Tableau: Tell us a little about Piedmont—how large is the practice?
Mark Jackson, Business Intelligence Manager: Piedmont Healthcare is a five-hospital system. We have 400 employed medical staff members, and we have approximately 1,200 affiliate physicians with more than 100 physician practices across North Georgia.
Tableau: Wow, so you must handle a lot of data in your role there! Can you tell us more about that?
Mark: Well, I’ve been with the company now going on four years. I started out with the Heart Institute, which is a cardiovascular service line within the organization.
And we started to become a more data-driven organization. I mean, that’s kind of where the industry is going. You have to be able to respond rapidly to changes in the market, with all the changes that are coming in healthcare legislation. So it’s really a requirement for you to be able to adapt.
Tableau: Has Tableau helped you with that?
Mark: Well, I just needed a way to do my job quicker, faster, to be able to drive insight. And the first thing that I did with Tableau was to take what I was doing in Excel—a project with our catheterization (“cath”) and electrophysiology (EP) labs—and use Tableau to track the throughput for those labs.
I first started doing all this in Excel, which was taking me four hours a month—and could only be produced as information in aggregate, and always led to more questions and more work. Tableau really enabled me to fully automate that process. Now it runs itself.
Tableau: That’s great! So what happened next?
Mark: Well, as I continued to develop the work that I was doing, transitioning a lot of what I previously did in Excel to Tableau, I got recognized by corporate. And they said, "Okay, we're struggling with the same problems. We've got to be able to deliver this insight faster. We can't get the information to the decision-makers quick enough for us to be able to rapidly change in this environment."
They had these books that they would print out. They were monthly operating reports, and they were 133 pages long—and they produced one of those for every hospital. So by the time they reached the decision-makers’ hands, they were out of date. Using Tableau, we were essentially able to replace 2,400 pages with a single dashboard.
Tableau: Wow! And have you built other dashboards since then?
Mark: Yes. The second project that we did with Tableau was this project called Patient First. And the first phase of that project was about reorganizing physician schedules. We had more than 100 physicians and mid-level extenders and we needed to reorganize their entire day. We had them basically spread out doing many different things, creating some inefficiencies.
We leveraged Tableau to help us reorganize those schedules. And one of the effects of that was that we were able to cut out a third-party physician-consulting service and some additional third-party work that was being done on the weekends. That, in total, saved us a little over $2 million a year.
Tableau: That’s great. Has that impacted quality of care?
Mark: We’ve seen some real benefits from the Patient First project that we did, which allowed the physicians to really focus all their attention on one patient-centered activity at a time.
We’ve seen a 50-percent improvement in our variance from benchmark in cardiovascular length of stay. We’ve also seen a 10-percent reduction in heart failure readmissions, and we’ve seen a 12-percent reduction in readmissions for patients who are heart attack victims. And overall, that’s improved our patient satisfaction scores by 7 percent.
In addition to helping us setup the new physician schedules, Tableau has helped us track the results of our efforts for all phases of Patient First.
Tableau: That’s amazing! Are you using Tableau in any clinical spaces?
Mark: One of the cool things that we are doing is that we mine our imaging system for a certain set of criteria that makes somebody a good candidate for valve surgery. We can find potential candidates and talk to the person who ordered the imaging study, who can go back to their patient and say, “Hey, you might want to come in and see our vascular surgeon. You might benefit from a certain procedure.”
That's the kind of care that I want to know I'm getting when I'm going into any hospital system—that I'm not going to fall through the cracks, and they're not going to miss something, because people are on the back end doing smart things with all the information that they're collecting about me.
Tableau: Absolutely! So how many Piedmont employees use Tableau?
Mark: Well, with Tableau getting its start in the Heart Institute, we have several people in that entity who are pretty well-versed in it and doing lots of creative things. We have users in every entity and most departments at this point. So from one original license, we’re up to 85 licenses and we have a 16-core server implemented and about 250 unique users every week.
Tableau: Do people like the self-service aspect of Tableau?
Mark: Self-service has been a big buzzword in the BI space right now—and we’re definitely taking advantage of that.
What Tableau has allowed us to do is take what used to require this huge back office to maintain—this huge, lumbering BI infrastructure with these giant data warehouse projects that might take a year to complete—and our speed is so much faster. We can rapidly deploy these small data marts and put them in the hands of the end users who are facing the problems individually. And as we continue to develop with this project, I think people are going to be more and more completely self-reliant.
So it’s not been a top-down mandate where people have to use this. People are genuinely excited about it. And it’s all grown organically. We’ve given out a few licenses just to groups that we thought would be interested in it. And from there, people are doing their job more efficiently.
Tableau: Can you tell us more?
Mark: Well, I sent out a survey recently and the results tell me that, on average, people have seen a 23-percent improvement in their efficiency.
If you extrapolate and assign an hourly rate to those resources, looking at the amount of hours that they spend per week working on those things, that’s about $250,000 to $650,000 that we’ve reallocated.
One of the best success stories was a user in our IT department who automated work that used to take her 3.5 weeks per month when she did it all in Excel. We recreated those reports in Tableau and now she maybe spends 3 days per month on it and has time to focus on her new role instead of running reports every month.
Tableau makes me feel empowered, like there’s nothing that I can’t really accomplish here. I can take any problem, and as long as the data is available, we can solve it. And probably pretty quickly too.
Tableau: How about you?
Mark: Personally, I can say I’ve increased my efficiency probably 80 to 90 percent—and that’s just time to initial insight. Because what Tableau enables you to do is fully automate that insight going forward. So all the insight maintenance work that I would do throughout the year, well, I don’t have to do anymore. It’s amazing now when I look at the amount of data that we’re pushing through Tableau.
We have more than 70 data sources that we’re publishing to the server right now. And so I started taking the columns and the rows that we have, and it’s more than 20 billion points of data that we push through Tableau every day. So it’s this massive amount of information, but honestly, it just doesn’t feel like it’s that big.
Tableau makes me feel empowered, like there’s nothing that I can’t really accomplish here. I can take any problem, and as long as the data is available, we can solve it. And probably pretty quickly too.