REACT 2

react-2-twitter-image

REACT 2: Immediate total-body CT scanning versus conventional imaging and selective CT scanning in patients with severe trauma: a randomised controlled trial

Sierink. Lancet 2016, published on-line June 28, http://dx.doi.org/10.1016/S0140-6736(16)30932-1

Clinical Question

  • In patients with severe trauma does immediate total body CT scanning compared with standard workup with selective CT scanning, reduce hospital mortality?

Design

  • Randomised controlled trial
  • Computer randomisation
  • Stratified by centre
  • Non-blinded
  • Modified intention to treat analysis
    • Patients excluded post randomisation if:
      • Did not fulfil inclusion/exclusion criteria (n=203)
      • Declined participation (n=112)
      • Language barrier (n=5)
  • Power calculation: 539 patients in each group would give the study an 80% power to detect an absolute risk reduction of 5% with a false positive rate of 5%

Setting

  • 4 hospitals in the Netherlands and 1 hospital in Switzerland
    • All level 1 trauma centres and academic teaching hospitals
  • Data collected: April 2011 – January 2014

Population

  • Inclusion criteria: Adults trauma patients with potentially severe injuries, as defined by:
    • Abnormal physiology
      • At least 1 of the following at hospital arrival: Respiratory rate ≥30 or ≤10, HR ≥120, BP≤100, estimated blood loss ≥500ml, GCS≤13, abnormal pupillary reaction
    • OR clinical suspicion of any of the following injuries:
      • Fractures from at least 2 long bones
      • Flail chest, open chest or multiple rib fractures
      • Severe abdominal injury
      • Pelvic fracture
      • Unstable vertebral fractures/spinal cord compression
    • OR severe mechanism of injury:
      • Fall from height (>3 metres or 10 feet)
      • Ejection from vehicle
      • Death of occupant in same vehicle
      • Severely injured patient in same vehicle
      • Wedged or trapped chest/abdomen
  • Exclusion criteria:
    • Patients referred from another hospital
    • Clearly low-energy blunt trauma
    • Stab wound in 1 body region
    • Too unstable to undergo CT and required either CPR or immediate operation
  • 1403 patients randomised of whom 320 excluded post randomisation  (including 60 patients who did not meet inclusion criteria, 107 patients who had low energy blunt trauma, and 112 patients who declined participation)

Comparing intervention with control group

  • Significantly more patients with ISS≥16: 67% vs. 61%, p=0.045
  • Significantly lower:
    • Hb (median): 129 vs. 133, p=0.003
    • Haematocrit (median): 0.38 vs. 0.39, p=0.003
  • No significant difference in:
    • Age (median): 42 vs. 45
    • ISS (median): 20 vs. 19, p=0.4
    • % that received blood transfusion: 27% vs. 28%
    • Blunt trauma: 98% vs. 98.3%
    • Traumatic brain injury: 32.9% vs. 27.9%

Intervention

  • Immediate total-body CT scanning (vertex to pubic symphysis)
    • Without prior conventional imaging
    • Potential life saving interventions performed prior to imaging including: intubation, chest tube insertion, pericardiocentesis, haemorrhage control
  • Protocol violations
    • 1% of patients in intervention group ‘crossed-over’ to control group
    • 9% of patients had other protocol violations

Control

  • Selective CT scanning
    • Chest and pelvic x-ray and FAST scan performed during primary survey
    • Selective CT scanning. Indications were pre-defined according to local protocols, listed here
    • 46% of patients underwent sequential CT scans of all body regions comprising a total body CT scan in the end
  • Protocol violations
    • 3% of patients ‘crossed-over’ to intervention group
    • 11% of patients had other protocol violations

Outcome

  • Primary outcome: In-hospital mortality – No significant difference
    • 15.9% in the intervention group vs. 15.7% in the control group, p=0.923, Fragility index -22 patients with 117 patients lost to follow-up (Declined participation or language barrier)
  • Secondary outcomes:
    • Radiation exposure was significantly increased in the intervention group
      • Radiation exposure in the ED (median): 20.9mSv (IQR 20.6-20.9) vs. 20.6mSv (IQR 9.9-22.1, p<0.0001
      • Radiation exposure during hospital admission (median): 21mSv (IQR 20.9-25.2) vs. 20.6mSv (11.8-27.6), p<0.0001
      • In the control group 45% of patients had a radiation dose that was below the lowest dose of patients who underwent a total body CT scan
    • Time to imaging completed in the ED – significantly reduced in the intervention group
      • 30min vs. 37min , p<0.0001
    • Time to diagnosis – significantly reduced in the intervention group
      • 50min vs. 58min, p<0.001
    • Time spent in the trauma room – no significant difference
      • 63min vs. 72 min, p=0.067
    • Direct medical costs – no significant difference
      • 24,967 EUROS vs. 26,995 EUROS, p=0.439
    • Number of missed injuries found during tertiary survey – no significant difference
      • 8.8% vs. 10.1%, p=0.45
    • Life threatening event during scanning – no significant difference
      • 0.6% vs. 0.2%, p=0.374
  • Pre-determined sub-group analysis
    • No significant difference in hospital mortality for:
      • Patients with ISS ≥16: 22.4% vs 24.8%, p=0.457
      • Traumatic brain injury: 38.2% vs. 43.7%, p=0.311

Authors’ Conclusions

  • The use of immediate whole body CT scan did not reduce mortality compared with standard radiological work up

Strengths

  • Randomised controlled trial
  • Multi-centre
  • Clear criteria for performing CT scans in control group
  • Allocation concealment maintained with use of computer randomisation

Weaknesses

  • Large number of patients excluded post randomisation. This is a major weakness. Clinicians would have known what group the patient had been assigned to when the decision was made to exclude them. This may have introduced bias to the results.
  • Inclusion/exclusion criteria subjective – although this could be argued to be pragmatic, I feel this is a weakness when so many patients were excluded following a disagreement over whether these subjective criteria had been met. The inclusion/exclusion criteria were also modified after 62 patients had already been enrolled.
  • Differences in baseline characteristics with more patients in the intervention group having an ISS score of ≥16. On first glance this may appear to be a weakness but in reality it may just be that the whole body CT scan diagnosed more injuries and therefore these patients had a higher ISS score.

The Bottom Line

  • For patients with suspected major trauma who are managed in a high functioning level 1 trauma centre, the use of immediate whole body CT scan compared to focused CT scanning resulted in no change in mortality but an increased radiation dose. Focused CT scanning was ultimately a total body CT scan for 46% of patients in the focused group due to the combination of segmental scans.

External Links

Metadata

Summary author: David Slessor
Summary date: 4th July 2016
Peer-review editor: Duncan Chambler

11 comments

  • Pingback: LITFL Review 239 – FOAM Ed

  • Nick Tarmey

    So the “radiation exposure during hospital admission” was significantly lower in the control group, but look at the absolute difference: 21mSv (IQR 20.9-25.2) vs. 20.6mSv (11.8-27.6). Is a mean difference of 0.4 mSv something we need to worry about from a total dose of around 21 mSv in both groups?

    • David Slessor

      Thanks for your comments. The differences in radiation exposure were differences in the median levels. As 46% of patients in the focused groups had the equivalent of whole body CT scans it is expected that the median dose of radiation will be similar. However for the patients who did not have whole body CT scans then the radiation dose was often significantly lower, as demonstrated by the lower IQR.
      I think that trauma CT is very important tool in the management of trauma. however, there is often ‘investigation creep’ with the use of trauma CT for more liberal indications. I think that for some patients the risks outweigh the benefits, however the challenge is deciding which these patients are. I would also have some reservations about applying this data to units other than high functioning level 1 trauma units as I think that trauma CT will help to minimise differences between excellent and poorly performing units.

      • Thanks David and Nick, we also noted the very small difference in radiation exposure overall and felt that thie essentially meant that there was little difference.

        David – you rightly point out that the distribution of scores is different and that raises an interesting question. Are we trying to reduce individual doses or population doses of radiation.

        On an individual basis the risks are still small, very small in fact. At these levels we will not cause cancers, se just raise an individuals (small) additional risk.

        We would argue that it is the population risk that matters here as it is the overall radiation dose to the population of patients we see that is the real difference – and yet there isn’t really a clinically significant one.

        vb

        S

        • David Slessor

          Thank you Simon for your comments

          For this study I find it difficult to compare the population risk based on median values. With nearly half of the intervention group having the equivalent of whole body CT the median radiation exposure does not provide very useful information. What I want to know is, in the population that you decide not to do a whole body CT immediately following your primary survey, what radiation is this population exposed to?

          In the control group 45% of patients had a radiation dose that was below the lowest dose of patients who underwent a total body CT scan. This suggests to me that the total radiation dose for this population may be reduced by a clinically significant amount. As there were no complications related to this strategy it may be an appropriate strategy for units to adopt. However, as mentioned before this may not apply if you do not work in a high functioning level 1 trauma centre. If you were to change practice on the basis of this paper then you would need to collect data on your practice to ensure that there are no missed injuries or delays in e.g. getting to theatre, as any complications would negate the benefits of what as you mentioned on the individual basis if a very small change in radiation dose.

      • Duncan Chambler

        I agree with David’s comments: a whole body CT scan reviewed by a consultant radiologist is a safer approach in smaller hospitals that don’t have 24/7 expert skills in trauma management and point-of-care radiology/ultrasonography skills. This trial suggests that both strategies are reasonable for patient outcomes and safety, so for me it now comes down to system design and staff skill mix.

  • Pingback: JC: Do we always need a whole body CT in trauma? St.Emlyn's - St.Emlyn's

  • Pingback: CME 11/08/16 - REACT 2 Trial - Imaging in Blunt Trauma - Charlie's ED

  • Pingback: Total-body CT of conventionele beeldvorming met CT op indicatie bij traumapatiënt. Wat kies jij? | fan of EM

  • Pingback: SGEM#181: Did You Ever Have to Make Up Your Mind, Pan Scan or Leave Other Scans Behind? | The Skeptics Guide to Emergency Medicine

  • Pingback: emDOCs.net – Emergency Medicine EducationThe Reign of the “Pan-Scan”: Whole Body CT vs. Selective Imaging in Trauma - emDOCs.net - Emergency Medicine Education

Leave a Reply to Simon Carley Cancel reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.