Broomedocs is on the SGEM: RePHILL Trial

This week I am making a guest appearance on my Canadian mate Ken Milne’s amazing podcast. Below is the written version of the chat we had about the recent RePHILL trial: prehospital blood products for shocked trauma patients.

If you are not already subscribed – please go over to the Skeptics Guide to Emergency Medicine to hear the banter and subscribe to this amazingly frequent and thorough resource.

Reference: Crombie et al. Resuscitation with blood products in patients with trauma-related haemorrhagic shock receiving prehospital care (RePHILL): a multicentre, open-label, randomised, controlled, phase 3 trialThe Lancet Haematology 2022

Guest Skeptic: Dr. Casey Parker is a Rural Generalist that includes in his practice emergency medicine, anaesthesia and critical care. He is also now a fully fledged “sonologist”. Casey currently splits his time between Broome, a small rural hospital in the remote Kimberley region of Western Australia, and a large tertiary ED in sunny Perth.

Case: You are working in the emergency department (ED) and receive a call from the Advanced Care Paramedics who are at the scene of a stabbing. Apparently, two rival gangs (Jets and Sharks) had a rumble. The young man has been stabbed in the abdomen and lost a lot of blood. The patient is tachycardic (120 beats/minute), hypotensive (80/60 mmHg) and looks very pale. They have two large bore intravenous (IV) access and are planning to bring them to your ED as soon as possible. The paramedic asks you, “we have saline, and we also have red-cells and this fancy new lyophilised plasma.  Should we give our shocked patient saline or plasma / red cells en route to the ED?”  What do you advise him? 

Background: The use of fluids in trauma resuscitation has been studied in a number of trials in recent years. A lot of observational data has been collected from the battlefields of Iraq and Afghanistan.

The Control of Major Bleeding After Trauma (COMBAT) Trial was published in the Lancet in 2018. It was a pragmatic, randomised, single-centre trial done at the Denver looking at the use of plasma in the prehospital setting. This trial did not show a statistical mortality benefit within 28 days of injury. First10EM and REBEL EM both did a review of the COMBAT trial. 

The Prehospital Air Medical Plasma (PAMPer) trial was published in NEJM, also in 2018.  The goal of this trial was to determine the efficacy and safety of prehospital administration of thawed plasma in injured patients who are at risk for hemorrhagic shock. This trial did report that prehospital administration of plasma was safe and resulted in lower 30-day mortality. PAMPer was reviewed by First10EM and The Bottom Line

The traditional teaching in trauma is to replace blood with blood products, so we would expect that we should see a benefit if we used blood and plasma instead of saline alone for the initial resuscitation.


Here is a PICO summary of the trial:

  • Population: Adult patients 16 years of age or older suffering traumatic injury resulting in shock believed to be due to a traumatic haemorrhage. Shock was defined as a systolic BP less than 90 mmHg or an absent radial pulse. 
    • Exclusions: Patients known to refuse blood produces, those who received transfusion of prehospital blood products before assessment for eligibility, pregnancy (known or apparent), isolated head injury without evidence of major haemorrhage, and prisoners. 
  • Intervention: Up to four units of blood products boluses one unit at a time.  Units were alternating between units of O-negative packed red cells (PRBC) or or reconstituted lyophilised plasma to a maximum of two units of either.
  • Comparison: Up to four boluses of 0.9% saline (250ml/bolus)
  • Outcome: 
    • Primary Outcome: Composite outcome of mortality from time of injury to hospital discharge or the failure to clear lactate by 20% within the first two hours after randomisation.
    • Secondary Outcomes: Individual components of the composite primary outcome, fluid volumes, measures of coagulopathy, 30 day mortality and common side effects of transfusion such as reaction and adult respiratory distress syndrome (ARDS)
  • Trial: This was a multi-centre,open-label, concealed, single-blinded, randomised controlled trail

Authors’ Conclusions: The trial did not show that prehospital PRBC–LyoPlas resuscitation was superior to 0.9% sodium chloride for adult patients with trauma related haemorrhagic shock. Further research is required to identify the characteristics of patients who might benefit from prehospital transfusion and to identify the optimal outcomes for transfusion trials in major trauma. The decision to commit to routine prehospital transfusion will require careful consideration by all stakeholders.”

Quality Checklist for Randomized Clinical Trials:

  1. The study population included or focused on those in the emergency department. No
  2. The patients were adequately randomized. Yes
  3. The randomization process was concealed. Yes
  4. The patients were analyzed in the groups to which they were randomized. Yes
  5. The study patients were recruited consecutively (i.e. no selection bias). Unsure
  6. The patients in both groups were similar with respect to prognostic factors. Yes
  7. All participants (patients, clinicians, outcome assessors) were unaware of group allocation. No
  8. All groups were treated equally except for the intervention. Yes
  9. Follow-up was complete (i.e. at least 80% for both groups). Yes
  10. All patient-important outcomes were considered. Unsure
  11. The treatment effect was large enough and precise enough to be clinically significant. No
  12. Lack of financial conflicts of interest. Yes

Results: The trial randomized 432 participants. Median age was 38 years, 82% were male and 62% of patients had blunt force trauma from a motor vehicle collision. 


  • Primary Outcome: Composite outcome of mortality from time of injury to hospital discharge or the failure to clear lactate by 20% within the first two hours after randomisation.
    • 64% in the blood product group vs 65% of the saline group 
    • Adjusted risk difference –0·025% [95% CI –9.0 to 9.0], p=0·996)
    • Adjusted risk ratio 1.01 [95% CI 0.88 to 1.17]
  • Secondary Outcomes: 
    • Mortality: 43% in the blood products group vs 45% in the saline group (Adjusted risk ratio 0.97 [95% CI 0.78 to 1.20]; p=0·75) 
    • Failure to clear lactate: 50 vs 55% (Adjusted risk ratio 0.94 [95% CI 0.78 to 1.13]; p=0·52)
    • None of the other secondary outcomes reported were statistically different
    • Serious adverse events were similar between both groups

1. Mixing POOs and LOOs: Is it appropriate to place LOOs (lab-oriented outcomes) and POOs (patient-oriented outcomes) together in a composite primary? This trial used a composite primary outcome that included  mortality (a big POO) with lactate clearance (a lab-oriented outcome that may or may not be a surrogate marker for mortality).  Usually one might design a trial with softer LOOs if one were trying to detect a subtle difference in a population where hard patient-oriented outcomes are rare. However, in this trial nearly half of the patients died.  Mortality is very objective and important to patients, and it would therefore seem better to stick with a single primary outcome and power the recruitment to that end.

2. Dose: The RePHILL trial participants received less than a litre of fluid in total prior to arriving in hospital. That could either be all saline in the control group or a mixture of saline and blood products in the intervention group.  The fact that this was a “negative” trial may mean that there truly is no difference between saline and blood products in pre-hospital resuscitation. Or, it could mean that at this dosage there is no detectable potential benefit or potential harm observable. For example, If we only gave 10 mg of aspirin to patients with ST elevated myocardial infarction and observe no mortality benefit, but also no GI bleeds we could conclude that aspirin has no effect. However, we do have high-quality data showing that there is a benefit to giving 162.5mg of aspirin to STEMI patients with a small increase in harm ( So is the RePHILL trial telling us there is no benefit, or that the doses are not adequate to give an effect that we can observe? This is why it is good to be cautious not to over or under-interpret the data. The most accurate conclusion is that the intervention provided in this cohort of trauma patients with shock did not demonstrate superiority over the control group. This is different than concluding blood products do not work for trauma patients in the prehospital setting. 

3. External Validity: This trial used a team of pre-hospital physicians and critical care paramedics in the UK. The team traveled by helicopter or land based rapid response vehicles with a blood product that is relatively new and hard to come by (lyophilised plasma).  If RePHILL had been a positive trial and shown a clear benefit then we would be stuck with the issue of external validity, especially in rural areas like Canada and Australia which are much larger countries geographically.  We do not have the systems or access to this product. It would be very difficult to bring a physical-led plasma wielding team to the roadside in most parts of the world. Rural clinicians know about the tyranny of distance.  The goal should be to get the best medical care to rural patients which can be logistically tough at times. However, it should not be the knowledge of the team that results in lesser care.

4. Too Good to Be True: The original power calculation for RePHILL was based on the consensus that a 10% absolute difference in the primary outcome. This is a large difference for a complex disease like trauma using a simple intervention. However, the previous PAMPer trial did show an amazingly high 9.8% mortality reduction.  The RePHILL trial was unfortunately interrupted by COVID19 after 432 patients were randomised. It is unlikely that if RePHILL has enrolled 68 more patients for a total of 500 it would have shown a 10% absolute reduction in the primary outcome. We know that science usually iterative and moves in baby steps. In diseases like sepsis or trauma where there are complex interventions and systems involved it is really very unlikely that any single intervention will have such a large mortality impact.  This suggests that the PAMPer trial is an outlier and we should remain skeptical as these results seem too good to be true. Trials with results that seem too good to be true are rarely reproduced on subsequent studies.  We think RePHILL is such a case. It could be viewed as a partial repeat of the PAMPer trial and is more consistent with the COMBAT trial results demonstrating no superiority.

5. The Goldilocks Zone of Mortality: When looking at mortality data there are at least three possible populations:

  1. Mild Traumatic Injuries: Patients who are not too sick and 100% will survive, no matter what intervention. This is why we do not put ankle sprains into trauma trials like RePHILL.
  2. Severe Traumatic Injuries: These are the super sick patients whom will die no matter what we do. For example if you only studied patients with hypotension due to penetrating heart injury then you would find it difficult to prove any intervention helped them
  3. Goldilocks Zone: These patients are the ones that are not too sick, but not too well to get a clinically meaningful benefit from an intervention. The patients in the the RePHILL trail seem to be within the Goldilocks zone – high mortality for sure, but with potentially salvageable injuries that need to get to definitive care

Comment on Authors’ Conclusion Compared to SGEM Conclusion: We generally agree with the authors’ conclusions.  This trial does not demonstrate any benefit to the use of small volumes of prehospital blood products over the use of normal saline. 


Case Resolution: I will advise the prehospital team to use whatever fluid they have and feel is necessary to resuscitate the patient whilst expediting transfer to the ED.  A trauma call is put out to activate the lab, surgical teams and radiology in order to provide rapid care to this very sick patient.

Clinical Application: The choice of pre-hospital fluids remains an open question.  We simply do not know if there is a real benefit to early infusion of blood products in the field.  

There are certainly some patients who will benefit from the use of blood products – although at this time it seems that we can limit the use of blood products to the ED and hospital phase of care. Further research is needed to define the group patient whom may benefit from earlier blood in the prehospital environment.

Having said that – trauma is a team sport, and having a seamless system of care between the prehospital providers, the ED, and the operating room is crucial.  

What Do I Tell the Patient? You have been stabbed and lost a lot of blood. We are going to start some intravenous fluids and get you to the hospital as quick as possible. They will be standing by with a team to help you on arrival. 

OF course, if you work in remote Australia or Canada where the transfer to definitive surgical care is counted in days, not minutes… then you need to look at the system and recognise that this trial cannot tel you what to do for those long hours. Blood with blood still makes empirical sense, but we work in an evidence free zone out here! This is where you get to be a real doctor and treat the individual in front of you!

Add a Comment

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