Assessment of Risk for VTE and Risk of Bleeding

Deepak Bhatt, MD, MPH: This whole idea of informatics I think is an important one, don’t you think, Alex?

Alex C. Spyropoulos, MD, FACP, FCCP, FRCPC: I really take my hat off to Greg and your group for paving the way for how we should look at informatics, both being embedded within the EHR [electronic health record], but also the use of the electronic alerts, at admission, but also there’s recent work at discharge as well.

In our system actually we’re doing exactly that, Greg. What we’ve done is we’ve implemented and embedded the IMPROVE tool, not just for VTE [venous thromboembolism] risk, but for bleed risk. The tool automatically is automated and has a logic process built in. The score is automated with some small exceptions. And then the hope is ultimately that—and we’re in the process of building this throughout the entire Northwell Health system—is that it spits out a score for VTE risk, spits out a score for bleed risk, automatically calculates VTE and bleed risk in terms of net clinical benefit, and then is tied to an order entry set that states, your patient is at so-and-so risk, these are our recommended prophylactic strategies.

It could be an anticoagulant strategy, or if they’re high bleed risk, it could be a mechanical strategy. But the most important thing in my view, and I think Greg had hinted to that, is that now this is also being done at discharge. All of a sudden we have the same process being done after 3 or 4 days, which is the crucial point, in my point of view. The same process is alerting the physician that your patient is still at risk. Are you willing to consider extended prophylaxis? I really think this is where the direct oral anticoagulants have a huge advantage. Now patients, in my view, from a patient-centered approach have real options.

Before we just had injectables. Now we say, take a pill every day, and once a day for 30 days. I just want to emphasize the incidence that we’re talking about is a 30-day incidence. I think Mike, I’ve seen a seminar that you’ve talked about this, where we multiply this over let’s say a 12-month period, all of a sudden these incidences start looking like high-risk cardiovascular patients. This is only a 30-day incidence.

C. Michael Gibson, MS, MD: We did see continued divergence, the curves from 30 days on out beyond that, say 90 days. I think we underestimate what I call the legacy effect, where if you reduce the amount of clot there, well that’s the gift that keeps giving after you stop the drug as well. Again, the cardiologist side of me says, “Why aren’t we treating these people longer?” We’ve done some trials, 30, 35, 42 days. I’m thinking we really need to consider 60, maybe 90 days, once we get the buy-in that this is a good thing.

Gregory Piazza, MD, MS: We may actually be stopping the process of developing thrombosis at an early stage and that feeds in. There’s now some thought in the surgical VTE space that a lot of these patients are actually forming clots during their operation in the OR [operating room]. I’ve heard some very compelling hypotheses that medical patients actually are forming their clots before they come to the hospital. We’re treating patients after they leave and during the hospital, but they may actually be forming these due to their medical illness at home before they’re admitted. We have to rethink where the pathophysiology of thrombosis is really spending most of its time, because it may be starting where we’re not recognizing it.

Deepak Bhatt, MD, MPH: It’s amazing that such a fundamental thing we don’t really know, that we’re still learning what seems like the basics.

C. Michael Gibson, MS, MD: Illness is probably a systemic disease, and we’re not treating it as a systemic disease, mostly inflammation.

Deepak Bhatt, MD, MPH: Elevated markers of inflammation, elevated markers of thrombosis. You’re right.

Alex C. Spyropoulos, MD, FACP, FCCP, FRCPC: When you look at the accrual of events in the MARINER trial, because we actually then collected events even after 45 days, another 30 days after that, so up to 75 days, the VTE events continue to accrue in the placebo arm for 60 days after patients were discharged from the hospital. Around day 60 or so, there was starting to be a plateau effect of the number of accrued VTE events. I think those are very good points that the patients seem to be at risk for much longer than even 30 days, and maybe in the future we may consider, especially Gary to your point, if our options are safe, to even extend this beyond 30 days. At the very least these patients should be getting some type of post-discharge prophylaxis, and in a subset of these patients they should be getting extended post-discharge prophylaxis.

C. Michael Gibson, MS, MD: I’d also like to say that machine earning can be helpful in saying, “Well you know, you made it to 35 to 42 days. You didn’t bleed, so you’ve passed your bleeding stress test, you’re at a different risk moving forward.” We need to iteratively assess risk, not just at baseline but at different times and then we can make decisions, do we stop or do we continue? I think we’ve got to do a better job in terms of informatics.

Gary Raskob, PhD: To Mike’s point earlier, we should move away from where we’ve been in venous disease for a long time, which was focusing on venous thromboembolism; we should take more holistic patient look. That’s even further rationale for extending it, because we talked about the risk factors being the same, we’re just getting patients with markers of a thrombotic tendency. It’s no comfort to the patient to say we prevented your PE [pulmonary embolism], but you had a stroke. If we can reduce all of those thrombotic events, and anticoagulants do have impact on that. There’s a lot more work to do here, but I would see the tendency toward longer rather than shorter, and patient selection will be key, but how we do that….

Transcript edited for clarity.