In addition to the breaking trials at the #CCR meetings a number of other important trials were presented that you should seek out and review. We’ve not done full critical appraisals on these trials, that’s up to you to do, but we’re happy to signpost them here.
The TRACT trial1,2
Presented by Prof. Kath Maitland, this study of transfusion thresholds in African children demonstrated that we can accept low thresholds for transfusion in children. The World Health Organization recommends not performing transfusions in African children hospitalized for uncomplicated severe anemia (hemoglobin level of 4 to 6 g per deciliter and no signs of clinical severity). The question posed was whether immediate transfusion or delayed transfusion was possible and then whether transfusing 20ml/Kg was better than 30ml/Kg in these children.
The main results are presented below. It’s worth mentioning that lots of these kids were really sick.
Severe and complicated anaemia plus no fever
- 30ml/Kg halved mortality vs 20ml/Kg at a cost of $20 per life saved
- two thirds of children in the trial
Severe and complicated anaemia plus fever
- 30ml/Kg transfusion strategy doubled mortality
On another note, why has Kath Maitland not been given an OBE or more yet? She is just amazing and has transformed our approach to treating the critically unwell and injured. Can someone please nominate her….
The COACT trial3
The Coronary Angiography after Cardiac Arrest without ST-Segment Elevation trial was published in the NEJM earlier this year and we covered it on the blog earlier this year4.
In the ED we often focus out learning and skills on improving our ability to get ROSC, but what next? If we get ROSC should the patient go to ICU, CCU or direct to the cath lab? This is a real question as moving patients to the cath lab is not always straightforward as they be quite unstable and in some centres it may be tricky to manage things like temperature control.
For those patients who have ST elevation on the post ROSC ECG I don’t think there is any argument – they go straight to the cath lab. What about those without ST elevation though?
The COACT trial looked specifically at this. An RCT of 552 patients randomised to either immediate cath lab transfer or a more delayed approach.
At #CCR20 we learned that the long term outcomes were similar in both groups.
The editorial presented by Chris Nickson at #CCR20 takes the approach that clinical context is important here and that there are inevitably issues in open label trials. There are issues with selection bias perhaps with quite a large number of screened patients failing to be included in the study – we don’t know why. Chris also pointed out that it’s probably underpowered. In terms of sample size the aim of reducing mortality by 13% is unrealistic, combined with a lower than expected coronary lesion rate it does look underpowered. It’s worth considering delta inflation in a lot of critical care trials.
So the bottom line appears that there is no real need to go direct to cath lab in this group. Probably because they represent a broad range of pathologies and that’s important because this does not mean that in practice none of these patients should be directly transferred. It does mean that you should stop and think. For example, the patient with prior ischaemic symptoms is different from the 24 year old athlete who collapsed whilst playing football. The first should go, the second maybe not just yet.
This trial of dexmedetomide vs. usual sedation in ICU patients failed to show a difference in 90-day mortality.
At #CCR20 we saw some subgroup analysis that looked at the older populations which we can’t report here as it’s embargoed, but there is some interesting data to look at that we hope to see in print soon. It’s only a secondary analysis though so the it’s hypothesis generating and there will be a SPICE IV (Ed – or perhaps ‘Old Spice’).
Read more about this trial on the bottom line blog5 which raises some important questions. Why was mortality the primary outcome here? Sure, it’s patient centred but how would we pathophysiologically explain choice of sedation to this end point as opposed to the quality of sedation and complications? In that respect the secondary outcomes are not favourable to dexmedetomidine. Celia Bradford on the bottom line blog concludes that there is little compelling here to use dexmedetomidine unless we are considering some specific subgroups of highly agitated patients.
Named after a greek god, the HYPERION trial looked at the use of moderate hypothermia in post cardiac arrest patients with non-shockable rhythms. This is important as most of the other targeted temperature trials have looked patients with shockable rhythms.
They analysed 581 patients with a primary outcome of survival with a favourable neurologic outcome, assessed on day 90 after randomisation with the use of the Cerebral Performance Category (CPC) scale.
ICU-ROX10,11 and MEGA-ROX
This is the Conservative Oxygen Therapy during Mechanical Ventilation in the ICU trial published in NEJM in 201910. It compared a liberal oxygen strategy in ventilated patients vs. a more conservative one. They found no real difference in ventilator free days (the primary outcome) and little difference in any of the secondary outcomes.
The rationale for this trial is that high levels of oxygen is detrimental to health and there is some data to support this12,13. If you want an overview from the main author then follow the links below.
The sub group analyses also suggested that there is a potential benefit to a conservative approach to oxygen therapy, but these should be considered hypothesis generating, although if you’re looking for evidence to support a conservative approach I suspect you will find them somewhat supportive, especially the mortality data in the patients with ischaemic brain injury where there was a significant benefit for the conservative management of ICU patients.
The bottom line is that it does not seem to matter what strategy you use, but there are patho-physiological arguments for us to aim for as low as required FiO2 in ICU ventilated patients.
Paul also talked about MEGA-ROX which will look at an even more liberal use of oxygen which is a more complex trial using an adaptive trial design. We’ve not seen many of these in EM/CC but there are significant reasons why this may be a way forward for some of the more complex clinical questions we face.
MEGA-ROX is designed to be very easy to recruit to and will be another large scale international study. It will also have planned co-trial enrolment into sepsis, cardiac and brain injury trials of oxygen therapy. We don’t have a huge amount of detail as yet, but this really looks like a study that will answer some fairly fundamental questions about oxygen therapy on the ICU. This is based on some quite different outcomes for these subgroups in the ICU-ROX patients14,15. Patients with sepsis did worse with conservative management, whereas brain injured patients did better. These were big changes and therefore those hypotheses need to be explored in MEGA-ROX
Panel discussion on trial design.
A theme through day 2 and indeed the world of FOAMed is the interpretation of results in trials. Even common stats such as the p-value are poorly used and understood. For example many people don’t know that the p-value is based on the null hypothesis and thus relies on some assumption of this. We also see the p-value used in a very dichotomous way, which is rather limited and misses out the richness of the data. Journals such as the NEJM and Nature are taking an active approach to not simply putting p-values in the trials to be replaced with confidence intervals or estimate of effects 16,17 . Here at St Emlyn’s we do agree that p-values are limiting, but there is a coherent argument that the primary outcome should be tested. However, papers often have swathes of p-values that are not adjusted for multiple analysis and that cannot be right. Howard Bauchner reminded us that journals are looking for clarity from authors that they did what they said they were going to do. He also discussed how approaches to statistical analysis goes through phases and perhaps the current vogue for bashing p-values may just be a fashionable phase.
We then moved onto the classic threshold of p<0.05 as a marker of significance. The clear message here is that using this level for everything is crazy. Some interventions may need a higher standard of proof before we expose our patients or ourselves to the findings. We do need to consider moving away from p<0.05 as being a universal threshold of significance but rather tailor it to the question we are asking in each trial. Josh Farkas has a great blog on this at the EMCRIT site18.
Chris Nickson reminded us that we should interpret new studies in the light of past evidence. No trial stands alone and that may influence how we think about the strength of published results.
We also need to be careful to look for trials that are registered in advance and who publish their statistical methods in advance. Most of the major journals do this, but even then there are problems. Triallists are now putting huge numbers of secondary outcomes on their pre-planned analyses. We could say that this is just sensible, or we could say that this is cynical ploy to ensure that something comes up as positive.
The fragility index had a bit of a bashing on stage with some logical arguments. Personally I still like it from a communication perspective. Humans are naturally better at interpreting natural frequencies and whilst statisticians don’t need it, many people do. I think we should keep it in the interests of communication and learning.
The pros and cons of social media were debated (again). I think the conclusion is that we need to engage in this space in order to champion science. That’s something that we would agree with and would hope that sites like St Emlyn’s and Critical Care Reviews are part of the solution and not the problem.
A fantastic first day in Belfast. We’re off to the social event on the Nomadic tonight, the only remaining White Star Line ship in existence. We’ll be back tomorrow.
Also check out our more in depth analysis of the 65 trial here. The 65 trial was published today at #CCR20 and shows that it is probably fine to aim for 60-65mmHg MAP in septic patients aged over 65. Our blog on the Global Sepsis Study will be up shortly too.
- 1.Maitland K, Kiguli S, Olupot-Olupot P, et al. Immediate Transfusion in African Children with Uncomplicated Severe Anemia. N Engl J Med. August 2019:407-419. doi:10.1056/nejmoa1900105
- 2.Ingelfinger JR. Immediate Transfusion and Transfusion Volume in African Children with Severe Anemia. N Engl J Med. August 2019:475-476. doi:10.1056/nejme1908869
- 3.Lemkes JS, Janssens GN, van der Hoeven NW, et al. Coronary Angiography after Cardiac Arrest without ST-Segment Elevation. N Engl J Med. April 2019:1397-1407. doi:10.1056/nejmoa1816897
- 4.Carley S. Top 10 papers 2018-2019 for RCEM. St Emlyn’s. https://www.stemlynsblog.org/jc-top-10-papers-2018-2019-for-rcem-annual-scientific-conference-st-emlyns/. Published 2019. Accessed 2020.
- 5.Bradford C. SPICE III. The Bottom Line. https://www.thebottomline.org.uk/summaries/icm/spice-iii/. Published 2019. Accessed 2020.
- 6.Shehabi Y, Howe BD, Bellomo R, et al. Early Sedation with Dexmedetomidine in Critically Ill Patients. N Engl J Med. June 2019:2506-2517. doi:10.1056/nejmoa1904710
- 7.Farkas J. Hyperion. EMCRIT. https://emcrit.org/pulmcrit/hyperion/. Published 2019. Accessed 2020.
- 8.Lascarrou J-B, Merdji H, Le Gouge A, et al. Targeted Temperature Management for Cardiac Arrest with Nonshockable Rhythm. N Engl J Med. December 2019:2327-2337. doi:10.1056/nejmoa1906661
- 9.Olusanya S. Hyperion. The Bottom Line. https://www.thebottomline.org.uk/summaries/hyperion/. Published 2019. Accessed 2020.
- 10.Conservative Oxygen Therapy during Mechanical Ventilation in the ICU. N Engl J Med. October 2019. doi:10.1056/nejmoa1903297
- 11.Bradford C. ICU-ROX. The Bottom Line. https://www.thebottomline.org.uk/summaries/icm/icu-rox/. Published 2019. Accessed 2020.
- 12.Chu DK, Kim LH-Y, Young PJ, et al. Mortality and morbidity in acutely ill adults treated with liberal versus conservative oxygen therapy (IOTA): a systematic review and meta-analysis. The Lancet. April 2018:1693-1705. doi:10.1016/s0140-6736(18)30479-3
- 13.Girardis M, Busani S, Damiani E, et al. Effect of Conservative vs Conventional Oxygen Therapy on Mortality Among Patients in an Intensive Care Unit. JAMA. October 2016:1583. doi:10.1001/jama.2016.11993
- 14.Young P, Mackle D, et al. Conservative oxygen therapy for mechanically ventilated adults with sepsis: a post hoc analysis of data from the intensive care unit randomized trial comparing two approaches to oxygen therapy (ICU-ROX). Intensive Care Med. November 2019:17-26. doi:10.1007/s00134-019-05857-x
- 15.Perner A, De Jong A, Shankar-Hari M. Trials on oxygen supplementation in sepsis: better late than never. Intensive Care Med. November 2019:116-118. doi:10.1007/s00134-019-05874-w
- 16.Harrington D, D’Agostino RB Sr, Gatsonis C, et al. New Guidelines for Statistical Reporting in the Journal. N Engl J Med. July 2019:285-286. doi:10.1056/nejme1906559
- 17.Amrhein V, Greenland S, McShane B. Scientists rise up against statistical significance. Nature. March 2019:305-307. doi:10.1038/d41586-019-00857-9
- 18.Farkas J. Demystifying the p-value. EMCRIT. https://emcrit.org/pulmcrit/demystifying-the-p-value/. Published 2015. Accessed 2020.