PlayPlay Podcast downloadDownload Podcast Take Post TestTake Post Test

Transcription:


Welcome to episode one of Evidence-Based Medicine Breakdown. I am Matthew Delaney. I'm Charles Khoury, and we are both Emergency Physicians at the University of Alabama at Birmingham. Look, I've listened to plenty of shows that talk about what's in the literature. One of the things that drive me absolutely crazy, is the hosts will go through this elaborate paper. It seems like it's a practice changer. It's from a big journal and then their conclusion is something along the lines of, yeah, we need future studies to guide us further, or this really isn't ready for prime time. I understand that, that's likely an accurate statement, but when I'm at the bedside with a patient, thinking about future studies is really not helpful.

It gets me nowhere. On this show, what we're going to try to do is shine a light on studies that we can go through, digest and then take to the bedside the very next shift when we're seeing patients. Part of this is each month, we're going to have some special segments. Every month, we're going to get one hot take, and this is going to be a paper that for one reason or another reason, really got my blood pressure up. And the second segment, that's going to be recurring every month, is called the pitch. This is going to be a paper that not only gives us a clear clinical takeaway, but is strong enough, it hits hard enough, that this should actually be a game changer. This is something that we should take and change our current practice or at a minimum, I'm going to pitch you on why I think that this paper is that impactful.

Charles Khoury, MD: To kick things off, we're going to look at two papers that have come out recently involving IV fluid. Now I'll be very honest, typically the IV fluid that my patient gets is chosen based on what's available at the bedside and what the nurse happens to grab. I've always operated under this assumption that there's not much of a clinical difference between a balanced solution like LR versus normal saline.

And I feel like all hospitals are always struggling with some sort of random supply chain issue, that's going to limit which one you have at which time. So I've always been pretty ambivalent about which fluid we're going to give as long as we're giving something. But a lot of the critical care folks where I trained were pretty LR heavy, and they would cite studies that said that normal saline exacerbates acidosis, and that there was this possible risk of worsening hyperkalemia, in a patient who is already hyperkalemic.

And I also remember being taught that even though normal saline is cheaper, that LR was only like 25 cents more. So it couldn't have been that big a deal. So I've gone back and forth on the fluids I use, but it's been largely influenced by guilt and shame, and by not understanding the actual literature. I know that there's been a lot of recent debate about whether we should use balanced solution as opposed to normal saline.

But in the past few years, there've been some observational studies and two unblinded single center studies in the US that suggested that giving balanced crystalloids resulted in better patient outcomes. However, the results were pretty mixed and we needed larger randomized multicenter studies. So along comes this first paper.

Dr. Delaney: Yeah, I'm glad we have this paper because I've always found this debate to be very nerdy, very irritating, and it just doesn't seem like it should really matter. But to help us answer the question we've got two papers by the same author group, Zane Pierre et al. The first one is Effective Intravenous Fluid Treatment With a Balanced Solution Versus 0.9% Saline Solution nr Mortality in Critically Ill Patients.

This is the BASICS randomized clinical trial in JAMA 2021. And this was a double blind factorial randomized clinical trial conducted across 75 ICUs in Brazil. And the factorial design means that they looked at does the fluid type matter? And that's this paper. And then they also kind of two birds, one stoned it and said we're going to also look at does different infusion speeds during a fluid challenge matter.

We'll hit that paper next. So you're in an ICU in Brazil, you were randomized to get plasma light 148, which is the balanced solution. That's like lactated ringers. I know it's not exactly the same, but it's close enough. Or you've got normal saline. It's a big study. So 10,520 patients were available. If you look at who are these people in the Brazilian ICU's, they were not exactly what we'd see in the emergency department.

So 48.4% of the patients were planned surgical admits, 68% of all patients got a fluid bolus before being admitted to the ICU and greater than 45% got over one liter. So these patients were in the midst of a resuscitation and they were sick, so 60.6% of patients were hypotensive and or needed vasopressors ,44.3% required mechanical ventilation at the time they were put into the study.

And if you look at both groups, so they come into the ICU, they're sick, they're already getting resuscitated. They're randomized to get one of two fluid types. And the overall fluid they got during that first day, after being enrolled in the study was about 1.5 liters in both groups. So some fluid. They're not drowning these patients.

And if you look at the accumulated mean fluid administration, so basically how much fluid did you get before you got to the ICU? And then once we put you in this study, how much fluid did you get? After the third day, following their enrollment is about 4.1 liters in both groups. So these, these are patients who are sick who are getting a decent amount of fluid.

They looked at 90-day mortality as a primary outcome. And in the balanced solution group, the 90-day mortality was 26.4%, in the normal saline group was 27.2%. So there was not a difference between these two groups. Charles, I think this is interesting and they went and did a bunch of secondary or subgroup analyses, really trying to see is there any patient population where a balanced solution is going to be beneficial. I think these people really wanted to see that work. And as much as they really wanted to dog on normal saline, there was not a single subgroup that did better when they got a balanced solution.

Dr. Khoury: One of the really odd things from this paper was that there was a signal of possible harm for patients in the balanced solution group with a traumatic brain injury and a worse neurological SOFA component score at day seven.

So this could be due to a measurement error because we know it's suboptimal to assess GCS in sedated patients. This comes out in a secondary analysis. So I give it a little less emphasis, but I do think it is, at a minimum, hypothesis generating. And so I'm looking at these two options and it seems like there's no benefit to balanced solution and there's potentially a downside in patients with a TBI. So for me, I don't really have a strong argument about why we should use balanced solutions.

Dr. Delaney: Yeah. And again, the secondary analysis doesn't get us the full answer, but the numbers are shocking. So 90-day mortality, if you had a TBI with balanced solution, it was 31.3% compared to similar patients, and the only difference really being that they got saline solution. And if that's the case, their mortality was 21.1%. So there is a difference when you look at that subset. So I think you said it well, Charles, for me, the bottom line here is that if you look at critically ill patients who need a fluid challenge, the type of fluid whether we go with a balanced solution versus a normal saline, did not seem to impact 90-day mortality, except for in the group who had a TBI where balanced solution might actually increase mortality.

Dr. Khoury: Yeah, I totally agree. I actually want to give full credit to these authors. They run this big study and then they somehow get two publications in JAMA, out of what is essentially the same study. The first paper that we discovered, looked at the fluid choice. With the second paper that we're about to look at, actually it looks at the rate of our fluid bolus. And I don't remember the last time I thought about the rate of a fluid bolus, especially in the emergency department. Typically, I tell nurses, give it fast, but obviously this isn't precise. But when I started digging around, there's actually not much evidence-based guidance when it comes to how fast we should give a fluid bolus or a fluid challenge.

There's this theoretical risk that a faster infusion rate could rapidly expand the intravascular space. This could maybe lead to third spacing and edema, but we really have nothing to go on other than these theoretical risks.

Dr. Delaney: So the paper is same period. All same author group, same patients, same paper yet, it's got a different title and it gets a separate publication. So Effective Slower versus Faster Intravenous Fluid Bolus Rates on Mortality In Critically Ill Patients. The BASICS randomized clinical trial in JAMA 2021. Again, this was the subset of patients, sick folks, across 75 ICUs in Brazil. And you were randomized to the slow group, which was you got your fluid at 333 milliliters per hour.

The control group was 999 milliliters per hour, which I didn't know was like a precise thing. I just thought that's how you maxed it out. When you look at that same group of patients, right, these were ICU patients. They were pretty sick. They were often on vasopressers, a fair proportion of them were mechanically ventilated. When you randomized the fluid to the slow group versus the fast group and looked at 90-day mortality, it was the same, same. Slow group, 26.6% control group, 27%. There was a little bit of debate about the slow infusion group actually doing better. So they looked at day three SOFA scores and they said, yeah, maybe you were less likely to need pressers at day three in the slow group, but by day seven, the groups still were the same, same. So this is simple. This is easy. The bottom line here I think, is among critically ill patients who require a fluid challenge, the rate of administration of fluid 333 cc's per hour versus 999 cc's per hour did not result in any difference at 90-day mortality.

Dr. Khoury: All right. So at this point it seems like this podcast is just becoming a big hype page for all of Zane Pierre's et al's papers. It's like, we're almost like rick, rick-rolling you guys for, but for medical podcasts. But I do think we should probably change subjects at some point, unless you want to go through one of his earlier works.

So, you know, we've been friends, you and I Delaney, for a really long time and I have to confess something. You know, one of my favorite things to do in life is to get you fired up about something. It can be anything, but it's really, really fun to see you fired up. And if you have a close friend that you love making irrationally angry, you know what I'm talking about, you know what I'm about to do.

And for some reason your thing that gets you fired up more than anything, is TPA literature. It's kind of my equivalent of cracking an ammonia packet under your nose and just seeing what happens and not that papers on IV fluid administration aren't exciting. We've already talked about how much we love Dr. Zane Pierre, but I was really looking forward to hearing you get fired up about our next topic.

Dr. Delaney: Look, that brings us to the hot ticket of the month. I don't think there's been a paper that I've read, I put in the notes in the past 10 years. I would say I've never read a paper that makes me angrier than the one we're going to discuss. I think everybody listening will agree if you come to the emergency department and you have neurologic symptoms that are not coming from an ischemic stroke, then there is no way in the entire world that you would ever benefit from getting IV thrombolytics. Now I work in a stroke center.

I understand there's this emphasis on seeing patients who could have a stroke quickly and I'll confess that's a tricky thing to look at a patient and say, yes, this is an ischemic stroke, or no, this is not an ischemic stroke. It's actually a stroke mimic. This just looks and acts like a stroke. I personally have messed up this distinction. This is a tough call clinically. The literature tells us that all of us struggle with this distinction of stroke versus stroke mimic. The problem is, not only are we making an error in incorrectly diagnosing somebody, but if a patient comes in with a stroke mimic and we think it's an acute ischemic stroke; we all know that the gears start turning and now we're potentially exposing a patient to IV thrombolytics.

So they don't have a stroke. We're going to potentially give them medicine for stroke. So now they're going to get all the risks that comes with thrombolytics. On the surface, this paper, we're going to talk about, looks at what happens to stroke mimics in this situation where everybody's thinking stroke, we're treating stroke, but the patient ultimately doesn't have a stroke. They ultimately have a stroke mimic. The premise of this paper is very reasonable, but sit tight because this paper gets freaking insane.

Dr. Khoury: The paper today is by Ali Ahmed, et all. And it's Titled Intravenous Tissue Plasminogen Activator in Stroke Mimics. Findings from the Get with the Guidelines Stroke Registry. It was published in Circulation Cardiovascular Quality Outcomes in 2019.

Dr. Delaney: So this was a study that looked at the Get With the Guidelines Stroke Registry. So this was over 72,000 patients who came in with suspected stroke, who were then treated with IV alteplase. And this was done across 485 hospitals in the US and what they said is all these people entered in the databases we thought they had a stroke, we treated them with thrombolytics, yet they found 3.5% of the patients ended up having stroke mimics. So they were lysed, but they weren't having a stroke. That happens. You gotta crack some eggs to make an omelet. This is a hard clinical diagnosis. What happened to these people?

So we know that the risk, if you get thrombolytics is you can bleed and you can die. The risk or the rate of symptomatic intracranial hemorrhage in the stroke mimic group was 0.4% compared to in the acute ischemic stroke group of 3.5%. If you look at in-hospital mortality, if you were having an acute stroke, it was 6.2% compared to 0.8% in the stroke mimic group, those are the numbers.

And I want to start this discussion of this paper with a quote that's directly from the authors in this paper. And they say the complication rates for stroke mimics were nominal, nominal, and outcomes were favorable. And they then go on to say that, because this risk is quote, "nominal," that whatever happens to the stroke mimic patients, whether they bleed in their heads and then die, that is a quote "reasonable trade-off" for earlier treatment time of acute stroke patients.

Basically, it doesn't matter what happened to the mimics because it lets the other folks get treated quicker. Charles, I'm gonna ask you a question. If you have a migraine and you show up in the emergency department or say you have an isolated seizure, you're postictal. What is your expected mortality from that event from your migraine headache?

Dr. Khoury: It's zero, right? I mean, people don't die from migraine headaches, but we continue to have clear evidence from this. And many other studies that people with migraines are often misdiagnosed with having a stroke, when in fact they're having a stroke mimic.

Dr. Delaney: Right? So when I look at this study on first pass, I understand. We don't always nail who is a stroke mimic versus who is an acute ischemic stroke. But if you look at these numbers, if a patient comes in with a migraine, which is a common known cause of stroke mimics, and we misdiagnosed this that this is a stroke and give them IV thrombolytics; they now have a 0.4% chance of having a symptomatic bleed in their head and they have a 0.8% in-hospital mortality. These numbers are unacceptable. These should be zero. This is a patient with a migraine headache. There is not a nominal rate of this patient bleeding or dying that's acceptable. I just want to beat the drum again. I know that this is an error that we're all prone to make. Is this a migraine versus stroke?

But holy smokes, the thought of giving somebody thrombolytics and killing them with a migraine is unbelievable. It's, it's unacceptable. That's my beef. That's my issue with these authors is that they have the gall to say that this is nominal. They spend this study not to say we should really try to figure out who has this versus a stroke mimic, but actually say don't care about the mimics. What you need to focus on is treating strokes more quickly. Don't worry about the mimics. Don't worry what happens to those patients. And when you look at the database that the authors use for the study, very clearly we see as time has progressed since thrombolytics have come on the market, more and more patients each year are getting treated with IV thrombolytics. It's been a very successful marketing campaign by the makers of these medications. So what that means is each year more people are thrombolyzed, each year we are exposing more patients who have potential stroke mimics to a potentially life changing or life ending therapy. So the numbers are going up, I think, and I worry things are only going to get worse from here.

Dr. Khoury: Yeah, I think this push to see patients faster and faster in this way is potentially dangerous. We have evidence from patients with STEMI's that shows that as door to balloon times decrease, the number of false positive STEMI activations also increase, which leads to this increased in-hospital mortality with patients with false positive STEMI's.

I think that's similar to this phenomenon we're seeing here in these patients with stroke mimics, and it's almost like these authors are gaslighting us into thinking that we're the crazy ones for not being okay with overusing TPA on these potential non-strokes.

Dr. Delaney: Yeah, these, these authors, the term nominal makes me furious. You could say regrettable, you could say unfortunate. There are dead people that are being described as that's a nominal number of people that are dead. And I don't, I'll tell you this. I don't trust these authors numbers for a couple of reasons. They tell us that in this study that they ran, the rate of stroke mimics was 3.5%.

Previous studies have said it could be as low as 1.4. It could be as high as 15.5%. Those that's a wide range of numbers and the authors kind of skip past this with very little explanation. But what this means is that when we're seeing a patient, the chance that they're having a stroke mimic could actually be five times higher than these authors are reporting.

And, I don't think the authors are trying to mislead us with this. I do think that they're minimizing some of the inherent issues with these database studies and think of it this way, this is a database that people were voluntarily entered into. If a patient comes in and I give them thrombolytics and they're not having a stroke and they bleed into their brain and they die; are you going to put that in the database? Are you, is there any benefit to you if you report that case into a national database? I'm, I'm not putting my name on that. One of the big questions here is reporting bias, which means we really don't know how often stroke mimics show up. And how often are we hurting them with IV thrombolytics because we're relying on someone to say, yeah, we screwed up, let's put it into the database. I recognize that a lot of good intentioned people are doing that, but human nature has to kick in at some point. So we, I think we're probably not getting a true grasp of how often are mimics getting lysed.

And the other thing that's interesting from this database, is smaller hospitals are excluded. So if you show up with a TODD's paralysis to a small community hospital and get lysed and bleed and die, you're not even in this database. So methodologic issues here, makes me really worry that the numbers that we're seeing from these authors, that they're describing as nominal rates of death for patients with migraine headaches, actually in reality, are probably much higher than what we're seeing in this paper.

Dr. Khoury: Yeah, I want to add something else to this conversation and it is that I worry that studies like this can be misinterpreted. And then all of a sudden your hospital has implemented a new guideline that hasn't really been vetted. These large medical centers rely pretty heavily on activation pathways and power plans, which is usually fine because these things often do decrease rates of error.

But the real error is when patients are pushed onto this rollercoaster ride without being vetted initially. So the system and the process itself might be streamlined and refined, but it's really important that we don't just start letting anyone get on the roller coaster. You know, you must be this tall to ride the code stroke TPA activation pathway.

Dr. Delaney: Yeah. For the bottom line, I think I could not disagree with these authors more. They say that the rates of adverse event for stroke mimics is low. I would say that that rate should be zero. When a patient does not have any potential to benefit from the treatment, any harm I think is unacceptable. I'm not saying it can be fully avoided, but we cannot consider this to be a nominal risk. We have to take this seriously.

The next paper is great. I remember in med school being forced to memorize how chemotherapies worked and not doing well on that block of pharmacology. And I have never once needed to know anything about chemotherapy in my practice. And I take care of patients who are actually on chemotherapy. So I was taught all this stuff I don't need. Yet, if you put a gun to my head and say, what were you taught about how do you treat an ankle sprain? What is the best medication to get that patient feeling better? No one has ever talked to me about that in my entire career.

Dr. Khoury: Yeah. Great point. If I'm working with a new medical student, I've never worked with, I always sit down and go through the different classes and dosages of routine pain medications, and it's clear that it's the first time that they've ever heard this stuff. It's a really good teaching point to students and to residents. And it's something that we should probably be talking more about in medicine.

Dr. Delaney: Theis study is by B. Jore et al. It's a randomized trial comparing the efficacy of five oral dose analgesics for treatment of acute musculoskeletal extremity pain in the emergency department. This was from Annals of Emergency Medicine in 2021. This is a randomized clinical trial, out of two urban emergency departments where they looked at 600 patients that were predominantly Latino men with musculoskeletal pain, where the clinician planned to treat them with oral analgesics and get imaging of the painful extremity.

Dr. Khoury: All right. I got to stop you here just for a second. Because the paper is a little vague on how exactly they've recruited so many Latino men. And I think it's hilarious to imagine that there's a research assistant or a first-year medical student, just walking around the ER, just staring at random men in various states of pain. It's becoming kind of a slightly weird process, but anyway, keep going.

Dr. Delaney: Yeah. It's, it's not addressed in the paper. That being said, these this group of predominantly Latino men was randomized to get one of five different medications. Either 400 milligrams of ibuprofen with a gram of acetaminophen, 800 of ibuprofen with a gram of acetaminophen, 30 milligrams of codeine with 300 milligrams of acetaminophen, five milligrams of hydrocodone with 300 of acetaminophen or five milligrams of oxycodone with 325 milligrams. Or five milligrams of oxycodone with 325 milligrams of acetaminophen. And the primary outcome was change in pain before getting the medication to one hour in the numeric rating scale that zero to 10. And essentially everybody changed by about three points. Their pain got a little bit better. And statistically there was no difference between any of these medications at that hour mark, when it came to how much better was your pain. They looked at a secondary outcome, which was how often did patients need a rescue medication at one and two hours post baseline? So you got the first thing. How often did you need a second dose or a different medication? And they found that in the patients who got the combinations of ibuprofen with acetaminophen, that the rates of needing rescue medication were higher. It was about 1 to 2.5%. Whereas the patients who got five milligrams of oxycodone with 325 acetaminophen, did not need rescue medication at one or two hours.

Dr. Khoury: All right. So the codeine acetaminophen combo or the Tylenol #3, I've just, I have a comment about this. I can't stop myself here, but it's something that I've never, ever prescribed in my practice. And I don't really know anyone who's prescribing Tylenol #3s. I'm actually surprised they were able to add that one to the protocol because, you know, it seems to me that whenever kids go off to college, their parents tell them, you know, hey study really hard, make good grades. And also you're allergic to penicillin and codeine.

And so I'm not sure how or why Tylenol #3's were included. Delaney, maybe you and I should perform some kind of clinical trial on Darvocet in the near future. That also seems like a drug that could have made it into this, this study, that I think would be useful.

Dr. Delaney: Or ether. I mean really just go back a couple of decades here. So I like this study, Charles. I think it it's, these are patients. We see. Something is wrong with an extremity. It's not obvious that a bone poking through, so we're not reaching for IV analgesics. And I liked the bottom line, that no matter what you do, whether it's a combination of ibuprofen with acetaminophen or something that involves a narcotic, that the results at one hour, are about the same. You know, one of the interesting things here was that the dose of opioids they use was fairly low. So this is basically five milligrams of opioids, whether it's hydrocodone or oxycodone, and that's similar to what we use, but also if a patient has an acutely painful condition, we know that higher doses of opioids are going to get patients a little more bang for their buck when it comes to pain relief.

We know that with ibuprofens and acetominophens, we get a ceiling dose. So there's a max. So if we're looking at a patient and we're saying, I think one dose will be fine for now, it doesn't really seem to matter what we pick. But I think if we're thinking there's going to be an ongoing need for analgesics or they're in a lot of pain, that might be where a higher dose of opioid may get us a little more pain relief.

One other little quirk here is they talk about this combination of acetaminophen and ibuprofen, which is what I take when I have pain. And I tell patients to take this, but we don't have a commercially available combination pill in the US. And so what I would love is to write a prescription for this is a combo seat, and this is, this is you go to the pharmacy and you get one pill that has ibuprofen and acetaminophen, because I worry a little bit, if I just tell them, go get two over the counter agents, take them both together, a little bit of the magic, a little bit of the placebo impact we get from giving them a prescription and have them go to the pharmacy, may go away. But overall, I think this paper is encouraging and tells me that whatever I'm reaching for when a patient has acute musculoskeletal pain, probably works as well as what you're reaching for. The differences don't really seem to matter.

Dr. Khoury: Yeah. What I've learned from this is that I should probably be giving more combination therapies in the emergency department. I think, I write for either ibuprofen or Tylenol and I'm going to start writing for both at the same time.

Dr. Delaney: Charles before we shift to the pitch, I want to tell you a story about my friend Seth from college. So Seth, he's a cool guy. He was really into Lord of the Rings and in fact, he spent all of the money he'd earned working one summer to get him and his brother's name, put in the end credits Lord of the Rings.

And I saw the movie and I was like, I didn't see your name. And it turned out, you had to really watch and it scrolled for like 17 minutes. And then Seth and his brother's name kind of popped up at the end. And I thought that was the weirdest thing about Seth until one year at Halloween, we would throw a big party in our house and we're trying to get ready. And Seth kind of comes out of his room and he's like, I really, my chest hurts. I'm having trouble breathing. And we're college kids and we're like, well, I guess you better go to the doctor. So Seth walks to the hospital and we lived in Charlottesville. So he basically to walk up and down a mountain to get to the hospital and we didn't see him for a day and he came home and he's like, yeah, my lung collapsed or popped.

And they had to like stab me with this knife in my chest and get the air out. And we're like, that's really unfortunate. I'm sorry that happened to you. And a couple of weeks later, we were going to go have a date function. And Seth comes out of his room and is like, I think that thing to my lung happened again.

And we had learned that we're like, yeah, you should definitely go to the doctor. So again, Seth goes schlepping over the Hills of Charlottesville, Virginia to the hospital, and he had yet another pneumothorax or a recurrence of initially his initial pneumothorax. When he came back that second time and he was pretty frustrated.

And I think he was correct that two times he'd had a potentially life-threatening issue that none of us were willing to drive him to the hospital for. And that when he talked about being stabbed in the chest, we kind of minimized it, we're like, yeah, well, I'm glad you're better now. And as I started my practice of medicine, I really thought, gosh, there's gotta be a better way for patients like Seth, who are ambulatory, physiologically stable, who get a spontaneous pneumothorax, there's gotta be a better way than walking to the hospital, getting a large test tube, because as we know, that's a very morbid, very painful procedure for patients. And we've started to see this shift for spontaneous pneumothoraces where the were the idea is if Seth shows up and he's otherwise well appeating, can we just observe him, can we do a conservative management and avoid putting a tube in his chest?

Dr. Khoury: So today's pitch is a watch and wait approach to the treatment of pneumothoraces. The paper is by Brown et al, and it's called Conservative Versus Interventional Treatment for Spontaneous Pneumothorax. And it was published in New England Journal of Medicine in 2020.

Dr. Delaney: This is a multicenter prospective randomized non-inferiority trial at 39 hospitals in Australia and New Zealand. Patients were randomized to either get immediate interventional management. So it was the less than or equal to 12 French Seldinger technique, chest tube verse the conservative observational management.

So they watched them for greater than, or equal to four hours in the ED, got a repeat chest x-ray and if that was stable, they just send them out. So overall they recruited 316 patients and then at eight weeks they couldn't find some of them. We'll come back to that, but they end up with complete analysis of 256 of these cases.

These were Seth. These patients were Seth. So the average age was 26. The mean pneumothorax size and they get into this Collin's formula was 65%, but basically that's a decent size pneumo. You and I look at the x-ray and say, yeah, that's definitely a pneumo there. And their primary outcome was re-expansion within eight weeks. In the intervention group, you got a chest tube, it was 98.5%. And then in the conservative group, it was 94.4%. And so the way they set this up as a non-inferiority trial, basically that means the conservative approach was thought to be non-inferior, that it wasn't quite as good. The success rate wasn't quite as high as the intervention group. Yet the math holds up and that watching these patients, observing these patients is a non-inferior. It's an okay way to go. They looked at the secondary outcomes and consistently they were worse in the intervention group. If you look at pneumothorax recurrence rate, which is interesting, it was 16.8% in the folks who got a chest tube versus 8.8% in the conservative group.

The adverse events rate, serious adverse events were 12.3 in the chest tube group compared to 3.7 in the observation group. Inpatient days, this makes sense, but six days in the hospital, if you've got a chest tube versus 1.6 days in the observation group. And then hospital revisits 26.6% in the chest tube group compared to 17% in the intervention group.

Before we jump into discussion, the math here is tenuous. So the devil really is in the details here. This is, I think it's important to realize this is a very specific patient population. So, overall they looked at over 2,600 patients yet could only find 316 that they thought were appropriate to be randomized.

So there's some selection bias potentially here. So I worry a little bit about taking this study and applying it to every patient that shows up with a pneumo. Also 30% of patients were lost to follow-up. I don't think they died. I don't think something horrible happened, but that does the math starts to break down a little bit.

So, Charles, it's an interesting paper. I think they tried to do a robust study. I do worry a little bit about some of these issues with the math, with the methodology. What's your clinical take on this paper?

Dr. Khoury: Yeah, I mean, this definitely makes me think more. We do want to start looking for reasons to avoid chest tubes because they really hurt and there are complications that come along with them. So if a patient comes in and has a 25% spontaneous pneumothorax on the right, they are otherwise pretty healthy; this is someone who I would OBS. I would admit to the hospital and I would OBS and kind of see what happens. What about you?

Dr. Delaney: Yeah, I think that's the tweak. So in this study there OBSing you in the ED predominantly and then sending you out. I am not brave enough to do that yet. And I actually think that some of those kind of nerdy methodologic issues make me hesitant to do that. But I absolutely, the next time I see a spontaneous pneumo, and I saw this paper when it first came out and I've actually done this once or twice you tell the patient, look, we can either put a tube in your chest. We can stab you with a large knife like they did to Seth twice. Or we can just admit you for a night and in the morning, if you're fine, I think it's very reasonable at that point to say, you're going to go home and almost certainly going to do as well as you would, if we had put a chest tube in your chest.

Dr. Khoury: Yeah. There's this almost non quantifiable aspect of this argument. Like how many patients after receiving a chest tube, do you think you could ask hey, would you have rathered that we didn't do that to you? I mean, of course every one of them is going to say, well, yeah, absolutely. You mean, I didn't have to go through that. So it is important to kind of put yourself in the patient's shoes here and ask them what they would want to have done, if at the end of the day, the outcome could be the same.

Dr. Delaney: Yeah. And this paper, warts and all it gives us ammunition to say that approach, the I'm going to watch you. I'm going to not stab you is non-inferior and I think is something we can start doing. Again, I think with the caveat that I would be careful about how long we watch these patients. So Charles, we flew through some stuff. We hit Zane Pierre fast and furious because he's obviously our favorite. What are your five takeaways from this month?

Dr. Khoury: Yeah. So five takeaways, number one fluid choice, probably doesn't matter. Number two rate of fluid bolus, probably doesn't matter. Number three, we are at risk of killing patients if we keep giving thrombolytics to people who aren't having strokes, number four, combination analgesics all seemed to work. And then number five, do you really want a chest tube if you have a spontaneous pneumothorax?

Dr. Delaney: I like it. Five papers, five easy clean takeaways. We will be back next month. If you've got a topic, if you've got a paper, if you've got something that makes you hot, if our take on something, gets your blood pressure up, drop us a line, let us know, but we will talk to you next month.