Education News and Reviews

Read our latest blog articles and Journal Club reviews, catch up on our Quants or post your own article

Latest education news from CCCEG

Quisque eleifend elit a mauris malesuada rutrum. Quisque fermentum nibh in libero gravida auctor.

Presented by Capt (MD) Dimitrios Sidiras HAF


In patients undergoing cardiac surgery, who have a moderate-to-high risk of death, is a restrictive transfusion strategy (trigger = 75 g/L) non-inferior to a liberal strategy (Trigger = 95 g/L intraoperatively/ICU or 85 g/L on the ward) with respect to a composite outcome of death, myocardial infarction, stroke and acute kidney injury requiring dialysis at 6 months after surgery?


The most common situation leading to massive transfusion is cardiac surgery, but trauma, where physical injury and blood loss combine, remains the best-studied example and trauma patients are the other large group of patients consuming large amounts of red cells.

We currently follow the NICE guidelines for red blood cells transfusion which sets a threshold of 70 g/L and a hemoglobin concentration target of 70–90 g/L after transfusion (for patients with acute coronary syndrome, the trigger is 80 g/L and the hemoglobin concentration target is 80–100 g/L after transfusion). Our local TBI protocol sets a threshold of 90 g/L. This does not apply to major haemorrhage, history of acute coronary syndrome or chronic anaemia.

Trauma patients in need of the major haemorrhage protocol are transfused in the Emergency Department, aiming for a haemoglobin concentration target above 80 g/L. However the suspicion of major haemorrhage is not based on haemoglobin concentration, but instead on the mechanism of injury and the vital signs.


At present most centres are using the 2017 EACTS/EACTA Guidelines on patient blood management for adult cardiac surgery (or something very close to that). In addition, we need to find a balance between optimal patient care and the appropriate use of a limited resource.

The two approaches to blood transfusion currently are:

  • The restrictive approach, which is based on limiting the infectious and non infectious (TRALI, pulmonary oedema, multiorgan system dysfunction etc.) risks of transfusion
  • The liberal approach, which is based on the fact that patients who are at high perioperative risk may be more susceptible to anaemia-induced tissue hypoxia

When it comes to the brain…..the evidence is conflicting!

There is clear clinical and guideline agreement that Hb less than 70 g/L in critically ill patients with TBI requires red blood cell transfusion (RBCT). However, the exact threshold between 70 and 100 g/L remains a contentious issue. Recent data from a randomized controlled trial (RCT) and meta-analysis found no difference in neurological outcome between the restrictive and liberal transfusion strategies, but the overall quality of the evidence was low. Several observational studies have shown an association between anemia and poor outcomes in patients with TBI. However, other studies evaluating anemia and TBI outcomes have not demonstrated a consistent risk of harm (see further reading section).

Additionally, several studies have shown that RBCT administration in TBI is associated with increased mortality, decreased functional outcomes, increased ICU length of stay, and impaired cerebral autoregulation. A liberal transfusion strategy applying a threshold trigger of 100 g/l was associated with an increased risk of progressive cerebral hemorrhagic injury and thromboembolic events. However, evidence from other observational studies in patients with TBI is conflicting, with data to support a lack of association between RBCT administration and worse outcome in TBI.


Patients undergoing cardiac surgery often have significant transfusion requirements and there is wide variation in transfusion practices among clinicians and institutions. Robust evidence has emerged in other patient populations (general critical illness, septic shock, etc) that a restrictive transfusion strategy is at least as good as (and possibly better than) a liberal one. If a restrictive strategy is shown to be noninferior to a liberal one, and this can influence practice, there is the potential for significant cost savings and decreased use of a scarce and limited resource.


  • Should we compare transfusion practices in cardiac surgery with those in trauma and TBI?
  • Transfusion thresholds for trauma and TBI. We aim for higher than other critically ill groups. Why? Do latest studies agree with that?
  • This study is unusual in that it is about patients at a higher risk of death and other complications than some of the other studies.
  • It also shows that a RCT overturned TITRe2
  • It focuses on transfusion of red blood cells only
  • No data exists for transfusion and TBI. How would data like CENTER-TBI answer this question in TBI?
  • How often do we ignore guidelines and stick to personal experience or instincts?


We are already following a rather restrictive strategy when it comes to transfusing red blood cells, so we are probably going to the right direction according to this study. Furthermore, we base our practice on a multivariable approach that includes fluid resuscitation and coagulopathy.

Maybe we should wait for the next big “transfusion in trauma” or “transfusion in TBI”  study focusing on non-inferiority of restrictive transfusion strategy and then check how cost/effective is our practice. In any case, excessive concern about cost-effectiveness shouldn’t change our approach to cases where we highly suspect major haemorrhage: We should have a low threshold for transfusion.


The Bottom Line

Hematology Times


Transfusion practices in traumatic brain injury

Variation in blood transfusion and coagulation management in TBI at the ICU

RBCT in Patients With TBI: A Systematic Review and Meta-Analysis

Journal Club: BREATHE (JAMA 2018)

Effect of Protocolised Weaning With Early Extubation to Noninvasive Ventilation vs Invasive Weaning on Time to Liberation From Mechanical Ventilation Among Patients With Respiratory Failure

JAMA, 2018. doi:10.1001/jama.2018.13763


Presented by Dr Joel Davis, ACCS CMT


In adults in whom weaning from invasive mechanical ventilation is “difficult” (ie a spontaneous breathing trial has failed), does protocolised weaning with early extubation to non-invasive ventilation reduce time to liberation from ventilation, compared with protocolised invasive weaning?


Invasive mechanical ventilation is one of the most common lifesaving interventions performed on our unit, but prolonged ventilation is associated with increased morbidity and mortality.

The authors of this paper state that 33% – 86% of invasively ventilated patients are “difficult to wean”, and this concords with our experience on the JVF. To provide a snapshot example, on the unit today we have 13 ventilated patients, of whom 8 have undergone tracheostomy – most of them for actual or anticipated difficulty weaning.

These patients consume a disproportionate amount of our resources. Any intervention that can reduce the morbidity, mortality, length of stay and cost associated with this would be very welcome.


Several previous studies have tried to address this question, but the majority of patients enrolled had COPD. In that cohort, weaning with NIV has been shown to reduce mortality as well as the rates of tracheostomy and ventilator-associated pneumonia (see the meta-analysis by Burns et al However, it is not known whether these results are generalizable to a modern general ICU population, in which fewer COPD patients are now invasively ventilated.



The earlier trials referred to above suggested that weaning to NIV does produce clinically important and statistically significant benefits for patients with COPD, but this does not reflect the case mix of ventilated patients on our ICU, who may have different pathological processes and lung mechanics hampering their weaning.

Importantly, this new trial also sought to mitigate a key source of potential bias by ensuring that both the control and intervention groups underwent protocolised weaning. In most previous studies, a protocol had been adopted only for the intervention group, making it difficult to distinguish whether any beneficial effects arose from the actual intervention or simply from the protocolisation.


  • Considering they had all failed an initial SBT, the patients in both arms of this trial had a surprisingly low duration of mechanical ventilation (2.9 days on average). This may have been because of the strict protocols, that called for weaning assessments q2 hours. It might not be feasible to offer this to every patient on the JVF, but should we at least consider protocolising our weaning plans more consistently?
  • Does it matter that only a small proportion of eligible patients were included (and we don’t know why); or that almost half of the patients were recruited from only 3 centres?
  • Although no benefit was seen in the primary outcome of liberation from all ventilation, the secondary outcome of total invasive ventilation time was reduced. This is surely beneficial – should it alone prompt us to change practice?
  • On the other hand, extubation to NIV did bring with it an increased risk of reintubation – is this a major concern? And what about other risks? (No mortality difference was seen, but the study was inadequately powered to detect one. The authors accept that reintubation is risky – see
  • What are our patients’ preferences?? (MyICUVoice could be very helpful here!)


As explained in an accompanying JAMA editorial (see below), the importance of this paper is in challenging the “dogma” that patients who fail an SBT necessarily require ongoing invasive ventilation. As we all know, reducing the total time that patients spend intubated could have major benefits for both the patient and for our ICU resource allocation, so these results suggest that we should strongly consider extubating to NIV in such cases – provided we recognise and mitigate any potential risks, particularly the increased risk of reintubation.

In fact, on the JVF we are increasingly extubating patients onto high-flow nasal oxygen (HFNO) when we anticipate that they will require extra support. In a 2016 paper, Hernandez et al showed HFNO to be non-inferior to NIV for patients they defined as ‘high-risk’ (although they had not necessarily actually failed a SBT). So, could HFNO provide another alternative strategy for extubating patients who fail their SBT?

The questions we should consider going forward are:

  • In spite of the negative results of this trial, are there cohorts of patients who would be liberated from ventilation more quickly if extubated to NIV after failing an SBT?
  • Even if the answer is no, should we consider it anyway, to reduce the time (and associated risk, resources etc) of invasive ventilation? Or is the reintubation risk unacceptable?
  • Whenever we find ourselves considering extubation to NIV, could we achieve comparable outcomes with HFNO instead? (More resource-friendly and almost certainly a less traumatic experience for the patient!)



A nice editorial of this article and its importance –


2013 meta-analysis of previous trials –


Summary of the Hernandez trial (HFNO in high-risk patients) –

And a useful lit review on the same topic –