In this conversation, Eric and I discuss the importance of and process of data collection at SpecialtyCare and how this information is analyzed and converted into peer-reviewed research papers.

Speaker 1: Bringing you conversations with leaders within the operating room and healthcare community, this is Scrubbing In.

Todd Schlosser: Hello and welcome to this episode of Scrubbing In, a podcast powered by Specialty Care. I’m Todd Schlosser and today my guest is Eric Tesdahl, senior biostatistician at Specialty Care. In this conversation Eric and and I discuss the importance of and process of data collection at Specialty Care, and how this information is analyzed and converted into peer reviewed research papers. Enjoy the conversation.

Thank you for joining us here on Scrubbing In. I am Todd Schlosser and of course I am joined by our special guest Eric Tesdahl, who is the senior biostatistician at Specialty Care. I did a little stalking on your Linked In profile. I saw that you actually have a doctorate in philosophy from Vanderbilt.

Eric Tesdahl: That’s right, yeah.

Todd Schlosser: When did you graduate?

Eric Tesdahl: 2013.

Todd Schlosser: Okay. And you’ve been working sort of in the biostatistician realm since then I’d imagine?

Eric Tesdahl: Yeah. It’s kind of an interesting story. I’m trained as a social scientist, as a researcher.

Todd Schlosser: Oh.

Eric Tesdahl: Yeah. While I was doing that, while I was a student, you know you have to learn a lot of statistics to do that job.

Todd Schlosser: I’d imagine.

Eric Tesdahl: A lot of statistical modeling. I started to get contacted by people in the medical center at Vanderbilt to do consulting in biostatistics. It’s something that
kind of evolved as a side business because of skills that I had as a researcher that really kind of took off when I finished my post-doc.

Todd Schlosser: Right. When you started at Specialty Care you started I guess as a senior biostatistician.

Eric Tesdahl: That’s right.

Todd Schlosser: And you now take these numbers that come in from the people in the field, our clinicians in the field, and sort of answer questions around the data that you’re getting.

Eric Tesdahl: Absolutely.

Todd Schlosser: I wanted to talk a little bit about that. You told me the story of sort of what a statistician does in general. You collect all this data from stories that come from in our case the clinicians. Well you sort of analyze that data, and then you have to turn that analyzed data back into a story. That comes out as like a research paper and you sort of publish those.

Eric Tesdahl: Yeah.

Todd Schlosser: I’d like to talk about that lifecycle of what those research papers are and how they come to fruition I should say. How do you collect the initial stories?

Eric Tesdahl: The nice thing about working in a big organization that’s as amazingly staffed as it is is that I don’t collect really the stories or the information. In a lot of sense the thing you were mentioning about a statistician is I’m kind of a middle man. I’ve got a lot of tools for dealing with data, but that’s kind of its own specific realm. So when I want to do that I look to people who actually know phenomenon really well. Here in Specialty Care we have people who are really the best in profusion, or neuro monitoring, or anything to do with surgical services. They’re the ones who know the questions. They have done their practice for decades. They know the important questions to ask.

The information comes from all the people, clinicians, who are doing procedures every day. They’re entering that into the scope database and over time our clinical experts, Al Stammers or Tony Sestokas, they’ll say hey there’s a really important question that I think we have enough data to answer it. That’s the genesis of the question and that’s where I get involved. People want to say how do we understand what’s the best practice to use in this situation, how do we really help our patients to get to the other side of this surgery as safely as possible.

Todd Schlosser: And that’s all coming from the clinicians in the field in our case?

Eric Tesdahl: Yeah, the clinicians in the field or other discussions that are going on in journals, or things that payers want to know about. There’s a lot of parties that are involved, as you know.

Todd Schlosser: Sure. Yeah so if it’s profusion it might be like an Al Stammers or a Linda Mongero type figure in the I guess realm of profusion that says hey we should focus on this.

Eric Tesdahl: Absolutely.

Todd Schlosser: Is that … And then like a Tony Sestokas for IONM.

Eric Tesdahl: Exactly right, or Brian Willett.

Todd Schlosser: Oh yeah, absolutely. We’ve actually interviewed all of those people on Scrubbing In.

Eric Tesdahl: Exactly right. Yeah, if you have a conversation with them and you ask them, they could probably list a half a dozen or more just pressing research questions
that they have built up in their mind because they see where the field is at and where they think it needs to go. I get involved with the tools and the knowledge that I have about what it takes to really answer some of those questions with data. There are modeling techniques that we do. There are ways to take the data and visualize it and help us think more deeply about what we’re seeing. Then at some point this often morphs into a conference presentation. We talk about the lifecycle-

Todd Schlosser: Where you actually go into the conference?

Eric Tesdahl: Yeah. We talk about the lifecycle of one of these papers, it almost always starts out as a conference presentation. There’s an upcoming conference, there will be a deadline, please submit an abstract. That’s just kind of short summary of some research that you’ve done. We’ll start in on one of these questions, get some data. I’ll do some analysis. We’ll have a bunch of back and forth conversations. I’ll say this is what it sort of looks like, does this make sense, and we’ll kind of interpret it together and pursue the analysis. Then we’ll get to some good working draft of what we have and submit that. Folks at the conference will review it. They’re very knowledgeable in the clinical areas. They’ll say oh this is going to add value to our conference, we would like to invite you to come and make a full presentation.

Then at that point we go back to what we started and really flush everything out, address all the what if questions that come in. You deal with something like cardiac surgery, it’s a really mature field. A lot of people have working in it and answering questions for a long time, so there are a lot of intervening factors you need to account for if you want to answer one particular new question. You have to account for all the old questions, and then say okay here’s my new question. Scope is really an excellent tool for that because-

Todd Schlosser: Scope being the database where we collect all this information.

Eric Tesdahl: Exactly.

Todd Schlosser: It sort of houses all the information.

Eric Tesdahl: Exactly. Thank you. It’s an excellent resource because we have a lot of patients that we help, so that means we have a lot of observations. As a statistician it’s really important you have big sample size so you can be sure of what you’re seeing. But we also collect a lot of individual pieces of information. Each of those helps us to say accounting for X and Y and Z and the other thing, setting that all to the side, we can then answer a new particular question. It’s not because of age or gender, or how big a person was and their body size. Yeah, so that’s a really nice aspect of working for Specialty Care.

Todd Schlosser: Sure absolutely because you have a lot of, the pool of data you have to play in is much bigger I’d imagine because Specialty Care is the largest provider of things like profusion and IONM. Just the data sets we’re collecting and housing in Scope for you and your team to look at is bigger.

Eric Tesdahl: Absolutely.

Todd Schlosser: You can avoid things like confirmation biases and stuff like that.

Eric Tesdahl: That’s right.

Todd Schlosser: Let me ask this. When you’ve turned that story into data and you’ve analyzed it like you have and before you actually make a research paper, how do you go from the process of analyzed data to telling the story again in a research paper?

Eric Tesdahl: That’s right.

Todd Schlosser: How do you make that conversion?

Eric Tesdahl: Yeah. Again, in very close collaboration with our clinical experts. What I’m able to say from my perspective is here’s what patterns we can see in the data, but apart from understanding those patterns there’s still a big gulf to get to what’s the story. It takes somebody like Linda, or Andy, or Al to be able to say okay let’s put these numbers in context. They’ll start to say oh that’s really high, that’s high for a patient to experience, or that’s a low value for a patient to experience, or this value really does not conform to more people’s expectation. They have the background of all stories and the way people kind of see this practice, and then our patterns are kind of set against that and say is this really what we thought we would find, is this really significantly different.

Then you start to put the story together. It’s just really a back and forth aspect. Usually I’ll supply the analysis, they’ll take time to interpret and digest it and write about it, and then I’ll look at what they’ve written and I’ll put my spin on it and say you know what we kind of change it a little here, a little there. Then what we end up with is our preliminary version of what we think this data is saying, which then is reviewed and hammered by outside entities and everything else.

Todd Schlosser: Yeah, but that’s where you might start to go to conferences and things like that.

Eric Tesdahl: Exactly. Right.

Todd Schlosser: And start to present. Let me ask this. The person we’re actually getting each story from that makes up the large pool that is housed in the Scope database may see trends, but they only see trends from the people and the cases they’re actually touching. Having a Eric Tesdahl or Al Stammers, or Mongero, or Tony Sestokas who has years in the field and came up through profusion or IONM and are thought leaders in that space, and having you interpret the data and then them telling you sort of how to shape that into the story, I can see that as being incredibly beneficial because the person in the actual OR only has their experiences to guide what they should do next. Their experiences may be 100% accurate, but that is that confirmation bias where it actually works out. If you don’t have a team like what you guys have, you don’t get sort of a more rounded I guess scope of what we’re actually trying to do and the best way to prevent negative outcomes.

Eric Tesdahl: Yeah no, absolutely. The number is really only a stand in for the fact that one person can’t witness all of these happenings and keep in straight in their mind and sort it out. It’s just too much for one person. So we convert it to a number and we can do some things with those numbers, and then hopefully we’re faithfully converting it back to a story, which will help. In the end the whole point is to make surgery safer. If we’re able to engage in this process, the numbers really don’t have any value in and of themselves. They’re just representing a story that we want to move forward.

Todd Schlosser: Right. Once you’ve converted the data into a story and you’re starting to go to these conferences, I’d imagine that other people in the industry while they’re at the conferences, or maybe they get wind of it through just the networking that goes on at these things, they will have also their opinions that they will cast onto the research paper. How does that process work?

Eric Tesdahl: Yeah. It is difficult and it’s meant to be difficult.

Todd Schlosser: Sure because it’s not a white paper that anybody can put out and say hey this is what this looks like. It’s actually peer reviewed. That’s why research papers are so valued.

Eric Tesdahl: That’s right.

Todd Schlosser: And that’s why they’re so difficult to write and it’s a long process. But how does that process work?

Eric Tesdahl: Typically we make the initial abstract submission and that’s the first check. If it’s somehow missing something fundamental in the analysis it usually won’t be selected for a competitive refereed meeting. That’s one check. If you get invited to the meeting, which we most commonly do, then we go and make our presentation, which is generally representing a draft of a manuscript. We start with something small and submit it. Then it gets accepted we go to present we have more of a full flushed out analysis and we submit a paper. At that point then it’s going to go to three anonymous reviewers. When someone’s reading the paper that we’ve submitted they don’t know it’s us, it’s Specialty Care. I mean they might be able to figure out if they really know the field and they see the sample size. They’re like okay this is probably them.

Todd Schlosser: That’s a large sample size it’s got to be Specialty Care.

Eric Tesdahl: Right. It’s their job to give an impartial review and say this isn’t about how I know or don’t know the people involved, this is about what constitutes a valid question and a valid analysis. Every time you get a review you get a bunch of really difficult critical pressing questions.

Todd Schlosser: Right.

Eric Tesdahl: Okay maybe sometimes if you work with somebody’s who really great they’ll come back and they’ll give you a clean paper and I’ll be like I like working with this author. As a statistician that’s always really nice, you say okay. But often you find, they’ll always find something because part of the job of the reviewer is to make the paper better no matter where it’s at. That’s the part, then we get that feedback back and sometimes it’s like oh this was unfair and you have to go through the ego process of they’re attacking my work, I can’t believe it all.

Todd Schlosser: Yeah, I think that part would be difficult for me.

Eric Tesdahl: Yeah. It’s really funny. There’s a whole running joke in academia about reviewer two must be stopped. It’s like there’s always one reviewer who is a zealot and really trying to go so far. In the end this is meant to improve everything and so you have to just set your ego aside and say if they found reason to criticize it likely other reasonable people would as well, and so I need to address it. Sometimes it’s a matter of changing how you’ve described something. They say well your data say this and then you’re saying a little something else. You need to be more in line with what your data say. Or sometimes they say you know what let’s change that analysis around. I think you probably need to change the focus and add another variable or something. That’s where I get involved and rerun the analyses. Then we resubmit it. If the reviewers are satisfied, say okay you addressed my issue, you answered my question, they’ll give it a big green check mark and send it back to the editor and it will be published.

Todd Schlosser: Awesome. Is it typically published just online, or are there medical journals they go into? I’d imagine even medical journals now have moved to online.

Eric Tesdahl: Yeah, most all the journals are online. Print is still … A big issue in access whether it’s online or print is pay access. That’s still in the medical field really just very traditional.

Todd Schlosser: Just hiding the research behind some sort of pay wall?

Eric Tesdahl: Yeah, that’s right.

Todd Schlosser: Okay.

Eric Tesdahl: The big publishers are good at setting up a system that facilitates the process and so in exchange we pay some access fees and things.

Todd Schlosser: I’d imagine that people who want access to the research paper also pays those fees because they want to make their surgery safer too. They need access to this information.

Eric Tesdahl: Yeah. Typically that’s through a, if you have a university affiliation, that university library has its subscriptions to this particular journal and then you can go get it. That’s the way most people find their way to these things. Sometimes we do make summaries of our research available. We wrote it, it’s our information and we can’t necessarily distribute that same version, but we can summarize it and say here’s what we found and refer to the published version, disseminate the information.

Todd Schlosser: Sure. Or you can go to conferences and give talks on it right?

Eric Tesdahl: Exactly right. It’s not a secret.

Todd Schlosser: Well and we’re not trying to keep it a secret. You want to let as many people know so they can make their surgeries safer because ultimately patient care is the focus.

Eric Tesdahl: Exactly right.

Todd Schlosser: We’re always trying to improve that. One thing that struck me when we were talking before was you’re not necessarily trying to answer the what question but the why question. That sort of struck me. I sort of wonder if I can get you to explain that, so how you find what the why question is and how you define what it is.

Eric Tesdahl: Yeah. I mean the general answer is that it all depends on the state of the research in your field. A good contrast is actually the IONM and the profusion. Profusion has a longer history-

Todd Schlosser: Yes it does.

Eric Tesdahl: Because cardiac surgery has a longer history. Neuro monitoring as a discipline is much younger. What it takes to say something novel and useful in the field is very different because like I said in profusion you have a lot of people who have come before you.

Todd Schlosser: Yeah, it’s got like 50 extra years on IONM.

Eric Tesdahl: That’s right. So a lot of other ideas about what is important and what you need to account for and getting to a why question. The what questions are pretty straightforward. My fitness tracker can tell me what I did last week without much trouble. It can’t tell me why. I would need a lot of information from a lot of people.

Todd Schlosser: A lot of data set.

Eric Tesdahl: A lot of data. That’s what we have in when we go to answer a question in profusion we have to take a lot more of that approach. In IONM it’s a younger field and so while we are able to say some really sophisticated things, account for the kind of surgery you’re talking about, the region where it’s happening, where there still kind of developing their knowledge base a little bit more, so a different set of requirements.

Todd Schlosser: I’d imagine it’s a little easier in the IONM field just because a lot of the questions have already been answered for profusion, and IONM just being younger they just have less time to have answered those questions.

Eric Tesdahl: As a person trained with a research background I can tell there’s never a shortage of questions.

Todd Schlosser: Sure.

Eric Tesdahl: There will never be a shortage of questions because in any of these medical procedures you have risks that need to be attended to and minimized. Take profusion for example. There’s a really good example that a lot of patients get blood transfusions and those lead to downstream consequences. There’s certainly a lot of downstream consequences that as a profusion service provider we want to learn as much as we can about how we impact those, kidney injury, how long you stay in the hospital, the possibility of infection. Until those don’t exist anymore, until surgery has no bad outcomes, we’ll always have questions to work on.

Todd Schlosser: Fair enough.

Eric Tesdahl: This is about taking care of people.

Todd Schlosser: Yeah. I mean ultimately you want to reduce any sort of bad outcome and that’s going to take a very very long time.

Eric Tesdahl: That’s right, yeah.

Todd Schlosser: I’d imagine that some results can change outcomes downstream and they’re not always changed for the better, so you may have to change the way we look at something because of something we’ve changed.

Eric Tesdahl: Absolutely.

Todd Schlosser: There’s all kind of unintended consequences that could come up.

Eric Tesdahl: That’s right, yeah. Some of the more fun analyses that we’ve done on the profusion side are able to account for competing options for taking care of the patient. It’s like we’re going to do option A, or I want to be option B, or sometimes people do A and B. You kind of look at the combination of these things. It turns out some unpredictable results will happen. You can see from our data when people are pursuing each of these like well you may want to choose this particular strategy when you’re using these two tools together because improving one thing may lead to, like you say, a decrease or a worse outcome on something else.

Todd Schlosser: Yeah. Awesome. Well hey that’s all the time we have for today. I feel like I could talk about this for a very long time just because it interests me, but I really appreciate you coming in. It was very informative. Thank you so much for stopping by.

Eric Tesdahl: Absolutely. Thank you.

Speaker 1: Thanks for listening to Scrubbing In. Please take a second to give us a rating on your podcast app and subscribe so you won’t miss out on what we have coming up. See you next time.