Brandyn, thank you for coming down to talk to us and for this Coursera course on data and how we use data. Why don't you introduce yourself and tell us what you do? Well, thank you for the kind invitation. I'm Brandyn Lau. I'm an Assistant Professor of Radiology, and Radiological Science, and Health Sciences Informatics. I'm also a faculty in the Armstrong Institute for Patient Safety and Quality. A lot of my work as it applies to clinical informatics, focuses on quality improvement and healthcare organizations and largely around blood clot prevention in hospital settings. What's been a bit of blood clot? Well, blood clots are largely preventable cause of harm among hospitalized patients. More patients die every year than from AIDS, breast cancer and motor vehicle collisions combined and we can really reduce that number if we do the right thing. When a blood clot forms in the legs, in the veins of the legs, it can travel to the lungs and cause you to stop breathing. You said it's preventable. How is it prevented? It's prevented through multiple actions by multiple different providers. First, prescribers need to consider the risk of the patient. They need to assess the patient's risk when they come into the hospital. Based on the output of that risk profile, they need to prescribe the right medications or mechanical interventions to prevent the blood clots. Nurses need to administer the medications and patients need to accept the medications as they're prescribed. So, a lot of different players. It's real system intervention. It really is. So, you've been working on this for years. Years. Right. Have you made any progress? I've made a little progress. We've been able to implement a clinical decision support tools at multiple levels to assist prescribers, and nurses, and even patients about the harms of VT or venous thromboembolism blood clots. The importance of prevention practices which often takes the form of medication or mechanical boots that squeeze the legs to prevent blood clots from forming. What are the measures of your success? The degree to which those physicians are following the guidelines? Yes. So, we actually have multiple measures of success. First, blood clot prevention is one of the most frequent measures of healthcare quality. So, there are a number of organizations like the Joint Commission, the Centers for Medicare and Medicaid Services, the agency for Healthcare Research and Quality, that are measuring how we're doing on preventing blood clots. So, we're beholden to multiple different organizations that have standardized measures. But within our health system, we look at a couple of different things to determine how we're doing. Those take the form of process measures, and outcome measures, and process linked outcome measures. When I talk about those, let me go into detail there. When we talk about process measures, we wonder what percent of patients are risk assessed when they come to the hospital? Our goal for that as always 100 percent. We want to make sure that 100 percent of patients have their risk assessed. You might be high risk, you might be low risk, but we don't know that unless your risk is assessed. Next, we want to see what percent of patients are prescribed risk appropriate prevention therapy or prophylaxis? Our goal for that again is 100 percent. If you're particularly low risk or you have a high risk of bleeding, we don't want you to be prescribed anticoagulants. If you are at low risk for clotting or you are at high risk for bleeding, we want to make sure that you're not prescribed anticoagulant. So, what we want to make sure is that patients are prescribed the right thing for their risk profiles. So, we always want that risk appropriate prescription to be 100 percent as well. So, I am hearing both data and intervention. Yes. So, these measures you need data to know how you're doing. You need data to know to figure out who's at risk, and that you need to intervene with the decision support. So, you're involved in both sides of our little construct loop for informatics. So, say more about the role that data plays in your work. I think that data plays an absolutely critical role in understanding what it is that we're doing in hospitals at large. So, currently we're collecting more data in health care than has ever been collected in the history of medicine. It's our responsibility to use it to understand how it is that we're currently practicing and what are opportunities to improve on. When I first got involved in blood clot prevention work, we took a snapshot of data to look at how we were doing in the hospital. We found that less than 30 percent of patients were prescribed the right medication to prevent blood clots. Your target is 100 percent? Our target is a 100 percent. But, it's not good. It's not, especially for a condition that kills more people than AIDS, breast cancer and motor vehicle collisions combined. So, we saw that as an incredible opportunity. We first needed to find ways to improve risk assessment. With that initial 25 percent of patients that were being prescribed the right thing, nobody actually underwent a formal risk assessment, that was just what happened to get prescribed for them. In 25 percent of the cases it happened to be right. So, our first effort was to make sure that patients were actually risk assessed when they came into the hospital. So, we implemented a mandatory clinical decision support tool that required risk assessment for every patient when they came into the hospital. You designed it so well that even though it was mandatory they didn't hate you. Well, they did hate us probably for the first few months, but then it becomes a part of their workflow. It becomes a standard practice and they start to hate it less as we implement other things for them to hate. But also because if you start doing the work you don't get the punishment of the decision support, so it's sounds a little bit more carrot than stick. Well, we also offered an additional benefit. When they completed the risk assessment, they were presented with what the risk appropriate prophylaxis regimen was, displayed for them on the screen based on the data that they entered. It gave them the option of prescribing that prophylaxis regimens. I misspoke. I meant that it sounds like you do more stick, but this is a good carrot. I think that when you consider what health IT does for busy clinical providers, a lot of times it adds extra burden. What we want to make sure is that we're providing something back to the decision-makers that lessens the burden of what we're asking them to do. Because when you think about providers coming into the hospital, nobody wants to do the wrong thing for their patient. Nobody comes into the hospital and says, "I want to harm my patients today." Everyone wants to do the right thing. But as informaticians, it's our job to empower clinicians and patients to do the right thing, with the right information, at the right time, in the right format. I heard those five rights before. So, thank you.
