How can we use decision support for quality improvement? What approaches can improve VTE Prophylaxis? It would be great if we could just go give talks like this or lectures to residents or physicians or nurses and just tell them what to do. Unfortunately, passive dissemination of guidelines is unlikely to improve VTE Prophylaxis practice. Just handing out cards or laminated cards or putting posters up on the wall, those strategies don't work alone to improve prevention for VTE. We need active strategies used in combination to incorporate different approaches to improve optimal outcomes. So at Johns Hopkins Hospital, we've been working on this project for literally over a decade and I could make a strong suggestion that you go and read this paper that we wrote in the British Medical Journal, BMJ. It's our lessons learned working on this collaborative over five, six or eight-year period. And it really tells the story of how we were able to make significant change at our institution. We started with paper order sets. So this was well before we had a computerized order entry system. We came up with algorithms based on the best quality evidence of what we should be doing for patients. We had one for surgical patients, one for medical patients, one for orthopedics, and we had them on the floor and people would use these, fill them out on paper to remind them of what's the best practice prophylaxis for patients. And this can work. So if you're in a hospital that has this capability, that's what you should be doing. If you're still on paper order set or paper orders, this is certainly an approach that can make a difference. However, it didn't fix the problem completely and we're still having patients not being prescribed appropriate prophylaxis. So we went on and we took a more advanced approach and we used a mandatory risk stratification tool that got built into our computerized provider order entry or ordering system, and it used advanced computerized clinical decision support. And what I'd like to do is show you what that looks like. So, these are some screenshots from an older version of our system but it will give you an idea of what it looks like. So this is an example of the general surgery admission order set. Here's the parent order set. And as you can see, there's a piece that says VTE prophylaxis for general surgery. So general surgery patients get different prophylaxis than medicine patients. So there's a specific algorithm just for those general surgery patients, and it's mandatory. So here's what the back end looks like. This is a flow diagram of what you should be doing for general surgery patients. The computer programmers out there love this. It's very analytical. The average frontline clinician physician nurse doesn't love looking at something this complex. So what we've been able to do is build the complexity in the behind-the-scenes part of the computerized order entry system, so that the provider doesn't have to look at this complex flow diagram and can just answer a few easy questions. So we've taken this complex thing and put it into something that's relatively easy. So this is what the provider will view. They will view this part where all they have to do is click on a few boxes for what are the patient's risk factors you can see at the top and what are their contraindications to pharmacologic prophylaxis. So for example, if I have a patient who is undergoing major surgery or they have cancer, I clicked their risk factors. If they don't have any contraindications, I would click none or I can click that they have major bleeding, and then the computer would go back to that algorithm I just showed you and come up with a recommendation of exactly what to give. And right at the bottom, it would say, for example, give enoxaparin 40 milligrams once a day or give unfractionated heparin 5,000 units every eight hours. It gives the recommendation based on the algorithm based on the easy checklist that got filled out ahead of time. So, what are the benefits of this computerized system? It puts VTE prevention right into the workflow. As the resident physician or nurse practitioner or physician assistant is filling in the orders, they get guided on what they should be doing to give the best practice prophylaxis. It enables this rapid accurate risk stratification and risk appropriate prophylaxis right at the time the orders are getting written. It applies evidence directly to clinical care and that's what we should be doing. And on the back end, I'll show you what this looks like later, it allows for performance monitoring and reporting. We can actually go back and pull data to see if the providers are giving what they should. If they're following the recommendation of the computerized decision support. So then, we can see how they're doing. So what are the keys to success? You can't just turn this on overnight. I work with a great multidisciplinary team. I've worked with the same hematologist, Mike Strifer, for over a decade. I've worked with some of the same nurses, Deb Hobson, for over a decade. Same pharmacist, Peggy Crouse, for over a decade. Informatics, we have an expert, Brandon Lao, informatics expert. Our multidisciplinary team has been together for a very, very long time working on this project. And I don't think that we can just do doctors doing doctor things and nurses doing nurse things, you really need a multidisciplinary team to work together. It's important that leadership buy in. The dean of the medical school knew exactly who I was when he was the dean because of this project. He would say hi to me by first name and he had thousands of faculty members and tens of thousands of people worked for him. But he knew who I was. Leadership buy in from the president the hospital, the vice president of medical affairs, all the different levels. The leadership is very important. We collaborated with service teams. We went out and talked to orthopedics and talked to OB-GYN. We educated front line providers. We went out and told them what this is going to be like, how are they going to do it. And like I said, we can measure performance. We measured baseline performance and we also measured ongoing performance evaluations to get information about how people are doing using this system. So, does improving this prophylaxis change the outcome? We were obviously hopeful that the answer would be yes and I am going to show you that the answer truly is yes. And I'm going to give two examples both from Johns Hopkins, one of the trauma surgery service and one of Johns Hopkins internal medicine. So, the first on the trauma surgery service. We published this paper in Archives of Surgery and it was a single center Johns Hopkins Hospital study, a pre-post intervention study. We have one year's worth of pre-data and three years worth of post data and we obviously had IRB approval to do the project. And here's the main finding of the paper. The first thing I want you to look at is, in 2007, that's the pre-data, we were giving or writing appropriate prophylaxis orders for about 60 percent of patients admitted to the trauma surgery service. After we implemented our decision support tool, computerized decision support, we were at about 85 percent for the next three years on average. So we went up from about 60 to 85, a significant improvement in order compliance. VTE events trended down and moved down in the right direction. We were not powered to show statistically significant difference and we didn't, but it certainly moved in the right way. And then the last thing is when you define preventable harm as patients who have VTE event who had not been getting appropriate prophylaxis, this decreased significantly from the pre to the post period. We had a few random events in that last quarter or that last year of 2010. But for two full years, we eliminated preventable harm and had a significant drop in preventable harm from the before period. Similarly, we did another project on the internal medicine service. We looked at about a thousand patients pre versus post and patients being prescribed optimal prophylaxis went from about 65 percent to now 90 percent. And the patients being prescribed no prophylaxis dropped dramatically. Now, what about the VTE events in this case? The total VTE episodes decreased significantly from about two and a half percent to under one percent, also statistically significant. And more importantly, we were able to realize a zero preventable harm number. We had zero patients in the after period who had an event, who had not been getting appropriate prophylaxis. We still had events, we still had seven of them, but every single one had been ordered appropriate prophylaxis. Zero preventable harm from VTE can be a realistic goal. So, I would say we have a model system for this. We won the North American thrombosis Forum's DVTeam Care Hospital Award for this program. We have been cited by AHRQ as one of the three top examples of effective implementation and clinical decision support for the work that we've done. And you can read about this on the AHRQ website and here's the reference. And most recently, we won, with our team, the CDC Healthcare-Associated VTE Prevention Challenge. We were one of those champions for our multidisciplinary team approach to achieve perfect prophylaxis.
