JAMA Research at Critical Care Reviews 2024 (CCR Down Under)
17:46 Share

Apply now for the Morris Fishbein Fellowship in Medical Editing. This is a unique one-year fellowship offered by JAMA to introduce physicians to all facets of editing and publishing a major medical journal. The application deadline is January 5, 2025. For more information, including how to apply, go to jamafishbeinfellowship.com.

From the JAMA Network, this is JAMA Clinical Reviews, interviews and ideas about innovations in medicine, science, and clinical practice. Here's your host,

Hello, and welcome to our listeners around the world. Thank you for joining us today for this special edition JAMA podcast. I'm Dr. Kirsten Bibbins-Domingo, and I'm the Editor-in-Chief of JAMA and the JAMA Network. And I'm joined today by Dr. Christopher Seymour, an intensivist at the University of Pittsburgh Medical Center, JAMA Associate Editor, and the editor of the JAMA section, Caring for the Critically Ill Patient. Dr. Seymour,

We're also joined by Dr. Derek Angus, Chair of the Department of Critical Care Medicine at UPMC and Senior Editor at JAMA. We are discussing the articles published in JAMA and simultaneously presented at the Critical Care Reviews Down Under meeting in Melbourne, Australia, December 10th and 11th. These trials ask important questions. Can electronic sepsis alerts help lower in-hospital mortality?

Can different oxygen strategies change outcomes in trauma patients? Is high-flow nasal oxygen an acceptable alternative to non-invasive ventilation in patients with acute respiratory failure? And how can biomarkers help guide antibiotic duration in sepsis treatments?

Hello, Chris and Derek. Let's get started. Hi, Kirsten. Thanks for having us. Hi, Kirsten. So let's start with the sepsis trial. So the title of this trial is Electronic Sepsis Screening Among Patients Admitted to Hospital Wards. The SCREEN Trial, a Step-Wedged Cluster Randomized Trial. Derek, I'll start with you. Why don't you give us some of the summary of this trial and what the implications are?

Clinicians all over the world have been on the receiving end of lots of these so-called sepsis alerts. And there's been a lot of discussion about whether they really help or not. They cause alarm fatigue and so forth.

And there's already an expanding literature on them, but it's nearly always been observational studies. So what's interesting about this study, this is a randomized trial of the deployment of a sepsis alert. Their particular sepsis alert is one based on QSOFA that's firing on clinical signs that are available in the electronic health records.

but that maybe the clinical team hadn't quite put two and two together. So they designed a strategy where this alert would fire, and it would fire both to the bedside nurse and to the doctor on call. And they had a variety of quality improvement audits and feedback and trainings so that people knew what the alert meant.

And their design was to do a cluster design because when you're setting this alert up, you have to prepare the whole team and you have to turn it on for everyone that's being cared for over a block of time.

And they ran this design in over 60,000 patients, rolling it out sequentially in two-month blocks over about a year and a half, over what ended up being 45 separate hospital wards in multiple different hospitals. And the bottom line is the primary outcome was what happens to the overall hospital mortality rate.

for all of the patients cared for in those hospital wards when they were cared for under the alert being on or off.

And they found that there was a significant improvement in hospital mortality. The risk ratio was down to 0.85, which is significant. That's amazing. That's actually quite important. Of course, there are issues. How did it work? We know that it made them order more lactates, order more IV fluids. It didn't really change antibiotics because many of those patients were already suspected of infection and already on antibiotics.

And interestingly, if you looked at just the patients in whom the alert fired, they couldn't actually show that their mortality dropped. And there's a variety of issues to think about that, but including when you start changing alerts, you actually change the workflow for all of the patients. And you change the sense of surveillance and response to all of the different vital signs that are being triggered.

But nonetheless, we were really interested in this study because after all of this talk about sepsis alerts with all of these observational studies, now we have the first large-scale randomized trial, and the trial appears to be positive.

So an important trial, the first trial of these sepsis alerts with the results that it reduces mortality, but still questions that we're going to still be sorting through. So very interesting, very interesting. You discussed some of these issues in the accompanying editorial titled, Do Sepsis Alerts Help?

Let's go to the second trial. This is early restrictive versus liberal oxygen for trauma patients, the TROMOX2 randomized clinical trial. Chris, why don't you tell us about this trial? Sure. So it's a little change of pace here in focusing on our injured patients that come to the hospital. And as you suggested in the title, the investigators here were asking a question about how much oxygen we administer to patients with a trauma team activation.

Now, the issue of how much oxygen to treat patients with has been around for some time, but most of the studies have been in a controlled environment, such as the intensive care unit, maybe the operating room. And there's many large studies ongoing of hospitalized patients and oxygen targets. And right now, the evidence is inconclusive as to what's the best strategy.

These investigators at trauma centers in Denmark, the Netherlands, and Switzerland enrolled patients and randomized them to either receiving a restrictive approach based upon the oxygen saturation from the pulse oximeter or a more liberal approach.

And interestingly, over this eight-hour period, the investigators measured whether restrictive or a liberal approach changed death or respiratory complications at 30 days. And ultimately, they found no difference. And so this was a neutral trial, albeit quite large, multi-center. They hypothesized that

perhaps more individualized approach is needed and that a sort of one-size-fits-all oxygen strategy may not be the take-home message. Another important trial, a trial of an important topic, but one that is neutral. The authors didn't find a difference between these groups, suggesting that we need to do a little bit more research. What do you think this means for clinical practice?

Well, the guidelines right now in ATLS suggest that oxygen should be administered to these severely injured patients. But it doesn't say how much, by what route, and to what target. We learn a little bit more from this trial that

having every patient targeted at either a low or a high normal may not be superior. But there's still quite a few questions as to when to start oxygen in the pre-hospital phase and how much. And I suspect there'll be more trials in this space in the future.

Excellent, excellent. Okay, we're on to trial number three. This is high flow nasal oxygen versus non-invasive ventilation in patients with acute respiratory failure, the RENOVATE randomized clinical trial. Derek, how about this one?

So this is another really hot topic for acute care, whether you're in the emergency department, you're on the floor, or you're in the ICU. What do you do with the patient who looks like they may need to be intubated? They're clearly in trouble.

And of course, for decades, there were really only two options. You either intubated or you didn't. You put the patient on oxygen and then afterwards you could have decisional regret. Maybe I intubated the patient too early and now they're on the ventilator and they have to go to the ICU and all the problems of being on the ventilator. But on the other hand, you would feel terrible if you waited too late and the person got sicker and more agonally ill and so forth.

And then a few decades ago, there was a landmark change in practice by showing that if you could use non-invasive ventilation, putting like an emergency CPAP mask on someone's face, that could be like a hybrid solution. It could help the person get over the acute episode, but maybe avoid going on to being intubated. And that was first done in acute exacerbations of COPD.

And that became, oh, well, we should all do non-invasive ventilation. The problem is when you've done it once, you've done it once. People don't tolerate the mask well. There can be respiratory therapists who are more expert versus less expert. To be quite frank, it's a guzzle getting the mask on. And then when it's on, it's often uncomfortable and so forth.

And so people said for ages, hey, don't always intubate these patients. Put them on non-invasive first. And then other people would say, yeah, that's like hard to do.

That all changed a few years ago with the invention of high flow nasal oxygen. So this looks like you're just putting someone on oxygen, except it's much higher flow, much higher than traditionally given. And it's such high flow that not only does it provide superior oxygenation, but it's almost like providing ventilation. There's so much flow going in, you actually get positive airway pressure and so on. And so it ends up

being another mechanism a little bit like non-invasive ventilation. And it's much simpler to put on. It's just like putting on nasal prongs. So we started to see some small studies comparing non-invasive ventilation to high flow or high flow to just regular oxygen, et cetera, different combinations.

And we've also seen it in acute exacerbations of COPD, but other causes of acute respiratory failure, like heart failure. It's a little hard to keep track of when would you do one versus the other, and the studies have been small. So that's a lot of background.

But now we have the RENOVATE trial. So this large trial, it's arguably the largest trial in this space so far, run by colleagues in Brazil from a very well-known and mature clinical trials network.

And they enrolled about 1,800 patients or so all across multiple centers in Brazil, in the ED, in the ICU, or on the floor. And it was a sort of all-comer. If you think the patient is in acute respiratory failure, not so agonally ill that they must be immediately intubated. But someone in whom you're highly worried, they're acidotic, they're tachypneic, et cetera, then they were randomized.

to high flow or to non-invasive ventilation. And they had this interesting Bayesian design where they were going to try to look for different subgroup effects and try to work out would it be true or it was designed sort of as a non-inferiority. Is it adequately safe to give high flow or would you still need to do non-invasive?

The bottom line is, broadly speaking, high flow did pretty well. It broadly looks like it is as about as effective as non-invasive.

The problem is, even though they had a very sophisticated Bayesian model, they're probably still a little thin in sample size on some key groups like immunocompromised patients. Is it truly safe? It looks like it could be, but some may still have thoughts about this. But overall, the use of high flow

was practical and largely non-inferior across this broad swath of patients, which was an incredibly reassuring finding and has broad applications because high flow is quite deployable in a large number of settings.

Wonderful. So this sounds like an important trial for this field. It's also complex in its design and in the different patient subgroups and what it means for the different patient subgroups, right? So there's a lot to unpack here, but I like the way that you've summarized the main finding for clinical practice. We have two editorials that accompany this trial, one by a group who are really focused on the perspective of the emergency department and one really focused from the ICU perspective. Right.

What types of issues are they exploring in those editorials? They were very grateful to see this large study. They would probably echo that, broadly speaking, high flow appeared non-inferior, but there are still going to be residual worries about key subgroups in whom people will hold support.

strong existing preferences that sometimes it's the safer or more prudent thing to do non-invasive and they may feel that although the study is the largest to date it hasn't necessarily provided definitive evidence on some of these small but important subjects.

Excellent. Excellent. Okay. And we're in the home stretch. This is the final trial. This trial is titled biomarker guided antibiotic duration for hospitalized patients with suspected sepsis, the ADAPT sepsis randomized clinical trial.

So I'll take this one. We're saving the best for last. And so this is a nice multi-center randomized trial that's set in the NHS in the UK. And they were asking the question of how can we minimize unnecessary antibiotic use?

amongst patients with serious infection or sepsis. And in cooperation with antimicrobial stewardship, as well as clinical laboratories at many of the hospitals, the investigators designed a trial in which they were testing whether guiding decisions to stop antibiotics by procalcitonin, CRP, or just standard clinical characteristics was more effective at reducing antibiotics at day 28.

But the investigators also wanted to make sure that this wasn't leading to undue harm. And so they measured safety at the same time, and in this case, used a non-inferior design on all-cause mortality. So this was a large study. It was over 40 centers and over 2,000 patients

And in this three-arm trial, ultimately, they ended up finding that the procalcitonin-guided protocol, when giving daily advice to the clinical team, was superior in terms of reducing antibiotic duration. And on average, it was about one day less of antibiotics in the PCT arm compared to standard care.

And this was also non-inferior on the mortality endpoint. And so the investigators were concluding PCT shortened duration, but did so safely. But interestingly, CRP, or C-reactive protein, which is probably a less expensive biomarker to use, did not have a significant change in antibiotic duration and nearly missed on non-inferiority as well at mortality.

And so we're left now considering whether procalcitonin is really the test to measure almost every day in these patients. There have been other trials in the past that were inconclusive. Derek, of course, helped run one of those very large trials. We're certainly excited about this result, and it opens up a lot of questions about next steps.

So an intriguing finding here with the reduction in duration that seems to happen safely and for procalcitonin, but not the same finding for the CRP-guided approach to determining antibiotic duration.

So these have been a really interesting set of trials and a great discussion. Thank you for joining me here for this podcast, Derek and Chris. We hope for our listeners these articles inform your care of critically ill patients and give you a sense of the latest updates from the medical literature. Thank you for listening. You can find the links to these articles and their corresponding editorials in this episode's description.

To follow this and other JAMA Network podcasts, please visit us online at jamanetworkaudio.com or search for JAMA Network wherever you find your podcasts. I'm Dr. Kirsten Bibbins-Domingo. This episode was produced by Matthew O'Connor, Shelley Steffens, and Dr. Krista Roberts at the JAMA Network. Thanks for listening. This content is protected by copyright by the American Medical Association with all rights reserved, including those for text and data mining, AI training, and similar technologies.