Susan Wilcox, MD
Brian Fuller, MD, is an associate professor of anesthesiology and emergency medicine at Washington University School of Medicine in St. Louis and an emergency physician at Barnes Jewish Hospital. He and his colleagues recently published the LOV-ED study regarding mechanical ventilation in the ED in Annals of Emergency Medicine, and he presented their findings at the Society of Critical Care Medicine Annual Congress in January 2017.
Wilcox: Your recent study, the LOV-ED study, received quite a bit of attention at SCCM. Can you tell us about the study and what you found?
Fuller: Post-fellowship, when I started looking at how the ventilator was employed in our ED, I thought that it could have been done better. But if you want to see something done better, you need to put some data to it. We published a cohort study in Academic Emergency Medicine in 2013 to tell the initial story. The story was very little individualization of tidal volume based on gender and height, almost universal PEEP of 5 and FiO2 of 100%, and a low incidence of head-of-bed elevation. We followed that up with a four center, cross-sectional study that showed similar findings. Based on those results, as well as data from other institutions, we developed and implemented a lung protective protocol in 2014 in our ED.
The LOV-ED study was the investigation of the impact of that protocol. It was a before-after study on 1,192 patients ventilated from 2009 to 2013 (before group). We then implemented the protocol and had a 6 month run-in phase to educate providers and build collaborations. Then we prospectively studied 513 patients after the protocol was implemented. Our primary outcome was a composite outcome of ARDS and ventilator-associated conditions (VACs). The protocol focused on four simple elements identified from our prior studies that needed improvement - tidal volume, PEEP, FiO2, and head of bed elevation. So, it’s not really rocket-science type stuff - just bringing good, simple critical care down to the bedside in the ED with the ventilator.
With respect to our results, we saw a big change in how the ventilator was used in the ED, with excellent protocol adherence. The intervention was associated with a reduction in ARDS and VACs, and we had an increase in ventilator-free days and a reduction in mortality, as well.
Wilcox: Historically, management of mechanical ventilation has not been in the scope of EM practice, with many deferring to intensivists or respiratory therapists (RTs). Also, some have argued that patients aren't in the ED long enough for ventilation to matter. Can you elaborate as to why emergency physicians should care about mechanical ventilation?
Fuller: I think that was one of the things we had to overcome – to tell the story and get people to care about this within emergency medicine. When I was an emergency medicine resident, I can’t recall discussing the ventilator. It wasn’t until my critical care fellowship when I realized I had a lot to learn. But it’s not a historical fault with emergency medicine that the ventilator has not been a focus of training, because there has been a complete lack of data from the ED domain. I don’t think you can expect somebody to really care about something until you can provide them with data that says it should be important to them.
There are a couple of important reasons why emergency physicians should care about the ventilator. One is that it doesn’t really take a long time for ventilator associated lung injury to start. This has been shown with animal data and clinical studies alike. Even if you provide lung protection for a relatively brief period, there is a subsequent benefit. If you practice in an ED where wait times for a bed are long, then the time spent in the ED, in and of itself, can serve to initiate or mitigate ventilator-associated lung injury, depending on how the ventilator is set.
Another reason emergency physicians should care is because how you set the ventilator initially carries over to subsequent settings. Sometimes people are surprised by that. As intensivists, we like to pride ourselves on mastery of the ventilator. But if you look at study after study, adherence to lung protective ventilation in ARDS, a condition with an absolute indication for it, is poor. This is true, even in those centers that participated as ARDSnet sites. So I don’t think we are as good as we think we are in the ICU. If you set the ventilator well initially, it’s going to actively help patients in the ED. It’s also going to provide some passive help, as well, by influencing their chance of ever receiving lung protection after ICU admission.
Wilcox: Regarding the effect of ED tidal volumes on ICU tidal volumes, I found that relationship to be one of the most interesting aspects of the study. The ED intervention was the only significant predictor of the receipt of ICU lung protective ventilation. Was this a finding you expected?
Fuller: We’ve published several studies that show that how the ventilator is set initially, both pre-hospital or in the ED, carries downstream effects to those settings in the ICU. There is definitely “ventilator inertia”. So, even if you are in an ED that functions very smoothly, and ventilated patients are out of your ED very quickly, those initial settings are going to carry well beyond their departure from the ED. We sort of rolled the dice on the intervention by just intervening in the ED. It made operationalizing the intervention much easier, but there was a chance that the intervention’s effect would be diluted if there was no carry-over into the ICU. However, we were reasonably assured that there would be that carry-over effect once they went to the ICU.
Wilcox: Were there any barriers or challenges in implementing your ventilation protocol? What advice would you have for other EDs interested in implementing a similar protocol?
Fuller: We were really fortunate that the emergency medicine team was on board with it, and I was very appreciative of that. I think the biggest key for us was targeting the RTs in the ED. We have a respiratory therapist stationed in the critical care section in the ED, and we had a great working relationship with them before the study. They were really excited and motivated to implement the protocol. I think the biggest advice I could give to other EDs would be to first educate the physicians on how important the topic is. Then educate, motivate, and collaborate with the RTs to get it implemented. Like with anything else you would want to implement and change, relationships matter. If you can bridge those relationships, it makes things so much easier. And then another thing that sounds silly, but was a really important thing for us, is having a tape measure next to every ventilator to get an accurate height.
Wilcox: Although the findings for patients who develop ARDS are compelling, naysayers may point out that a small minority of patients ventilated in the ED go on to develop ARDS. Would you therefore recommend this lung protective approach for all patients?
Fuller: I definitely would not recommend lung protection for all patients. To take this question on a broader scope, I wouldn’t recommend really any treatment for all patients, regardless of what that treatment is. I think that if you rigidly adhere to protocols, then there will be regression to the mean, and some patients will be harmed. If you rigidly adhere to evidence based medicine, and do not account for the fact that evidence might not apply to this individual patient, then you are going to make a lot of clinical decisions without common sense. And then obviously on the other end of the spectrum, if you make all your decisions based on your experience or anecdotes, using no evidence or no protocols, then you’re probably going to harm some patients as well. I think you have to balance things. A good protocol, regardless of what it is, is one that reduces unnecessary – and I always emphasize that word – unnecessary practice variability.
So, in comparing this topic to early sepsis care, for example, if you practiced at a hospital that implemented a sepsis protocol or guideline for how sepsis should be detected and resuscitated in your ED, multiple studies have shown that this reduced mortality. I think it is highly likely that it did so because these protocols reduce the unnecessary practice variability around early sepsis care. And then if you practice at a hospital that never did anything like that, and then continued with wild-type sepsis care, then you probably saw a great degree of unnecessary heterogeneity in how sepsis was addressed.
I think you can apply the same thing to early ventilator care. There really shouldn’t be that much variability in how I do it and the next person does it, initially. I think that’s what this study shows. If you reduce the unnecessary practice variability, that’s probably going to help people.
But certainly, there are patients who shouldn’t have this applied. Patients with severe expiratory flow limitations, in status asthmatics, decompensated COPD - they are probably not going to tolerate it well and have air hunger and auto-PEEP. Patients with severe metabolic acidosis probably should not have their initial settings as lung protection.
You have to strike a balance between the preventative intervention and the disease or syndrome you are trying to prevent. If the intervention is really expensive or complex and you are trying to prevent something that is really rare or not that harmful, then what’s the point of doing it? I think in our scenario, lung protective ventilation is dirt-cheap, it’s tolerated well by the majority of patients, and I think that what we are trying to prevent is really a pretty big deal. ARDS and VACs are associated with increased mortality and morbidity. But, strictly speaking, maybe ARDS isn’t the best patient-centered outcome.
This is an area that would benefit from long-term, ED based, longitudinal cohort studies. That is an untapped research area for sure.
Wilcox: What do you see as the next steps with this work? Is there any role for further delineating tidal volumes, say 6 ml/kg vs 8 ml/kg of predicted body weight?
Fuller: Like a lot of single center studies, if you did what we did at Barnes-Jewish Hospital and expanded that to 20 different centers, I think you would see similar results to what we saw, but a smaller effect size - sort of classic for single center investigations. But for six versus eight, it would take a lot of patients if that were the only two variables between the groups.
Lung protective ventilation is an area that is ripe for implementation research - this is a much more important question than another tidal volume trial in my mind. There is so much literature out there to support doing good lung protective ventilation for everyone with ARDS, as well as those at risk for it. But I think what we miss in critical care research is implementation. There are the things we know work, simple things, but we continue to implement them poorly.
Return to Newsletter