systematic reviews with prognostic observational studies.

Outcome Need for neurosurgical intervention (critical)

Antiplatelet therapy

The level of evidence was downgraded by four levels due to retrospective design of all included studies, causing selection bias (-1, risk of bias); confidence intervals crossing both borders of clinical relevance (-2, imprecision); and indirect evidence for the effects of antiplatelet medication in mild TBI specifically (-1, indirectness). The final level is very low.

DOACS

The level of evidence was downgraded by four levels due to retrospective design of all included studies, causing selection bias (-1, risk of bias); confidence intervals crossing both borders of clinical relevance (-2, imprecision); and indirect evidence for the effects of antiplatelet medication in mild TBI specifically (-1, indirectness). The final level is very low.

Outcome Intracranial injury (critical)

Antiplatelet therapy

The level of evidence was downgraded by two levels due to the retrospective design of half of the included studies, a lack of analyzing potential confounders and the lack of evaluating the risk of intracranial injury in patients who did not receive a CT scan (-2, risk of bias). The final level is low.

A systematic review of the literature was performed to answer the following questions:

1. What is the diagnostic accuracy of clinical decision rules for determining whether a CT scan is needed in patients with mTBI within 24h after the accident for identifying intracranial injury and/or need for neurosurgery?        

Timing and setting: emergency physician, radiologist, neurologist within 24 hours after mTBI.

2. Is the use of new anticoagulants (DOACS), low molecular wight heparins in therapeutic doses, clopidogrel and combinations of platelet aggregation inhibitors associated with an increased risk for intracranial complications or need for neurosurgical interventions after mTBI within 24h after the accident?

Timing and setting: emergency physician, radiologist, neurologist within 24 hours after mTBI.

Relevant outcome measures

1. The guideline development group considered sensitivity as a critical outcome measure for decision making and specificity as an important outcome measure for decision making. Sensitivity was considered clinically accurate when being ≥95% for the outcome neurosurgical intervention. No minimally clinical accuracy of specificity was determined. For intracranial injury, no minimally important differences for sensitivity and specificity were determined.
2. The guideline development group considered need for neurosurgical intervention as a critical outcome measure for decision making. Intracranial injury was considered important. For both outcomes, a relative difference of 3% was considered clinically relevant (risk ratio (RR) of < 0.97 or > 1.03).

Search and select (Methods)

1. The databases Medline (via OVID) and Embase (via Embase.com) were first searched for systematic reviews with relevant search terms from 2000 until August 9^th, 2022. The detailed search strategy is depicted under the tab Methods. The systematic literature search resulted in 95 systematic reviews, 507 studies were labeled as RCTs and 635 as observational studies.
   
   A total of 60 studies were initially selected based on title and abstract screening. After reading the full text, 38 studies were excluded (see the table with reasons for exclusion under the tab Methods), and 22 studies were included.
   
   Studies were divided into the following categories:
   1. Diagnostic performance studies about the performance of validated clinical decision rules to identify intracranial injury or neurosurgical intervention.
   2. Diagnostic impact studies about the impact of applying validated clinical decision rules on patient relevant outcome measures.
   3. Association studies about the relationship between clinical criteria and intracranial injury or neurosurgical intervention. The level of evidence was not assessed because it is a descriptive overview of the literature. No GRADE assessment could be performed since no studies were included with at least internal validation of the prediction models. Thus, no risk of bias nor evidence tables were included for this part of the literature overview.
   
   
2. The databases Medline (via OVID) and Embase (via Embase.com) were searched for systematic reviews with relevant search terms from 2000 until May 19^th, 2022. The detailed search strategy is depicted under the tab Methods. The systematic literature search resulted in 72 systematic reviews. Studies were selected based on the following criteria:

• Systematic reviews including a systematic evaluation of all included studies (meta-analysis) and risk of bias assessment;
• Including adult (16+ years) patients with mild traumatic brain injury using antiplatelet medication before injury;
• Described at least one of the outcome measures as described in the PICO;
• Included at least 20 patients.
  
  Five studies were initially selected based on title and abstract screening. After reading the full text, two studies were excluded (see the table with reasons for exclusion under the tab Methods), and three systematic reviews were included.

Results

1. Clinical decision rules and additional criteria for CT scanning: 22 studies were included in the analysis of the literature:
   a. Diagnostic performance studies: Two systematic reviews and four observational studies.
   b. Diagnostic impact studies: One non-randomized controlled trial.
   c. Association studies: 19 observational studies.
2. Antiplatelet medication and anticoagulation as a criterium for CT scanning: three systematic reviews were included in the analysis of the literature.