Good morning, I'm Jeff Sorenson from the Synozervic Clinic, University of Tennessee. Disclaimer, I'm not an expert on this, but I'm here anyway because we wrote a meta-analysis about our experience, triggered by one of our experiences. So, you know, sometimes you'll have a trauma patient come in and there's not much going on. This is an example of one, got a little subarachnoid hemorrhage, a little epidural, maybe a subdural hemorrhage here. Not a lot, but you put them on DVT prophylaxis because they're polytrauma, maybe within 24 hours, and here we are three days in, we've got massive hemorrhage, you know, intercerebral, some subdural, midline shift, requires OR, and you think to yourself, wow, was that going to happen anyway, or was that the Lovenox that did that, you know, it's just an idiosyncratic reaction. We had another patient that, the one that actually triggered our meta-analysis, we did a routine craniotomy for like a convexity meningioma, everything went perfectly, completely dry at the end of the case, Lovenox, 24 hours in, ambulating, post-op day three, suddenly his scalp is completely swollen, he has an epidural and subdural hematoma, you go in there, it's oozing from everywhere, and you think, well, what happened, it was completely dry when we closed, was that the Lovenox, is this an idiosyncratic reaction to the Lovenox, so I told my resident, look, we need to review this, you know, so we initially wanted to find out, you know, does it increase intracranial hemorrhage, so we looked at all the RCTs, and I'll show you the results of what we found. So again, VTE in neurosurgery is a serious thing, it can be up to 34% of neurosurgical patients that have a VTE, 16% sort of an average that you'll find in the literature, or what we found, in 1994, they found that maybe 0-5% in that range will have a pulmonary embolus after surgery, in the neurosurgical population, with a 9-50% mortality, so when a PE occurs, it can be heartbreaking, patient's doing great, they have a PE, now they're dead, and that gets your attention, and so rightfully so, we're very nervous about PE post-op in neurosurgery, as we should be. In the 80s, autopsy, we found that, you know, a lot of people, you know, the report showed PEs in as high as 25% of neurosurgical patients who underwent an autopsy, with over half of those being the cause of death, of course this is a selected population, these are patients that were sent for autopsy, presumably because they died unexpectedly, and PE was a high cause of unexpected death after neurosurgical operations. So some of the risk factors, length of surgery, as we know, delayed ambulation, lengthy hospital stays, comorbidities, as you've heard, this is sort of a graph of all the different factors that go into it, as you've heard before. Mechanical prophylaxis has a modest effect, it helps compared to placebo, but we have chemoprophylaxis now, and so what we wanted to do with this study is look at comparing patients who are receiving chemoprophylaxis versus nothing or placebo, and do a systematic review and meta-analysis. So we used the PRISMA methodology, which many of you may be familiar with, we defined sort of what we're going after, we're going after patients who have neurosurgical procedures, had an intervention which is a chemical prophylaxis for DVT, co-intervention SCDs or not, and we want to look at rates of bleeding and venous thromboembolism. So are these rates different in patients that are receiving chemoprophylaxis versus mechanical prophylaxis or placebo alone? So, going through all the studies, we wanted the study to include a group of adult patients that received chemical prophylaxis, and another group that was treated with placebo or mechanical prophylaxis. And the use of chemoprophylaxis was the main difference between these two groups, and of course you need to report the number of events and the number of patients so we can do the meta-analysis. So two of my residents screened all of these articles as per the protocol. Nearly 4,000 articles came up, and this gradually gets weeded out looking at duplicates and excluding articles that are not matching your criteria. We came up with, it should be 9 there, 9 articles that were pertinent to our question. And we did the meta-analysis. So here we're looking first at DVT, and as you see the box and whisker plot here, as you are familiar with, the size of the box is sort of the weight of your study. How many events did you have? The whiskers are the confidence interval, and you can see there are some studies with very few events and very wide whiskers, and if it crosses over the line here, 1, with an odds ratio of 1, then there's no, as you know, no statistically significant effect. So here you see a couple of studies with a large number of events, right? 33 events, 22 events out of 160 or 130. Then here we have a study with only one event in 40 patients. So first thing that struck us is there's a wide variety of events among these different studies. And, you know, what explains that? You know, well, these patients here, the 1 out of 40, these were elective instrumented spine cases, largely in this series, whereas this series included a far larger percentage of cranial patients. But the other thing that you'll notice when you look at these RCTs is that they vary tremendously among screening. So postoperatively, some patients had daily ultrasounds in some studies. These had daily ultrasounds in venography. Some of them just did one ultrasound before ambulation. Some people did labeled fibrinogen. So there's a lot of heterogeneity in detection and, you know, could it be that the lower number of events, DVTs, in some of these studies are related to patient population, subpopulation? Or could it also be that they just went undetected because your screening wasn't as vigorous? We don't know the answer to that. But the results of the meta-analysis for DVT is that there's a statistically significant advantage to using chemoprophylaxis versus mechanical prophylaxis alone in neurosurgical populations. And the rate of DVT in the treatment group was 12.6% with chemoprophylaxis and 21.5% in the control group. So this gave an odds ratio of .51. A number needed to treat compared with placebo was 11 patients. So 11 patients treated with chemoprophylaxis to prevent HDVT. Which is interesting because in the military it's 22. But in that specific population of penetrating brain injury, it's massive. Although the incidence is higher. About 15% in that particular group. The other question that this begs is, does screening help prevent, does screening help us actually prevent these PEDs from occurring? That's a real good question. At our institution we don't do much screening and we don't have a rigorous protocol in place. But you'd like to think that if you could somehow do a daily ultrasound and find the DVT before it was a PE that you'd be able to do something about it, right? But that hasn't been shown, right? That's part of the problem that we've seen in the literature is that we haven't correlated screening with improved detection with improved outcomes. Yeah, so that's kind of disappointing. So major intracranial hemorrhage, this is sort of what we were interested in because we had a couple of disturbing cases that had unexplained hemorrhage after initiating chemoprophylaxis. And again, you see some variety in where they are favoring control versus treatment. Here's one that was favoring control. They had one patient, one group had five out of 46 with a major intracranial hemorrhage. It turns out they were giving this Lovenox preoperatively. And so they found an increase in major intracranial hemorrhage in their conclusions, although their whiskers are crossing the line. If you do the meta-analysis and look at all comers, all studies, you know, our diamond is crossing the line of non-significance. And so we can't statistically conclude from the available neurosurgical literature in this meta-analysis that there's a statistically significant increase in major intracranial hemorrhage, although there seems to be a little bit of a trend. The difference in the treatment group was 2.7% in major intracranial hemorrhage versus 1.6% in the control group. So maybe there's something there, but we can't resolve it with our current statistics. Major extracranial hemorrhage, again, the difference is 0.6% in the treatment group versus 0.8% in the control group. So really a wash and, you know, the diamond sits squarely in the middle of the line of non-significance. So extracranial hemorrhage, no significant difference, no trend even really. Minor bleeding complications, sort of similar, 3.6% in the treatment group, 2.6 in the control group. And again, not significant, although you see maybe a little bit of a trend. Our funnel plot looked reasonably okay. We had a, uh-oh, what happened? We don't think there's significant bias. We did a sensitivity analysis, so basically you take one of the studies out and do a whole new meta-analysis, and you do that for each study. So you perform K-1 studies by removing one study at a time, and that did not change the overall conclusions and results of the meta-analysis. So again, we had prospective randomized trials, moderate to good quality of evidence, which was rated as a 2 on the level of evidence here. That's pretty good quality studies that went into it. The problem is that there's a wide variety of, you know, what type of chemoprophylaxis were they getting, when were they getting it, the timing, the dose, the methods for detecting VTE varied a lot, and certainly sub- populations probably matter, which we couldn't really resolve with this meta-analysis. Based on level 2 quality evidence, again, chemoprophylaxis does seem to be effective in preventing VTE in neurosurgical patients with an NNT of 11. We couldn't find significance in increased bleeding events, but we do think there should be further study looking at populations, sub-populations, and maybe screening strategies, and as you've heard in previous talks, it looks like some of these studies are going on. It sounds like you have an institutional registry going now, and maybe a national registry ultimately might help us resolve some of this a little bit better, but I think there's still unanswered questions. It does help prevent DVTs and PEs, but can we, with better specificity, decide who needs what, when, and for how long, and who may be at risk for an idiosyncratic reaction, if that even exists? It seems like it does, and in our experience, a few patients do seem to, for no apparent reason other than the chemoprophylaxis developed, bleeding diathesis. All right, that's all I have. I appreciate your attention. Thank you.

neurosurgical patients

chemoprophylaxis

deep vein thrombosis

mortality

meta-analysis