I had the opportunity to attend the European Society of Intensive Care Society (ESICM) annual conference in Milan this month. It was an interesting conference with expert speakers from all over the world of critical care. The conference spanned three days, and all talks were only 15 minutes followed by 2 questions. This post is a summary of my key learning points:

1. SIRS isn’t dead.

Physicians tell us of hectic fever, that in its beginning it is easy to cure, but hard to recognize; whereas, after a time, not having been detected and treated at the first, it becomes easy to recognize but impossible to cure.

Machiavelli 1513

A political commentator 500 years ago seems to have summed up sepsis pretty well!

After the launch of sepsis 3.0 at the start of the year there has been significant discussion around the clinical utility of the new definitions for sepsis and septic shock. During a Q&A session, sepsis 3.0 authors Mervyn Singer and Jean-Louis Vincent simplified matters to the following statement:

Sepsis = infection and organ dysfunction.

And it’ll be the same for the next 2000 years! Remembering this will make it easier to avoid getting lost in a maze of complex definitions and criteria.

Key message: the use of SIRS criteria to identify and diagnose infection, followed by investigating for organ dysfunction, remains a perfectly reasonable approach.

FOAMcast on sepsis definitions

NICE guidance on when to suspect sepsis

2. Questions on qSOFA

Flavia Machado, from Brazil, presented data on the use of, and challenges associated with, the qSOFA in a resource-constrained environment. Prospective data over 5 months from intensive care in the public sector of Brazil showed 62% of septic/septic shock patients admitted to ITU did not meet the qSOFA criteria, but had a 40% mortality rate. This gives food for thought, certainly in similar settings.

The issue with qSOFA @FlaviaSepsis #LIVES2016 pic.twitter.com/URYi4krT5I

— ICU Management (@ICU_Management) October 4, 2016

3. Antibiotics

A  main aim of diagnostic criteria for sepsis is to allow for prompt treatment. Professor Lipman from Australia discussed antibiotic dosing for sepsis. He presented data on the volume of distribution of antibiotics used in septic shock and the importance of achieving early minimum inhibitory concentration (MIC).

In sepsis and, more importantly, septic shock, the volume of distribution is increased so dosing of antibiotics should reflect this. His recommendation is for individualised antibiotic prescribing in the critically ill; this may involve using a loading dose to achieve satisfactory levels early. The aim is to achieve MIC before bacteria can become resistant – some organisms can start to show resistance after less than 48 hours of sub-therapeutic levels.

We were also reminded that patients always require early culturing in order to tailor antibiotic regimens as soon as possible.

There was plenty of discussion around the issue of increasing global antibiotic resistance. Interestingly, multi-resistant gram negatives appear to present the greatest challenge, and colistin is now used for empirical treatment of suspected gram-negative sepsis in some parts of America.

4. Ultrasound in the critically ill

Ultrasound already has become a mandatory tool in assessing the critically ill patient. It is essential for guiding placement of central venous catheters and assessing patients with undifferentiated shock.

Ultrasound pioneer Professor Lichenstein presented on focussed diagnostic pathways using sonography. He talked specifically about his Bedside Lung Ultrasound in Emergency (BLUE) protocol which can diagnose the 6 most common causes of acute respiratory failure within 3 minutes, when the probe is in the hands of a skilled operator (see diagram). A number of publications have externally validated the BLUE protocol with impressive negative predictive values.

Should the BLUE protocol become a part of level 1 emergency ultrasound training?

CEURF website

5. Safety and systems

Chris Subbe from Bangor hospital compared early warning scores in hospital to warning systems in fighter planes. In a cockpit there are multiple instruments displaying the same information; the primary flight display has all the data required; however, should it fail, there are identical mechanical back-ups to allow safe flying conditions.

In healthcare there is often no back-up system. For example, a distracted healthcare assistant that manually records observations might incorrectly document a patient’s vital signs, preventing the busy doctor from detecting that the patient is particularly sick until they start crashing.

Our systems need to accommodate for the inevitability of human error.

Rather excitingly, a promising new automated system of monitoring vital signs was demonstrated in an industry sponsored session. The new device automatically recorded information used in the national early warning score (NEWS) and made use of built-in thresholds for directly contacting an appropriate clinician. In the trial data shown there was significant reduction in ITU deaths suggesting more timely and appropriate care. This could be a game-changer in the ED as well.

6. ‘Too much of a good thing?’

In the hot topics session, a recent JAMA publication – the ‘Oxygen-ICU’ RCT – was presented. This was a single-centre study from Italy comparing an oxygen protocol where target saturations were kept between 94-98%, and ‘usual care’ where 97-100% were aimed for.

This was not an emergency medicine population as enrollment required an expected ITU stay of greater than 72 hours. There were also a few methodological flaws, partly because an earthquake slowed down recruitment (!) and the trial was stopped early. Nonetheless, the findings are important.

The lower target oxygen group had mortality of  11.6% compared to 20.2% in the usual care group, giving a relative risk of 0.57. This somewhat correlates with the AVOID study in acute MI where infarct size was bigger in a group given supplemental oxygen. ICU-ROX is currently recruiting in Australia for a big multi-centre RCT to further explore this subject. It appears that too much oxygen might very well be a bad thing.

Interesting editorial on the ‘Oxygen-ICU’ paper.

7. Passive leg raise for dummies

Passive leg raise as a test for fluid responsiveness in shock was discussed in a number of sessions. Some recurrent points seemed to be:

• Use a protocol to standardise the procedure
• Blood pressure itself is NOT a good marker of fluid responsiveness – cardiac output monitoring is required.
• Cardiac output monitoring needs to be readily available/continuous – changes can be transient and less than 1 minute in duration.
• End tidal CO2 in a patient with a constant minute volume can be used as marker of cardiac output in patients without advanced monitoring.
• Use of echo to measure cardiac output seems reasonable, but does have drawbacks: it is operator-dependent and not continuous, and as such requires rapid repeat examinations during the leg raise.
• A further assessment of cardiac output when the patient is back in a semi-recumbent position is required to check for a return to baseline.

Passive leg raising: five rules, not a drop of fluid! Monnet and Teboul Critical Care (2015) 19:18

8. The future: CO2 removal

Extra-corporeal membrane oxygenation (ECMO) has uses for improving survival in acute hypoxaemic respiratory failure. However, it’s not available in all centres, requires large-bore central access and, without a large caseload, is prone to complications.

Extra-corporeal carbon dioxide removal (ECCO2R) devices now exist. Procedurally, this would be more straightforward as it requires lower blood flow rates and could potentially use a similar catheter that is currently used for renal replacement therapy. Oxygenation is achieved with smaller tidal volumes than are required for adequate removal of CO2.

Extracorporal carbon dioxide removal (ECCO2R) in a near fatal asthma attack https://t.co/0GoezxZyNH pic.twitter.com/yF2ZtvuhkY

— Biomedisys SAS (@BiomedisysCol) June 22, 2016

To date there are no large positive trials and caution around adopting new technologies without good evidence should always be exercised of course; but early preliminary studies show promise. Currently SUPERNOVA is recruiting to provide more answers. Some commentary was suggesting that in 10 years all chronic respiratory failure patients could be ‘dialysing’ CO2 at home.

Summary

All in all I had a fantastic experience hearing about exciting new research and clinical developments in the management of intensive care patients. Many of the topics discussed have relevance in the emergency department – because we see them first! In particular, the passion for improving sepsis care from providers all over the world was inspiring. I hope that this knowledge will improve my management of the next critically ill patient I see in resus.

John Gilbert
@johnnyg1986

Further reading

ESICM Blog