And what we're going to talk about now for the next probably 15 minutes is some basic information on how a guideline is developed, how you should look at it, and then how you should weigh a couple of recent critical care guidelines. None of this stuff will be on board exams. I hope that this will help you with your practice and with reviewing guidelines. And if you think this is totally useless, let us know, and we'll never do this again. So this is new for this year. It's a trial. We just wanted to see it. And I've given you my conflicts in that I have been on guideline committees. I'm the incoming editor-in-chief for Critical Care Medicine, so I will be publishing guidelines. And I've been part of the Board of Regents, which creates guidelines for SCCM. So we started out this conference with the PICO questions. I promised you you'd see this again. When you're looking at a guideline, you want to think about who does it involve, what population are you thinking about, what intervention are you doing, what you're comparing the intervention to, and what outcome you and hopefully the patient care about. And I'm going to give a couple of highlights about how to think about this. The other thing that I want to talk about is, what do you do? Every guideline has a process. You get a group of people who you hope know something about the content area. You get a group of methodologists who know something about evaluating systematic reviews and meta-analyses, a couple of librarians who dig out all the data, and then you follow a clear process to come up with either a recommendation, a suggestion, or a good practice statement. Guidelines are only as good as the evidence behind them. And I've had my name on guidelines that recommended things that later we found out to be harmful. So, for a number of years, we recommended giving tight glycemic control for patients with sepsis, which was based on a single center trial that was later proven not to be true. So, there are changes in guidelines based on new evidence. You'd like to think that people change their practice based on new evidence. The other thing that I think is reasonable to think about is that guidelines are great, maybe, but if following a guideline doesn't help a patient, if implementing the guideline in your hospital or somebody else's hospital doesn't improve care, then a guideline may not be all that helpful. It does eliminate some variation in care, and the analogy I usually give is that if you go to a restaurant and have a really good meal on Tuesday night, and there's a different chef in the house on Saturday, and they cook the dish a little bit differently, and your friend who goes there, because you said, I had a great meal there, goes there, and somebody prepares it differently, and maybe the waiter is a little bit different and doesn't do as good a job, that wouldn't be a great thing. You would think that if somebody comes to the hospital with pneumonia and respiratory failure on Tuesday, and I'm on service, and I give one treatment, and then my partner is on on Saturday, and they give a different treatment, that may not be the best thing for many patients. So one thing a guideline can do is to minimize unnecessary variation. Now obviously patients are different, right? You have to treat them all appropriately, but there's, at least in my own institution, your institutions, you may have everything down spot, but in mine, sometimes people get treated differently at different times of the night or day, and when different clinicians are on. So in theory, if you have a good guideline with a strong recommendation, that most people ought to give that to most patients. And usually, strong recommendations are based on randomized controlled trials in patients that look like the patients that you care for. Weak recommendations are based on less good evidence, and good practice statements or best practice statements are based on a lot of people sitting in a room saying, we all do this. And you'd like to think that that probably is right, but it's not based on any more evidence than a bunch of people saying, this makes sense. And a lot of stuff that we do, source control in septus, for example, are based on best practice statements. If anybody's really interested in this stuff, you can read this. The guide group that started helping people think better about guidelines has published a lot of stuff, and here, again, on your slides, you're welcome to review this. I'm not going to read through any of this, because my guess is it will put some of you to sleep. There are a lot of best practice statements in the guidelines that I'm going to show you moving forward. And so recently, the rapid sequence intubation in the critically ill patient's guideline was published. There are a lot of new studies that came out since this was published. And that's, again, a problem. We don't yet have living guidelines where the day after somebody publishes a new trial, somebody changes the guidelines. We can't do that that quickly. And that's part of the reason that this morning, when you heard that the guidelines on prevention of stress ulcers came out, and a couple weeks later, some of the people on them, some of the people on that guideline published the revised study in about 5,000 patients in which, looking at intubated patients, they showed that giving PPIs to intubated patients lowered stress ulcer and didn't change mortality and didn't change ventilator-associated pneumonia or C. diff or anything else. And so the guidelines that were published a month later already were out of date. And there's nothing you can do about that other than be ready to change your practice when new information comes out. So the intubation guidelines suggest that you give a neuromuscular blocker when you intubate. Can I ask, just out of curiosity, are there people here who prefer not to paralyze patients when they intubate by a show of hands? Does anybody do that? No? Okay. So you didn't need to have this. This guideline isn't going to change this group's practice. And even though there's a lot of concern about atomidate and possible adrenal suppression, Steve may touch on this when he talks about the new steroid guidelines. But they had no recommendation to make about that. The preoxy study that I mentioned back on Wednesday came out after this. And so they suggest that you preoxygenate either with heated high-flow nasal cannula or non-invasive ventilation prior to intubation. Again, the preoxy study didn't compare those two. We need another study that will compare them. But either of them are reasonable to do. Steve, you're going to talk about the fever guidelines. So I'm not going to talk about that. And here's another guideline that talks about failure to rescue and recognizing clinical deterioration outside the ICU. And the reason I want to put this particular no recommendation up is that there is absolutely no evidence to spend lots of money using a MUSE system or any other expensive system to monitor all patients in the hospital. You could do it if you have lots of time and money and want to do that. But the experts that looked at this did not have any recommendation on this, which suggests that either the 10 or 15 studies that have looked at that weren't large enough to show an effect size or it doesn't work. And those are the two options. And there are a lot of suggestions here that you should have a rapid response team or a medical emergency team and that you should talk about end-of-life issues and that families should be part of this so they can call if they're worried about a patient. But these are all suggestions. And over the next couple of years, there may be better evidence that will change these. Nobody has any idea what to do, how precisely to set glucose in non-neurologic patients in the hospital. And a strong recommendation went down to you probably should not give tight glycemic control for most patients. And you probably should aim for somewhere between 140 and 200. And so if anybody has a burning desire to do a trial that will answer these, we may have better guidelines. And having decision support tools to help people think about this is probably better than not doing so. So I mentioned that we have new guidelines on prevention of stress ulcer prophylaxis. And these are, again, they're already slightly outdated. But the high-risk patients here were coagulopathy shock and chronic liver disease, which is different than the patient population that Deb Cook's New England Journal article that was published two months ago looked at, which were intubated patients. It's reasonable to feed people. Feeding people is good. And in low-risk patients, it's probably enough. In high-risk patients, you probably should give some flavor of prophylaxis. And which prophylaxis you give probably depends upon side effects, cost, and your own particular institution. And that is my 10-minute soapbox speech about guidelines. Again, if this is totally useless, let us know. If it added value, let us know. And Steve, who was on two of the guidelines that were just published, is now going to talk about those two guidelines.

clinical guidelines

systematic reviews

randomized controlled trials

evidence updates

care variation