Eligible patients need to have any intracranial bleeding on CT scan OR, if no scan is available, a GCS of 12 or less. Why use a GCS of 12 or less as one of the eligibility criteria?
We know from the CRASH-2 trial that treatment with tranexamic acid is time critical so there should not be a great delay waiting for a CT scan. In some countries there is quite a long wait for a patient to have one or CT scanning facilities may not be available at all.
Therefore, an easy measure is needed allowing us to capture the patients with a high risk of having an intracranial bleed. A total of 7,796 patients (78%) of the 10,008 recruited in the CRASH-1 trial had a CT scan. There was evidence of intracranial bleeding (defined as the presence of evacuated haematoma, non evacuated haematoma, subarachnoid haemorrhage or petechial haemorrhages) in 49%, 64% and 80% of patients with mild, moderate and severe TBI respectively. The frequency of intracranial bleeding for patients with GCS <13 was 73%. In the CRASH-2 Intracranial Bleeding Study, 84% of patients with GCS <13 had an intracranial bleed (defined as the presence of parenchymal, subdural or epidural haemorrhage).
If a patient is randomised using the GCS and is later found not to have a bleed on CT, the treatment should simply be stopped and the data collected as per the protocol. As an intention to treat trial, these patients remain in the trial and their data will contribute to the final analysis. Recruitment of patients who subsequently have no bleed on a CT scan will contribute to this trial. This is the nature of a pragmatic trial.
Do hospitals taking part in the trial need to manage traumatic brain injured patients in the same way?
In a large trial involving hundreds of hospitals internationally it is inevitable that there will be some differences in the management of patients between hospitals. However, these will not bias the trial results. Providing that the trial is large enough, randomisation will ensure that the intervention and control groups are identical with regard to both known and unknown confounders.
It is possible that the size of the intervention effect may vary a little bit depending on the type of management given. Nevertheless, any differences would not affect the direction of the treatment effects. Patients in the future will almost certainly receive different forms of care than they do today. Treatments shown to be effective today may be more or less effective in the future, but the direction of the effect will be the same. Rather than standardise patient management, it is more important that the trials are large enough.
How will participants be randomised?
Randomisation codes have been generated and secured by an independent statistical consultant. Randomisations were balanced in blocks of 8 (corresponding to a box of trial treatment packs) so that each box contains either tranexamic acid or placebo in random order. The codes will be made available to the drug packing company explicitly for the treatment packs to be created in accordance with the randomisation list.
Baseline information will be collected on the trial entry form and after eligibility has been confirmed and appropriate consent process complete, the lowest consecutively numbered pack will be taken from the box of eight treatment packs. At the point when the treatment ampoules in both dose packs are confirmed to be intact, the patient is considered to be randomised onto the trial. The completed entry form data will be sent to the Trial Coordinating Centre as soon as possible. Once randomised, the outcome of the woman should be obtained even if the trial treatment is interrupted or is not actually given.
What is the standard treatment for TBI?
There is a wide spectrum of treatments for Traumatic Brain Injury (TBI). The treatment given is based on the clinical judgement of the doctor with primary responsibility for the care of a particular patient. There is no need to withhold any clinically indicated treatment in this trial. The trial treatment (tranexamic acid or placebo) would be provided as an additional treatment to the usual management of TBI. Giving the trial treatment must not delay any other clinical decisions (eg the need for surgery) or giving additional medical interventions.
Surgery is indicated in a randomised patient, what do I do?
The decision for surgery will be that of the doctor treating the patient. Inclusion of the patient in this trial will not alter that. As indicated in the Protocol and in the Investigator’s Manual of Operating Procedures, patients should receive ALL clinically indicated treatments (which may include surgery). There is no standard process for deciding when surgery is indicated and all investigators should follow their clinical judgment.
Can I randomise a patient that has been administered warfarin?
There is no absolute contraindication to randomising a warfarinised patient and we anticipate that many older patients are likely to bleed for exactly that reason. What we know is that there is no direct interaction between PCC (Factors II, VII, IX & X) and tranexamic acid. Factors II, VII, IX and X do not have lysine binding sites and this is the pathway by which tranexamic acid acts. Although PCCs have been associated with increased thrombotic risk in some settings, there is no specific contraindication to its use with tranexamic acid.
When should a professional legal representative (PrR) and follow-up consent be considered?
It would appear from our consent monitoring that the waiver of consent procedure is being used well. However, using a PrR consent procedure seems to be problematic. It is only being considered as a last resort when all else fails. This should not be the case. Remember that written consent should be considered as soon as possible after the emergency is over. At this stage, neither the patient nor their personal representative may have the capacity to consent. PrR consent should be considered at this stage. Continue to re-assess the patient / relative for their capacity to consent and obtain informed consent where possible.
What monitoring procedures are in place to ensure patient safety?
In the CRASH-3 trial death, life-threatening complications, and prolonged hospital stay are pre-specified outcomes that are collected up to day 28. Therefore, each patient in the trial is assessed specially for these events on the Outcome form. To assure the safety of patients in the study, an independent Data Monitoring Committee has been appointed to oversee the safety monitoring. All outcomes (including mortality, vascular occlusive events and seizures) will be collected and presented to the Data Monitoring Committee for an unblinded review. The Data Monitoring Committee will meet on an ongoing basis to review accumulated data and advise the Trial Steering Committee regarding the continuing safety of current participants and those yet to be recruited, as well as reviewing the validity and scientific merit of the trial.
What happens if an adverse event occurs after discharge?
If a discharged patient is subsequently readmitted to hospital, has any untoward medical event, or is known to have died after she is discharged, this should be reported as an adverse event using the adverse event form, if it occurs within 28 days of randomisation. An Adverse Event is defined in the Protocol as ‘any untoward medical occurrence’. Please see the Protocol and the Manual of Operating Procedures for further details.
All patients randomised into the trial should be given a patient alert card if discharged before day 28. The patient and their representative should be advised that the patient should; carry the card for at least 28 days after entry to the trial (the date will be written inside the card) present the card to any healthcare provider she or he sees tell their family about the card so that in the event the patient is seriously ill a family member can present the card to the healthcare provider on the patient’s behalf.
Are there additional safety measures that should be considered for vulnerable subjects (expectant mother/mental impairment)?
The CRASH-3 trial involves vulnerable patients who will have mental impairment because of their significant head injury. Patients with head injury have a high risk of death and severe disability. It is important that they can contribute to clinical trials to improve the outcomes for patients with this devastating condition. We have considered the safety of the patients fully and the trial procedures do not involve greater than minimal risk for any research.
The intervention to be used has been well tested in thousands of patients and is in routine clinical use for over 40 years. The safety profile is well documented and we have provided full evidence of this in the Investigator’s Brochure.
Because of the high risk of death from a head injury, pregnant women are not excluded. The treating doctor will use the uncertainty principle to decide if the risks of the trial outweigh any potential benefit; if yes, the patient will not be randomised.
The trial will not interfere with any other clinical interventions. Patients will receive all clinically indicated treatments.
The trial treatment is a simple intravenous infusion – a common procedure to qualified doctors and nurses globally. All data collected is routine clinical data.
The consent procedure has been considered carefully and will allow the family, if present, to input as much as possible in the decision making even though there is insufficient time for fully informed consent (see Protocol Section 2.6 page 9). In addition, full information will be given to the family / patient as soon as possible and informed consent obtained. Patients or their relatives are free to withdraw from the trial at anytime without any effect on the patient’s clinical care.
Patients /relatives are not being paid for participation. This reduces the risk of coercion.
The proposed dose of 2 gm IV for all patients does not take into account patients with renal impairment. Is there evidence that the proposed dose is safe in patients with renal dysfunction?
The administration of multiple doses of tranexamic acid (TXA) should be done with care in any patient (irrespective of age) with renal impairment. This is because its administration is associated with a risk of accumulation in patients with renal dysfunction. In the CRASH-3 trial, a total of 2 grams of TXA is given over 8 hours therefore the risk of accumulation should be minimal because there is no repeated administration as anticipated in the Pharmacia’s Summary of Product Characteristics.
It is important to bear in mind that the CRASH-3 trial uses the uncertainty principle, whereby patients are excluded from the trial if the treating clinician believes there is any indication OR contraindication to tranexamic acid. Therefore if the treating doctor considers that a patient with renal impairment has a contraindication to TXA, that patient should be excluded from the trial. Each clinician is responsible for balancing the risks and benefits associated with the administration of TXA in a particular patient based on clinical condition of the patient and the information provided about the drug in the Investigator’s Brochure. The same principle applies whether the patient is elderly and no specific cut-off age is specified.
In this trial, it would not be appropriate to mandate a renal function test on which eligibility will be based. This trial reflects the real world scenario. A patient with Traumatic Brain Injury is in a critical emergency condition and may be at a high risk of death. It may not be known in advance if the patient has renal impairment as eligible patients will have impaired/altered consciousness due to their brain injury and no prior history may be available. Therefore, clinicians will need to make a judgement (as they would for any other treatment they initiate in the emergency setting).
Whilst we do not mandate blood testing pre or post randomisation, if the patient is found to have any subsequent abnormalities through routine blood testing, which would make the trial treatment contraindicated, the trial treatment should simply be stopped. There is no need to adjust the dose. After the initial 2 doses, which will be completed over approximately 8 hours, there is no further trial treatment. Tranexamic acid has a short half-life and is 90% eliminated within 12 hours of administration, therefore the risk of accumulation is very low.
It is worth noting that the dose of TXA used in this trial was used in 20,211 patients enrolled in the CRASH-2 trial and was found to be both effective and safe in reducing the number of death without any increase in complications. In addition, we are collecting information on renal failure which will be closely monitored by the Independent Data Monitoring Committee.
Is co-enrollment allowed?
The CRASH-3 trial is a pragmatic study and co-enrolment is allowed, if there is no biological interaction between the interventions and the other trial protocol allows for this. We are happy to discuss on a case by case basis if co-enrolment is a possibility at your hospital.