Speaker 1: Hello everyone, Jeff Dodson here, a solution architect with Qlik, and today I'm going to take you through our patient costing application. One of the challenges that people have when using analytics is how do we actually take our findings and report them back to our boss? How do we make these actionable insights? By leveraging Qlik Sense stories, you can drive your meeting from the data itself. A Qlik story is like Qlik's twist on a PowerPoint, and the advantage is that oftentimes when you're showing your findings, someone will inevitably question something or ask, wait, how did you get that number? That doesn't look right to me. And since we're presenting this from within Qlik Sense, I can actually right click and go to source, and we can see this insight within the exact context of the greater application and with the same filters applied that we may have had. We can look into the details, answer the question, any additional questions, and then get right back into the presentation. This eliminates the back and forth and allows us to take action from our findings by driving our meetings from the data itself and linking directly back to the dashboard. Now we can step through our findings. What we see here is that overall performance in 2020 is looking a little bit rough. COVID has been a huge impact, and profitability has not been good to say the least. We've identified a couple of problem physicians in particular who work in obstetrics, which was already known to be unpredictable even without a global pandemic going on. We've highlighted also an anomaly here that we noticed related to 20-year-olds, and what we want to do is explore that a little bit further. Here we highlight how we discovered that the largest loss is related to C-sections, and we know that there is an NHS-wide policy to reduce elective C-sections as they're often unnecessary and extremely expensive. Lastly, we highlight the three physicians who are doing many more C-sections than other physicians, including the two worst-performing physicians. By reducing the number of these, we can help our profitability, and this is especially important in these uncertain times. And once again, we link back to the source right here. We're linking our meetings with data, and we can examine the details and perform further analysis. Now the question becomes, what else can we discover in our data? Click helps you see where you think you are making money, and where you might not actually be making money. If I look at my selections and my filters here, I can see that highlighted in green are our selections. Then we have highlighted in white the results of our selections. But then everything else is in gray. So in the dark gray, we have unassociated items. And then in light gray, we have hidden or alternate selections, things that we could have selected as alternates to our current selections that we might want to be considering. So when is it important to see this gray? Well, let me tell you a real story about a healthcare customer that we were doing a proof of concept for. They had just gone through a bunch of cost-cutting measures, and they realized that they were using an expensive brand of surgical glue that they could save a ton of money on by replacing using a generic brand that had exactly the same clinical quality. During a proof of concept, as the customer was scrolling through all these surgical items, they clicked on this generic surgical glue. And suddenly, all of the counters went to zero. Every name in the list of surgeons became dark gray. They were unassociated. So what was the problem? Are the associations messed up because the architect built the data model wrong? Nope. In fact, the problem was because the cost-cutting committee had never told the surgical preference card builders to change the surgical glue that was being used. So when surgeries are going to happen, someone prepares the surgical trays with the items that the surgeon prefers to use, right, probably based off that specific type of surgery. And without those surgical preference cards being changed, the people preparing the trays continue to use that expensive surgical glue. The report writers often aren't asked to build reports showing you what isn't being used. It's always about counts of what is being used. And as you can imagine, with thousands of items being used for so many surgeries and with so many different surgical preferences, those reports never unveiled that the committee forgot to take the final step in the process, action. And that's why we want to highlight the gray. That helps us drive actionable insights. Now let's say that we are months into a project and we want to determine what we are really using our money on. As we move through this dashboard, we can clearly see that we were already patchy with regards to performance as a whole, but that we are especially losing money in 2020. We know that overall this is particularly the case for elective surgeries as a result of them being pushed back to save beds for COVID patients. So the question that I might want to ask is, how else is COVID affecting our numbers? I don't have COVID hotspot information in my EMR, but what I can do is pull this information from the Snowflake data market. It's easy to supplement what you're already doing with Qlik by leveraging data that is available with the data market. And that's exactly what I've done here. So this map was already showing us which states were doing well or not from a profitability and costing perspective. One's doing worse being in red. And I've added in a heat map layer to this to show the COVID hotspots from the data that we got from Snowflake. With this, I might want to ask the question of, are we losing money localized to a specific state or region? So maybe I want to drill into some place like California, where we were already not doing too well from a cost perspective and where there is a COVID hotspot. Now that we've zoomed in on California, I can see our patient locations as dots as well. I might want to dive into the details about specific counties, like Kings County or Kern, for example, where we have a lot of COVID cases and are struggling from a costing perspective. Then examining our patients. Is there a correlation between their relative location compared to the COVID hotspots? Is their distance from our locations affecting them? Maybe we want to add in drive times to look at this. And what about other states and hotspots? Is there any relation to COVID there? The point is, by adding in this extra data set from Snowflake that I didn't have access to previously, I've opened up a whole new world of questions that I can ask and analytics that I can dive into.