Outcome measures are an important part of EMS Compass, and NEMSIS version 3 has the capability to make these measures readily available to every agency, large and small. To see how some agencies are using outcome measures, we talked to Jeffrey L. Jarvis, MD, MS, EMT-P, the medical director of Marble Falls Area EMS and Williamson County EMS, both located outside of Austin, Texas. For nearly two years, these agencies have been receiving patient outcome data, facilitated by their ePCR vendor. We asked Jarvis about how having that information will help EMS and patients.
What’s your system doing in terms of connecting EMS and hospital data?
A. There are three main hospital chains in our region. One of those has been hooked up (to their equivalent to a Health Information Exchange) for about two years now … and the other two are in the process and should be coming online very soon.
Let’s say you ran a call before we were hooked up. There’s a patient with chest pain. You take him to the ED [and] traditionally when you drop him off, the patient vanishes. You never know what happened to him.
Now you’re able to (find out the disposition of the patient) on your next shift. You’ll log in (to the EMS reporting software) and on that initial dashboard, there’s an application that will pop up that says, “You have outcomes.”
How does that work?
A. You click on that and it will show you a list of every patient that you transported to one of those facilities that’s participating. Let’s use for example your chest pain patient. You thought he was having an MI but he was actually released from the emergency department with a diagnosis of pneumonia. And then you can see from looking at the lab that he had an elevated white blood cell count and they had a chest X-ray done which shows an infiltrate. And then it’ll actually show their length of stay in the ED was an hour and a half. Then you can link back to a PDF of your run record too, so you have access to all of that.
Are you doing that for every patient who’s transported?
A. Every patient who’s transported to one of the participating facilities. Right now we’re doing it with Saint David’s. It’s part of the HCA—Hospital Corporation of America—chain. Any HCA hospital will get that information
Jeffrey L. Jarvis, MD, MS, EMT-P
How are the medics reacting to it?
A. We initially thought that that was going to decrease the number of follow-ups (with medics) because they already had the information. It actually turns out it’s increasing the number of follow-ups because they (get this feedback) and it just gives them more questions to ask. ‘Oh that’s great, this patient only had pneumonia but jeez I wonder what the temperature was and I wonder what antibiotic they went out on.’
I actually look at that as a good thing—that they’re getting more information about their patients and it’s absolutely helping them go, ‘Oh, you know, the pain really was kinda on the left lower side … and now that I think about it they had a fever and they were coughing up this green nasty stuff. Maybe I should’ve been thinking pneumonia instead of MI.’ So, I think it’s a nice feedback tool for the medics.
Now in terms of formal quality improvement, that’s a step we’re taking right now. So the first one of these reports that we’re looking at relates to STEMIs. So what we’re trying to find out is what are our sensitivity, specificity, false positive and false negative rates and overall ability to recognize STEMI in the field.
What other benefits do you see coming from this?
A. The next thing we’re going to do is a sepsis alert program. But to really do it I need a tool to measure lactate in the field. And for a variety of reasons, those things are hard to come by right now. So there’ve been some suggestions that end-tidal carbon dioxide (EtCO2) may be a good surrogate measure for lactate. So we want to try and figure this out. What we’re going to do is look at all the outcomes data and say, ‘Show me everybody we transported to that facility or those facilities, and any diagnosis that included sepsis,’ and then we’ll go back and look at what our EtCO2 levels were to determine if EtCO2 is a valid surrogate for lactate. We’re also independently looking to see if there is an EtCO2 level that can predict sepsis.
Do you eventually want to actually do the same analysis with your dispatch data to see if the EMD has correctly identified the chief complaint?
A. Yes, absolutely … I’m also the medical director for our dispatch … we’re in the process of writing that report right now that looks at dispatch determinant code compared to EMS impression. That’s an easy report to write. We can do that now because we get a determinant code that gets pushed over from our CAD for every call. One report that’s more interesting is looking at the determinant code and comparing that to an ED diagnosis.
How are you using your outcome data with performance measures?
A. Dr. Brent Myers and the Eagles did a paper a few years ago that recommended some clinical benchmarks and related performance measures. I thought that was a great place (to start) and yes we’re absolutely reporting on those measures.
It’s only been in the last couple years that technology has allowed the interoperability of data, correct?
A. Absolutely. Having a platform that provides bi-directional data flow between different hospital medical record systems, such as EPIC, Meditech or Cerner, and different EMS ePCR systems is crucial. There’s no way you can ask 18 different EMS systems to change their systems and certainly no way to ask a hospital to change their EMR. It’s great having technology that makes this data connection in a platform-agnostic fashion. It really takes a lot of the angst and push-back out of the equation.
What would you say to your EMS colleagues around the country who have come up against hospitals unwilling to share data?
A. If you would’ve asked me this question four years ago when I was feeling a little more blunt, and less politically correct, I think I would’ve told you that what needs to happen is that the medical director and system director need to go to the hospital, and say, ‘Listen, these are our patients, we’re referring them to you, you either give us that data back or we’re not giving you our patients.’ And I think that will get their attention.
If you ask me now, I would say you have some version of that, but you do it in the least offensive way possible … you don’t break out that big stick. But we’re negotiating from a position of strength. Of course, if you’re only transporting to one hospital then you can’t make that argument.
We also have our champions through the trauma registries at the hospitals. They saw great value in automating some of the registries that they cared for. They stepped up to say, ‘Hey this is really important and this is why.’