Severe traumatic brain injury reguidelined

  • October 3, 2016

Guidelines for the management of severe traumatic brain injury

Brain trauma foundation: 4th edition, 2016

 

This is a brief summary and some of the wording is changed for brevity, hopefully without too much loss of accuracy.

Although not explicitly stated on the guidelines, ‘Severe brain injury’ is generally defined as a brain injury resulting in a sustained [for a variable number hours] Glasgow Coma Scale of 3 to 8.

Also look at the recent RESCUEicp trial that’s referred to.

Decompressive craniectomy (DC)

Level II A only

  • Bifrontal DC is not recommended in severe TBI with diffuse injury and ICP >20mmHg (for >15min in any hour). (re GOS at 6 months). May decrease ICU days.
  • A large, rather than small, frontotemporoparietal DC is recommended severe TBI (better mortality and neurologic outcomes)

NB RESCUEicp trial may change this (checkout this video re DC).

Changes from Prior Edition

New topic.

 

Prophylactic hypothermia

Level II B only

  • Early (within 2.5 hours), short-term (48 hours post-injury) prophylactic hypothermia is not recommended in diffuse injury.

Changes from Prior Edition

No change

 

Hyperosmolar therapy

Level I, II, and III

  • May lower ICP but insufficient evidence on other outcomes.

Old recommendations not brought forward but restated (acknowledging complexity and evidence gaps):

  • Consider Mannitol (0.25 g/kg to 1 g/kg) for ICP. Avoid arterial hypotension (systolic blood pressure <90 mm Hg).
  • Restrict mannitol use prior to ICP monitoring to patients with signs of transtentorial herniation or progressive neurological deterioration not attributable to extracranial causes.

 

CSF Drainage

Level III

  • An EVD system zeroed at the midbrain with continuous drainage of CSF may be considered to lower ICP burden more effectively than intermittent use.
  • Use of CSF drainage to lower ICP in patients with an initial Glasgow Coma Scale (GCS) <6 during the first 12 hours after injury may be considered.

Changes from Prior Edition

New topic.

 

Ventilation

Level II B

  • Prolonged prophylactic hyperventilation with partial pressure of carbon dioxide in arterial blood (PaCO2) of 25 mm Hg or less is not recommended.

Old recommendations, not brought forward, yet restated!

  • Hyperventilation is recommended as a temporizing measure for the reduction of elevated ICP.
  • Hyperventilation should be avoided during the first 24 hours after injury.
  • If hyperventilation is used, use SjO2 or BtpO2 monitoring.

 

 

Anaesthetics, analgesics and sedatives

Level II B

  • Administration of barbiturates to induce burst suppression measured by EEG as prophylaxis against the development of intracranial hypertension is not recommended.
  • High-dose barbiturate administration is recommended to control elevated ICP refractory to maximum standard medical and surgical treatment. Hemodynamic stability is essential before and during barbiturate therapy.
  • Although propofol improves ICP it doesn’t improve mortality or 6-month outcomes. Caution is required as high-dose propofol can produce significant morbidity.

Changes from Prior Edition

Wording changed only.

 

Steroids

Level I

  • The use of steroids is not recommended for improving outcome or reducing ICP (increases mortality)

Changes from Prior Edition

No changes

 

Nutrition

Level II A

  • Basal caloric replacement at least by the fifth day and, at most, by the seventh day post-injury is recommended to decrease mortality.

Level II B

  • Transgastric jejunal feeding is recommended to reduce the incidence of ventilator-associated pneumonia.

Changes from Prior Edition

Emphasised timing and route.

 

Infection Prophylaxis

Level II A

  • Early tracheostomy is when the overall benefit is felt to outweigh the complications associated with such a procedure. (Fewer ventilated days but no less pneumonia).
  • The use of povidone-iodine (PI) oral care is not recommended. (No less VAP)

Level III

  • Antimicrobial-impregnated catheters may be considered to prevent catheter-related infections during EVD.

Changes from Prior Edition

Removed Periprocedural antibiotics line. Also emphasised VAP is more common in TBI patients.

 

DVT prophylaxis

Level III

  • Low molecular weight heparin (LMWH) or low-dose unfractioned heparin may be used in combination with mechanical prophylaxis. However, there is an increased risk for expansion of intracranial hemorrhage.
  • In addition to compression stockings, pharmacologic prophylaxis may be considered if the brain injury is stable and the benefit is considered to outweigh the risk of increased intracranial hemorrhage.

No evidence for specific agent, dose, or timing of pharmacologic prophylaxis.

Changes from Prior Edition

Comment re compression stockings added.

 

Seizure prophylaxis

Level II A

  • Prophylactic use of phenytoin or valproate is not recommended (re late (>7 day) PTS).
  • Phenytoin is recommended to decrease the incidence of early PTS, when the overall benefit is felt to outweigh the complications associated with such treatment. NB early PTS have not been associated with worse outcomes.
  • No evidence for levetiracetam over phenytoin.

Changes from Prior Edition

No change.

 

ICP monitoring

Level II B

  • Management of severe TBI patients using information from ICP monitoring is recommended to reduce in-hospital and 2-week post-injury mortality.

Old recommendations not carried forward (weak evidence) but re-stated them:

  • Intracranial pressure (ICP) should be monitored in all salvageable patients with a severe traumatic brain injury (TBI) (GCS 3-8 after resuscitation) and an abnormal CT. An abnormal CT (hematomas, contusions, swelling, herniation, or compressed basal cisterns).
  • ICP monitoring is indicated in patients with severe TBI with a normal CT an 2 or more of:
    • age over 40 years
    • unilateral or bilateral motor posturing
    • systolic blood pressure (BP) <90 mm Hg.

 

Cerebral perfusion pressure monitoring

Level II B

  • Management of severe TBI patients using guidelines-based recommendations for CPP monitoring is recommended to decrease 2-week mortality.

Changes from prior edition

Separated threshold recommendations into different section.

 

Advanced cerebral monitoring

Level III

  • Jugular bulb monitoring of arteriovenous oxygen content difference (AVDO2), as a

source of information for management decisions, may be considered to reduce mortality

and improve outcomes at 3 and 6 months post-injury.

Changes from Prior Edition

Removed brain tissue oxygenation rcommendation removed due to conflicting evidence.

Thresholds

BP

Level III

  • Maintaining SBP at ≥100 mm Hg for patients 50 to 69 years old or at ≥110 mm Hg or above for patients 15 to 49 or over 70 years old may be considered to decrease mortality and improve outcomes.

Changes from Prior Edition

New evidence has refined the recommendation.

 

ICP

Level II B

  • Treating ICP above 22 mm Hg is recommended because values above this level are associated with increased mortality.

Level III

  • A combination of ICP values and clinical and brain CT findings may be used to make management decisions.

NB  RESCUEicp released soon after Guidelines formed.

Changes from Prior Edition

Updated after new study.

 

CPP

Level II B

  • The recommended target CPP for survival and favourable outcomes is between 60 and 70 mm Hg. Whether 60 or 70 mm Hg is unclear, decide for each patient.

Level III

  • Avoiding aggressive attempts to maintain CPP above 70 mm Hg with fluids and pressor may be considered because of the risk of adult respiratory failure.

Changes from Prior Edition

Refined after several studies.

 

Advanced cerebral monitoring

Level III

  • Jugular venous saturation of <50% may be a threshold to avoid in order to reduce mortality and improve outcomes.

Changes from Prior Edition

Removed recommendation re brain tissue oxygen monitoring.

 

btf4-changes

Sepsis 3: rise of the SOFA

  • February 25, 2016

Sepsis remains a syndrome without a validated criterion or standard diagnostic test but increasingly comprehensive data sets are making it easier to test validity of  potential criteria. The current instalment of this work in progress has refined the definition of sepsis in light of recent analyses and opinion.

Sepsis is life-threatening organ dysfunction caused by a dysregulated host response to infection

In more detail, sepsis is ‘a multifaceted host response to an infecting pathogen’. It may be significantly amplified by endogenous factors and it involves early activation of both pro and anti-inflammatory responses and major modifications in non-immunologic pathways such as cardiovascular, neuronal,  autonomic, hormonal, bio-energetic, metabolic, and coagulation.

Unhelpful truths

  • It’s clear that organ dysfunction, even when severe, is not associated with substantial cell death.
  • Patients with sepsis are in clinically very heterogeneous groups (age, co-morbidities, injuries, medicines etc.).
  • Sepsis-related organ dysfunction may be occult AND unrecognised infection may be the cause of new organ dysfunction.

Dear SIRS“Operationalization”

Current definitions have had problems with practical applicability in clinical situations. For example, at least 12% of septic patients don’t have SIRS criteria. And many non-septic patients tick the SIRS boxes.

Are you comfortable with SOFA?

Organ dysfunction is, now, an acute change in total SOFA score of 2 or more points, due to the infection. Assume baseline SOFA of zero, unless you know otherwise. In the presence of infection a SOFA of 2 or more is associated with a mortality rate of 10%.

SOFA JAMA

Further extensive multivariable regression of studies has led to the formulation of the quickSOFA score (qSOFA). Just three parameters can be used to identify patients who are at risk of long stays or ICU death.

A high risk patient has 2 of:

  1. GCS 13 or less
  2. Systolic BP < 100
  3. Respiratory rate > 22

Subgroups of sepsis

Septic shock has been “a state of acute circulatory failure” for many years now despite acknowledgement that this definition is too confined.

This consensus specifies septic shock as sepsis with underlying circulatory and cellular/metabolic abnormalities profound enough to substantially increase mortality (>40%). More pragmatically:

 Persisting hypotension requiring vasopressors to maintain MAP > 65 and

serum lactate >2 mmol/L

despite adequate volume resuscitation

Setic Shock JAMA

Severe sepsis is now felt not to be a useful term.

In summary

  • Sepsis is life-threatening organ dysfunction caused by a dysregulated host response to infection
  • We continue to be suspect infection in the normal ways; rash, consolidation, peritonitis, dysuria etc.
  • We will continue to wait for dysregulated host responses to be manifest as organ dysfunction.
  • SOFA and qSOFA scoring offer a tested means of risk assessing

So what do we do now with the literature on sepsis? Does it all need to be redone or re-analysed? 

In addition, this provides more impetus for the development of integrated patient monitoring and information systems. Automated SOFA scoring is surely the corollary of this work, but remains a distant prospect for many hospitals.

All the 2s? 
Beware:
  • SOFA: decline by 2
  • qSOFA: 2 off your GCS, resps of 22, and 2 figure SBP. Any 2 of these.

If it goes unnoticed, you might be calling 2222!

Comment elsewhere

EMCrit/PulmCCM sobering analysis of the new consensus definitions

RebelEM on Sepsis 3.0

StEmlyn’s on Batman and Sepsis 3

JAMA sepsis 3 explanatory videos

Fighting ‘traditional nihilism’ re haematological malignancy in ICU

  • November 7, 2015

A guidance group and a sounding board of 50 UK experts has published guidelines on the management and ICU admission of critically ill adult patients with haematological malignancies. Part of the driving force for this was the ‘traditional nihilism’ in the critical care community about these patients, meaning disagreements regarding the benefit of admission to ICU are not uncommon (~15%).

Consequently this guideline is a mixture encouragement to be less pessimistic and a lot of reiteration of common sense. This is my rough summary and take on it.

Making the case

Data from small, retrospective, single centre studies in the 80’s and 90’s suggested a very poor outcome in leaukaemic patients requiring advanced organ support; up to 100% mortality if mechanical ventilation was used. As a result consensus by 2000 was that patients with haematological malignancy and metastatic solid organ tumours were not ‘good candidates’ for critical care admission.

However over subsequent years larger studies, mainly from databases, showed improving survival and suggested markers of poor prognosis (high APAHCE, vasopressor use, BM transplant, advanced age, and both disease and genetic subtypes). This is a result of improvements in chemotherapy, BMT preparation as well as supportive care.

Mortality data from ICNARC published in 2009 revealed an ICU and hospital mortality of around 40% and 60% respectively.

Of note
  • Sepsis and respiratory failure remain the most common reason for ICU involvement.
  • Recent analysis suggests that older patients and those needing vasopressors do badly, and that dual antibiotic therapy (aminoglycoside + another) is protective. Survival rates had improved significantly over the course of the study.
  • Patients who are admitted and are deteriorating after the first 3 days are likely to do badly, particularly those starting pressors or renal replacement after day 3.
  • Even now, as with many of our patient groups, scores may predict how populations will do but none is accurate for your particular patient.
  • The concept of a trial of ICU may be of value in these patients. Look for the more powerful prognostic markers on day 4/5 (MV, pressors and renal support).
  • Aside from the standard considerations when making a decision to admit (premorbid state, prognosis, previous discussions re treatment limitations etc), remember to ask about:
  1. special transfusion requirements (e.g. irradiated products, cytomegalovirus (CMV)-screened)
  2. history of bleomycin exposure/radiotherapy
  3. isolation needs
 Legrande 2012

 

Recommendations

  1. An aggregate track and trigger system should be in place to monitor patients. (1C)
  2. Surviving Sepsis campaign guidelines and NICE guidance on neutropoenic sepsis should be followed. (1C)
  3. ICU referral should involve direct discussion between an ICU consultant and a haematology consultant. (1D)
  4. Patients who are clearly in the process of dying with irreversible illness should not be referred to critical care. (1D)
  5. Survival is largely determined by the acute critical illness. Patients appropriate for further life-extending treatment or with good performance status should be considered for an unrestricted critical care ‘ICU Trial’. (1C)
  6. Use of non-invasive ventilation (NIV) in the ward for these patients is not advised. (1C)
  7. Patients who have undergone allogeneic haemopoietic stem cell transplantation (HSCT) should be transferred to an ICU attached to a level 3 British Committee for Standards in Haematology (BCSH) unit. (1D)
  8. Inter-hospital and intra-hospital transfer should follow the Intensive Care Society (ICS) 2011 Guidelines for the Transport of the Critically Ill Adult. (1C)

 

 

Much like the exposure of misplaced pessimism surrounding COPD patients in ICU a decade ago, we need to make sure our reluctance to admit adult patients with haematological malignancy to the unit is based on up-to-date outcome stats and opinions.

Which group next for pessimism busting? Metastatic solid organ tumours?

NOAC? No problem – maybe soon.

  • November 3, 2015

Dabigatran is a direct thrombin inhibitor rapidly gaining popularity due to growing consensus that it is associated with less life-threatening bleeding than vitamin K antagonists. Measures to reverse the agent have a small evidence base and are limited to prothrombin complex concentrates, factor VIIa (less so) and haemodialysis/filtration (more so). The NIHR briefing from January summarizes near-current consensus.

A number of agents are currently being tested. 2 recent trials have looked at the most promising reversal agent, Idarucizumab, a monoclonal antibody fragment that irreversibly binds the drug.

 

In the Lancet, a randomised, placebo-controlled, double-blind phase 1 trial of Idarucizumab also recently appeared.

Four groups of healthy males (47 in total) were given dabigatran for four days and then received either placebo or 1 g, 2 g, 4 g or 5g of idarucizumab two hours after the last dabigatran (peak plasma time)

All doses immediately reversed the dabigatran although the 1 g dose didn’t achieve a sustained effect.

Interestingly plasma concentrations of total dabigatran initially increased, but not its anticoagulation effect. This is presumed to be due to diffusion of unbound extravascular drug (idarucizumab has a low Vd) into the vascular space. Although Dabigatran-idarucizumab complex is excreted by the kidneys a recurrence of anticoagulation is not expected.

A thorough look-out for adverse events found nothing concerning

Idarucizumab Lancet

In the NEJM, RE-VERSE AD was prospective cohort study of  the safety and efficacy of 5 g of idarucizumab in significant haemorrhage, or those requiring an urgent procedure. The primary endpoint was percentage reversal within 4 hours. The secondary end point was the restoration of haemostasis.

Dilute thrombin time and ecarin clotting time (recognized tests to monitor dabigatran levels) were normalized within minutes in the bleeding group. The median time to haemostasis was around 11 hours. In the group undergoing urgent procedures 33 of the 36 haemostasis during the procedure was sufficient.

There was only one episode of thrombosis, occurring 72 hours after the dose.

Of course, Idarucizmab won’t work for bleeding in the presence of factor Xa inhibitors (the ‘-Xabans’) for which the favoured treatment is Prothrombin Complex Concentrate. Parallel studies will hopefully offer encouraging results in the near future, for example, using Feiba.

So it looks good both in healthy volunteers and two small groups of patients. We await comparative studies and further safety data whilst managing these patients with our current ?suboptimal strategies. It may be infrequent that it’s really required but knowing it’s there would be reassuring!

(To impress on a ward round with your mastery of pronouncing awkward monoclonal antibodies, try “I dare you kiss Uma b”)


Other comment:

Incisive and more critical review from RebelEM

Intensivising death in the elderly

  • September 7, 2015

In admitting patients to the ICU we have, at some level, made a decision that we can benefit the patient. The decision is more difficult in the elderly where the difference between preventing and prolonging death can be cloudy. This is particularly the case where the patient looks appropriate on paper, but less so from the end of the bed.

3 articles on the elderly ICU patient, by Canadian researchers, have been published in the last couple of months. They shed light on current practice and its inconsistencies, encourage some personal practice refection and provide grounds for research into refining ICU admission decisions.

 

Long-Term Association Between Frailty and Health-Related Quality of Life Among Survivors of Critical Illness: A Prospective Multicentre Cohort Study

To look at the association between pre-admission frailty and long-term quality of life they performed a multicentre cohort study involving around 400 over 50 year-olds. A Clinical Frailty Scale score  >4 was used to diagnose frailty and a number of scores used to assess health at 6 and 12 months (including EuroQol and Short-form 12). 38% were lost to follow-up through death (or other reasons).

  • On multiple logistic regression, frail patients who survived had significantly less mobility, self-care ability, activity level and more pain and anxiety.
  • Additionally there was no real improvement between physical, cognitive and life quality scores between 6 and 12 months.

The Very Elderly Admitted to ICU: A Quality Finish

Nearly 1700 patients aged 80 or more, in 24 Canadian ICUs (mixed medical and surgical) were followed for a year. 30% had some degree of frailty. From a nested cohort they asked family members whether they would prefer ‘life-support” or “comfort care without life support”. Although only 51% said they would want life support for their relative, 85% got some MV, renal support or vasoactive drug infusions.

  • A third of patients died in hospital (median time to death was 10 days, median LOS was just 4 days); half of whom died on mechanical ventilation.
  • Half of the survivors made it home, rather than a care facility. So only a third of admissions were discharged to home.
  • In the cases where the family were unsure about the level of care, death was significantly prolonged.
  • Frailty was significantly associated with mortality.
  • Advanced directives or obvious frailty made little difference to ICU admission rates.

They discuss how the apparent overuse of intensive care is at odds with the current end-of-life care concern both in the media and literature.

 

Frailty

Recovery after critical illness in patients aged 80 years or older: a multi-center prospective observational cohort study

In order to look at 12 month outcomes, the same Canadian group followed 610 patients over 8 years-old who had  been on the ICU. These patients were compared with a parallel ‘hospital cohort’ of patients who were not managed on the ICU.

A 43 point frailty index was calculated according to the family’s reports of each patient 2 weeks prior to admission. Follow-up was done using the SF-36 form (a 0-100 scale with physical, cognitive and psychological components).

  • At 1 year, 50% had died. Importantly patients dying within 24 hours were not indcluded in the cohort.
  • Strikingly, only a quarter of patients had made it back to an equivalent state of health and quality of life by 1 year.
  • Frailty was a better predictor of death and disability than age, illness severity or comorbidity. For each 0.2 on the score, the OR for death was increased by 0.48.

 

Recap on frailty

Frailty is yet to achieve a formal quantitative definition. It represents loss of physical and cognitive reserve and is associated with poor outcomes after critical illness. It generally includes measurable and subjective components such as grip strength, easy exhaustion, unintentional weight loss, slow walking speed or low level of activity. The frailty syndromes that should prompt deeper concern and more formal assessment are:

  1. falls
  2. new immobility
  3. delirium
  4. incontinence
  5. susceptibility to side effects.

A comprehensive assessment involves a multidimensional look both at the patient just prior to admission and at what the trend has been in recent months. Despite growing appreciation and delineation of frailty it must be the case that the fragile patient who has been steady for years is different from the similar looking patient who has deteriorated in recent months.

This is in essence pulling apart and itemising that gestalt, end-of-the-bed impression that this patient ‘just won’t do’.

[important]
Take home messages:

For the very elderly admitted to the ICU mortality is high and few return to their previous health and quality of life within 1 year.

Frailty:

Probably predicts mortality and disability better than age, comorbidity and severity of illness.

Also predicts quality of life of the survivors of ICU.

Scoring is in its infancy.

Although it’s often difficult to gauge the patients wishes at the time of consideration for ICU admission it is probably the case that we too often treat more aggressively than the patient would have wanted.

ICU is not a pause button.

[/important]

 

STOP-IT sooner – short sharp anti-bug course for intra-abdominal sepsis?

  • July 3, 2015

Complicated intra-abdominal infection is a common case on the units admitting general surgical patients. By definition it is infection that extends beyond the viscus of origin into the peritoneal space, and is associated with abscess or peritonitis.
The traditional management approach has been to do the source control and then soak in antibiotics until signs of inflammatory response have settled, usually a good couple of weeks. This is despite guidelines moving towards a 4-10 day course in recent years (eg IDSA or WSES).

 

In this pragmatic, multicenter (USA and Canada), non-inferiority, randomised trial, published in NEJM, 518 patients with complicated intra-abdominal infection (source control achieved) AND fever or leucocytosis or peritonitis-related gastrointestinal dysfunction (not managing half their ideal nutrition), were given one of two antimicrobial plans:

  1. 4 day course post procedure.
  2. continued therapy until 48 hours of temperature <38, white cells <11, >50% standard nutrition,  or up to 10 days.

 

The primary outcome was a composite of occurrence of surgical site infection, recurrence of intra-abdominal infection, or death.

There was no difference in primary outcome (~20%), or any element of it (eg death – both around 1%).

There was no difference in emergence of C.difficile or resistant organisms. 

IA infection from NEJM

Potential problems with using this study to inform practice on ICU:

  • The patients were generally mildly unwell (APACHE 2 of 10) – perhaps not as sick as the cases we deal with. The subgroup with APACHE > 10 also had no difference in outcome but this is a post-hoc remark only.
  • The power calculation was based on a higher complication rate, leaving the study underpowered.
  • A considerable number of each group didn’t get the specified treatment (71/260 in control, 47/258 in experimental), mainly because the antibiotic course length was too short or long. Non-adherence can mean bias, in this case favouring the null hypothesis.
  • Funding disappeared after the fist interim analysis.
  • It doesn’t tell us what to do with immunosuppressed patients.

 

This suggests the 20% that develop complications don’t get them from a lack of antibiotic but probably form failed procedural source control.

All in all this supports, but not confirms, the practice of those in favour of short antimicrobial courses for intra-abdominal infection (post attempted source control). Saving pennies and perhaps looking after the microbiome are sensible considerations after all. 

Refine STOP-IT and set it on the ICU.

 

 

[notice]

As a reminder, the IDSA list the following as risk factors for source control failure in these patients:

  1. Delay in the initial intervention (124 h)
  2. High severity of illness (APACHE II score 15)
  3. Advanced age
  4. Comorbidity and degree of organ dysfunction
  5. Low albumin level
  6. Poor nutritional status
  7. Degree of peritoneal involvement or diffuse peritonitis Inability to achieve adequate debridement or control of drainage
  8. Presence of malignancy
[/notice]

 

 

PermiT underfeeding?

  • June 6, 2015

High calorie intake on ICU has long been known to be a bad idea. Topping up early enteral feeding deficits with parenteral nutrition is now felt to be non-life-saving and possibly dangerous. Animal studies have suggested that it is the protein element of feeding that keeps the mitochondria happy, not calories per se.

So, the PermiT study, a refocussed repeat of a previous trial, was designed to answer the question of whether targeting protein delivery but forgoing some calories would be beneficial in the critically ill.This was an unblinded, pragmatic RCT in Canada and Saudi Arabia on around 900 mixed specialty critically ill patients.

For 14 days, half were given 40-60% of the (Penn-State) calculated calorie requirement, and the others were given 70-100%. Both had 1.2-1.5g/Kg of protein each day and some multivitamins. Almost all was given enterally.

There was no difference between the 2 groups in primary end-point 90 day mortality, or any secondary outcome (28, 90 or 180 day morality, or ventilator-free days).Underfed

Interestingly, post-hoc analysis showed a lower fluid balance and less use of renal support in the underfed.

 

 

Reasons you should hesitate to form a strong opinion based on this are:

  • The randomised patients comprised just 14% of the potential eligible group.
  • The duration of intervention was 14 days. Nutrition after this was not controlled.
  • Low number of surgical patients.
  • Very low use of parenteral nutrition.

 

These weaknesses aside, this work adds weight to the growing refinement of calorie/protein/route feeding in the critically ill. Underfed but sufficiently fed (‘satisfed!’)? Perhaps we shouldn’t struggle too hard to get the calories in. And should we consider separating our feeding constituents and prioritising the protein?

 

FLORALI – breathe (HFHO2) in through the nose

  • May 28, 2015

High flow humidified nasal oxygen for respiratory failure has been gaining popularity both in adult and paediatric care in recent years. A number of studies have shown promising results.

Announced last year but published this month, FLORALI was a multicenter, sealed randomization, pragmatic RCT with a well-defined study protocol, performed in more than 20 ICUs in France and Belgium. Patients with hypoxic respiratory failure (Pa/FiO2 <39.9 KPa) almost entirely due to pneumonia) but without chronic lung disease, pulmonary oedema, shock GCS<12, hypercapnia or neutropoenia, were included. Obviously those needing emergent intubation were also excluded. 313 patients were randomised to receive:

  1. face mask oxygen at 10l/min
  2. high-flow (warm humidified) nasal oxygen (starting at 50l/min)
  3. NIV (for at least 8hr/day, PEEP 2-10, Vt 7-10ml/kg, high-flow nasal oxygen when not on NIV)

The primary outcome was intubation within 28 days.

Of 4777 admissions with respiratory failure only 512 were eligible, and 313 randomised successfully.

  • Overall there was no significant difference in the primary outcome between the 3 groups. There was an overall trend towards superiority of high-flow nasal oxygen, but the subgroup with more severe hypoxia (PaO2:FiO2 ratio of <26.7KPa) showed a significant benefit.
  • 90 day mortality was a lower in the high-flow nasal oxygen group. But what does this mean if the intubation rate was no different? It was also a secondary outcome so suffers from our usual concerns.
  • High flow nasal oxygen was clearly more comfortable.

FLORALI diag

 

 

 

 

 

 

 

Interestingly there was no significant difference among the groups in terms of the time until intubation or the reasons for intubation, suggesting NIV failure per se was not necessarily deleterious.

The lack of difference for intubation rates could have been due to slight under-powering from an over-estimation of the likely intubation rate when using standard oxygen therapy.Once intubated the excess mortality rate in NIV could possibly have been related to the relatively high tidal volumes in the NIV group, at around 9ml/Kg.

 

 

The exclusion criteria, although on the face of it sensible, actually meant just 525 of the 4777 patients were eligible for inclusion and only 313 received a treatment protocol. Being strict this study applies to around just 1 in 10 of our patients with respiratory failure. All in all a well-designed and well-enacted, important RCT that falls short of being practice-changing, so is hypothesis-generating.

Perhaps the benefits of additional pressure support are offset by the dangers of additional baro/volu/atelectatrauma. Maybe BiPAP subdues cough or gives rise to more persistent secretions.

High flow humidified nasal oxygen provides a tolerable way of delivering a small amount of PEEP, maybe helps clear CO2, and is at least beneficial as a first line therapy in the more severely hypoxaemic patient. The questions of whether early intubation is beneficial, whether NIV is actually detrimental

Many of us extrapolate the benefits beyond the tight confines of this trial already, but studies now need to focus on the subgroups in the exclusion set.

 

Others’ words:

The statistically awesome EMnerd review

PulmCrit – Pneumonia, BiPAP, secretions and HFNC: new lessons from FLORALI

Author JP Frat talks on the ESICM site

PulmCCM review

Bottom Line review