S1E31: Enterprise Imaging Strategy (ft. Cheryl Petersilge, Vidagos, Cleveland Clinic)

Cheryl Petersilge, CEO of Vidagos Enterprise Imaging and former Medical Director of Integrated Content and Enterprise Imaging at Cleveland Clinic, speaks naming conventions and standards as part of a holistic enterprise imaging strategy.

Transcript:

0:0:0.0 –> 0:0:10.590
Jordan Cooper
We’re here today with Cheryl Peter Silge, the founder and CEO of Vagos and Enterprise imaging strategy consulting practice. Thank you for joining us today. Cheryl.

0:0:11.30 –> 0:0:13.500
Petersilge, Cheryl
Thank you for having me, Jordan. It’s great to be here.

0:0:14.320 –> 0:0:44.200
Jordan Cooper
So Cheryl is also the former medical director of integrated content and enterprise imaging at Cleveland Clinic. She is currently one of the Co chairs of the hymns SIM Enterprise imaging community and describes herself as an enterprise pioneer, a practicing radiologist, and accomplished informaticist, and an engineers. Quite a mouthful of titles and roles, and she brings it very interesting perspective. I’m looking forward to speaking with you Cheryl today.

0:0:44.870 –> 0:0:46.170
Petersilge, Cheryl
Great. Great.

0:0:46.780 –> 0:1:15.710
Jordan Cooper
So we’re gonna be covering 2 main topics in our in our episode today. One will be setting up enterprise imaging for success as a peer component to the electronic health record to complete with to compete or complete the electronic health record. Sorry about that. And then we’re gonna delve into a few use cases without naming that particular customers. So with that, I’d like to open it up to you and tell us what should we be mindful of with setting me up the.

0:1:15.800 –> 0:1:17.340
Jordan Cooper
Enterprise imaging for success.

0:1:18.490 –> 0:1:37.630
Petersilge, Cheryl
So I think there are a couple things that are well known in, I’ll say the industry that are very important to the success of an enterprise imaging program. Number one is having good governance and I think this is an area where many organizations.

0:1:39.110 –> 0:2:8.360
Petersilge, Cheryl
I don’t wanna say our challenge, but it’s it’s something that they haven’t really gone out and developed and oftentimes they fall back on radiology and use radiology as their governance. But you really need a group of individuals and of course, radiology’s gonna be at that table. But you need a group of individuals who are thinking about the enterprise and what’s best for the enterprise, what’s best for our patients in terms of.

0:2:8.500 –> 0:2:39.710
Petersilge, Cheryl
How we’re going to manage all of our different medical images and then second is choosing a vendor partner. They should not be. It should not be a transactional relationship. You really want to be working with somebody who is partnering in your success. They want you to be successful. As someone said, we’re not successful two year successful and and that’s really what you want. And there are a lot of.

0:2:39.850 –> 0:3:11.0
Petersilge, Cheryl
Different strategies out there and different offerings among all of the vendors who are in the enterprise imaging space. And so going back to the governance piece and developing your strategy, know your organization know what’s important in your organization and what you want to achieve with your enterprise imaging program. And then when you understand that, you can really go out there and look and find that partner that really.

0:3:11.120 –> 0:3:13.580
Petersilge, Cheryl
The best pairs with your needs.

0:3:14.370 –> 0:3:45.240
Jordan Cooper
So Cheryl, I think in order to make some of your recommendations about good governance and selecting a vendor partner more concrete for our listeners, now might be an appropriate time to delve into some of your customer use cases. You’ve you’ve mentioned that one of your recent customers is a multi state system, does eager to develop a governance strategy like you just described. You said the process helped them understand the spec full spectrum of what enterprise imaging touches help them develop a framework for implementation.

0:3:45.380 –> 0:3:52.550
Jordan Cooper
And highlighted the need to help them guide existing programs, particularly their focus initiative. Can you tell our listeners more about that?

0:3:53.340 –> 0:3:53.580
Petersilge, Cheryl
So.

0:3:53.660 –> 0:4:23.120
Petersilge, Cheryl
So this is a client who has an awesome CMIO. Who was my point of contact and my just to stop there for a moment. My point of contact really varies along the organizations, but I really love that it’s the Chief technology Officer, the Chief Information Officer or the chief Medical Information Officer because that’s really the level that is looking across the enterprise so.

0:4:23.970 –> 0:4:34.780
Petersilge, Cheryl
Got to work with an awesome chief Medical Information Officer who realized that she needed to bring all of the different imaging specialties.

0:4:36.640 –> 0:4:49.290
Petersilge, Cheryl
From across the organization to work together to share the same vision, this organization had a very well developed pocus point of care ultrasound.

0:4:50.750 –> 0:5:7.920
Petersilge, Cheryl
Governance Committee amongst itself, but they were still operating in a silo. They were working with the tools they had, which did not include the software to support and encounters based workflow. So they were struggling with that.

0:5:9.240 –> 0:5:31.790
Petersilge, Cheryl
Didn’t look at how it was impacting other programs that were going on that the teams were managing and that’s a huge one that you hear a lot. You know we only have so many resources and we need to prioritize. So you know we needed to bring that focus to the team and then we needed to look at how they were.

0:5:33.50 –> 0:6:5.80
Petersilge, Cheryl
Labeling their images, which is something that’s really important to me, and I think oftentimes underestimated the importance of having good and call them study descriptors or procedure names as opposed to orders because orders is only one part of it. So it’s how are we labeling that imaging file? And you don’t want the imaging file that contains very similar images to be labeled differently depending on who acquired those images.

0:6:5.150 –> 0:6:11.860
Petersilge, Cheryl
Because our underlying philosophy, especially when it comes to point of care, ultrasound should be that.

0:6:12.770 –> 0:6:45.80
Petersilge, Cheryl
It’s not dependent on the specialty that’s acquiring the images. It should be dependent on the competency, the skills of the person that are acquiring those images. So if it’s an ultrasound of the right upper quadrant, very classic, right, if it’s the right upper quadrant and radiology, it should be the right upper quadrant in the emergency department so that you identify all of those as relevant and related images. So they this organization was very open to moving towards.

0:6:45.160 –> 0:6:55.450
Petersilge, Cheryl
A centralized study descriptor library, which is really important and then when you start to talk about interoperability among different health care organizations.

0:6:57.710 –> 0:7:25.740
Petersilge, Cheryl
That’s really a nightmare in the imaging space when you want to import them, and they come with poor study descriptors and you need to figure out how that outside study descriptor matches with the study descriptor that you use in your organization, because that’s what you’re hanging. Protocols are all based on, and you’re relevant exams, so it it seems so simple and trivial, but it’s really something that’s very important.

0:7:26.40 –> 0:7:30.310
Petersilge, Cheryl
In enterprise imaging, so they were open to all of this.

0:7:39.430 –> 0:7:39.880
Petersilge, Cheryl
Yes.

0:7:30.860 –> 0:7:50.200
Jordan Cooper
So it sounds like there are recommendations you’re making for naming particular files. It kind of in a taxonomy of how to save and label files that are often very similar. I imagine there might be best practice standards or guidelines issued by the American College of Radiology.

0:7:50.720 –> 0:7:52.930
Petersilge, Cheryl
You would imagine but.

0:7:54.250 –> 0:8:23.560
Petersilge, Cheryl
Not really. So study descriptors or procedure names are something that are developed at the local level now. Radiology has been working for a long, long time on Rad blacks to help standardize procedure names, and then that got rolled into Loinc and currently there is the US CDI, the United States core data set for interoperability that has been developed and that includes.

0:8:23.900 –> 0:8:32.310
Petersilge, Cheryl
Recommended standards that are used for all different kinds of documents that we share now currently.

0:8:33.730 –> 0:8:45.360
Petersilge, Cheryl
The recommendation applies to the radiology report and we we can spend the whole day talking about that, but they have recommended that loinc be the standard that’s used.

0:8:46.50 –> 0:8:47.640
Jordan Cooper
And what is lynx stand for?

0:8:47.730 –> 0:8:51.320
Petersilge, Cheryl
Ohh, I knew you were gonna ask me that. And.

0:8:51.820 –> 0:8:54.490
Jordan Cooper
Or what it how would you describe what it is? It’s sort of listeners know.

0:9:5.680 –> 0:9:5.910
Jordan Cooper
Uh-huh.

0:9:15.340 –> 0:9:15.750
Jordan Cooper
OK.

0:8:53.770 –> 0:9:23.230
Petersilge, Cheryl
So Loin developed as a standard ontology for labelling laboratory studies. So it’s really been a long existing standard in the lab space and with the incorporation of rad blocks from RSA, it now has a standard 4 radiology. It comes from the Reagan St. Institute that’s based in Indianapolis. It’s logical I I.

0:9:23.180 –> 0:9:24.620
Jordan Cooper
Well, that’s OK. Well.

0:9:23.450 –> 0:9:26.970
Petersilge, Cheryl
You know, be really wrong if I try and say what it stands for.

0:9:30.610 –> 0:9:31.230
Petersilge, Cheryl
Sure.

0:9:38.10 –> 0:9:38.480
Petersilge, Cheryl
Yes.

0:9:27.270 –> 0:9:57.380
Jordan Cooper
I’d like to so I just I wanna hone in on the interoperability piece here. So you said that so so you said interoperability is is that there are many difficulties associated with interoperability, but on the one hand, whereas many people will describe interoperability as a technical challenge right and and and that may well be and I’d like to ask you about that, you are describing interoperability as a challenge where we get the files and don’t know what it’s depicting because the descriptors.

0:9:57.670 –> 0:10:0.450
Jordan Cooper
Are a highly variable. Is that correct?

0:10:25.70 –> 0:10:25.440
Jordan Cooper
Umm.

0:10:1.120 –> 0:10:32.590
Petersilge, Cheryl
Absolutely, absolutely. So over here, we’re speaking one language and over here we’re speaking a different language. And so how how, you know, how do you interpret between the two different organizations? Because there really isn’t a standard in radiology. You know, some the basic I would say is maybe the CPT codes, but most organizations have moved well beyond CT of the abdomen with intravenous contrast to.

0:10:32.860 –> 0:10:36.760
Petersilge, Cheryl
CTIV PCT of the kidney.

0:10:36.920 –> 0:10:37.310
Jordan Cooper
Mm-hmm.

0:10:37.880 –> 0:10:40.730
Petersilge, Cheryl
CT, kidney stone. All of those.

0:10:49.770 –> 0:10:50.130
Jordan Cooper
Umm.

0:10:51.860 –> 0:10:52.330
Jordan Cooper
Yeah.

0:10:42.270 –> 0:10:57.550
Petersilge, Cheryl
Different names, but you know, if you’re looking at a CT of the abdomen and you see Arenal mass right now, I got renal thrown in there. You wanna get all those different outside studies that include kidney as anatomy?

0:11:0.30 –> 0:11:0.450
Petersilge, Cheryl
Yeah.

0:11:3.690 –> 0:11:4.400
Petersilge, Cheryl
Sure.

0:10:57.900 –> 0:11:28.630
Jordan Cooper
So I wanna do two things. I wanna be able to quickly soon shift over to the other use case. That’s a really interesting topic, but before we do for our listeners, I want concrete action items, right? So I’m a CIO of a large healthcare delivery system somewhere in the United States. I am listening to this episode right now and I’m becoming aware. Ohh. I wonder what this state is of our standards and I know we have a lot of partners and I wanna adhere to best practices as a CIO.

0:11:28.730 –> 0:11:36.940
Jordan Cooper
What can I do at my organization to help promote interoperability and be a good partner in care in the United States?

0:11:37.810 –> 0:11:50.860
Petersilge, Cheryl
So I would say you try using rad blocks. It’s it’s quite an extensive list so there should be study descriptors in there that meets your needs.

0:11:51.920 –> 0:11:58.50
Petersilge, Cheryl
There’s a huge effort going on in the HIMSS him enterprise imaging community to use anatomy.

0:11:58.760 –> 0:12:29.900
Petersilge, Cheryl
As a core piece of interoperability around study descriptors cause that’s radiologists we we sort of operate around anatomy, but actually if you think about most of medicine and operates around anatomy as well. So make sure that that’s a field in your DICOM header that is well populated. It’s not a mandatory field. And then I would say stay abreast of what’s coming out of the Hensim enterprise imaging community around the body parts standardization.

0:12:39.960 –> 0:12:40.290
Jordan Cooper
Umm.

0:12:30.300 –> 0:12:55.610
Petersilge, Cheryl
And then we’re gonna be taking it to the next level with recommendations around procedure and study descriptor standardization. They just don’t exist yet. But I think the first is just being aware. And you know what, if you’re 10 hospital system and you’re operating out of a single EMR, do you have a single study descriptor for your different radiology department that that could be the first place to start?

0:12:56.230 –> 0:13:3.760
Jordan Cooper
OK. Thank you. I think that is helpful for us for some action items for our listeners now moving on.

0:13:4.510 –> 0:13:5.360
Jordan Cooper
Oh, sorry, yes.

0:13:7.60 –> 0:13:7.550
Jordan Cooper
OK.

0:13:5.10 –> 0:13:13.490
Petersilge, Cheryl
I’m gonna have one more in there. So for those that are including photos, right, so photos.

0:13:14.980 –> 0:13:27.470
Petersilge, Cheryl
Our big component of enterprise imaging and some upload mechanisms allow for free text and you know free text is like.

0:13:28.800 –> 0:13:35.620
Petersilge, Cheryl
The wild Wild West and it is not going to support interoperability, so, you know, be aware of that.

0:13:36.940 –> 0:13:39.790
Petersilge, Cheryl
Component of your enterprise imaging program as well.

0:13:43.210 –> 0:13:43.820
Petersilge, Cheryl
Yes.

0:13:41.90 –> 0:13:46.110
Jordan Cooper
Maybe require discrete fields or a a pick list instead of free text?

0:13:45.200 –> 0:13:47.50
Petersilge, Cheryl
Yes, absolutely.

0:13:49.520 –> 0:13:50.60
Petersilge, Cheryl
Yes.

0:13:48.50 –> 0:14:5.430
Jordan Cooper
OK, thank you. Moving on to the next excellent use case here, you, Cheryl, you have a current client who is a health system working to serve the underprivileged. And I think you have some thoughts about issues you like to highlight with this customer.

0:14:6.210 –> 0:14:12.750
Petersilge, Cheryl
So they are really highly engaged with enterprise imaging.

0:14:14.150 –> 0:14:45.730
Petersilge, Cheryl
They are not, as you know, they they’re not an academic institution and so their choices are different and their resources are different. And so it’s been really great to work with them to help them define their priorities, but also to help them understand how important some of the investments in their imaging ecosystem are in terms of supporting that population. They still need to.

0:14:46.30 –> 0:15:15.970
Petersilge, Cheryl
Consolidate imaging systems. They still need to have standardization. They need to have a robust exchange mechanism when they’re sending patients to higher levels of care or when patients are coming from the Community into their organization. So the .1 of the points here is everyone benefits from an enterprise imaging ecosystem. It’s not just for big.

0:15:26.160 –> 0:15:26.480
Jordan Cooper
Mm-hmm.

0:15:16.50 –> 0:15:47.670
Petersilge, Cheryl
IDN it’s not just for the academic environment, it’s for every organization that’s serving patients. We wanna make sure that the provider has all of the information they need accessible to them and across, not just different points in time but across different geographies. So if you’re seeing over in clinic A, when you go to clinic B, the next time they have access to all of your information.

0:15:56.200 –> 0:15:56.510
Jordan Cooper
Umm.

0:15:48.10 –> 0:15:58.40
Petersilge, Cheryl
And and again, it doesn’t matter what size system you are or what patient population you’re serving, that’s very important.

0:15:59.240 –> 0:16:28.610
Petersilge, Cheryl
We also know that patient engagement is really important and there are many articles out there that show patients being able to visualize their disease process is very, very important for patient engagement. So what does that mean? If you have a patient sitting in your office and you can pull up their CTA of the chest that shows their lung tumor, you’ve now taken this.

0:16:28.720 –> 0:17:1.750
Petersilge, Cheryl
Computer screen which in the past could have been perceived as a barrier in the patient interaction, especially depending on how your office is set up and you now can turn it and show the patient their their tumor their you know what’s going on in their lungs. So now you’ve taken the computer and you’ve actually made it a tool to support the patient physician interaction and you’ve engaged in an activity which we know makes patient more engaged with their disease process.

0:17:1.890 –> 0:17:7.210
Petersilge, Cheryl
And more engaged with the treatment that’s going on. Lots of literature out there on that.

0:17:22.290 –> 0:17:22.830
Petersilge, Cheryl
OK.

0:17:28.160 –> 0:17:29.590
Petersilge, Cheryl
Kept sharp.

0:17:8.370 –> 0:17:38.520
Jordan Cooper
You know, I and I appreciate you delving into that topic with this particular customer. I do have a question that occurs to me, which is kind of out of left field. So you know, take it with a grain of salt, but I would like to touch upon the topic of excessive imaging in America. There is some sense that it may be driven by defensive medicine, the desire to avoid future potential litigation.

0:17:38.870 –> 0:18:10.700
Jordan Cooper
Leads to some providers excessively imaging they think ohh 90% chance there’s no reason to do this. Normally I wouldn’t do it, you know 30 years ago I wouldn’t have done it, but because there’s a 10% chance you know what’s, I’m just gonna order it anyway, just to cover my bases. Also, I’ve heard that recent medical school graduates are more likely to order particular tests and imaging as opposed to just doing a physical and actually touching the patient. Would you speak to this particular issue since we are?

0:18:10.820 –> 0:18:12.420
Jordan Cooper
Talking about imaging in America.

0:18:12.290 –> 0:18:15.700
Petersilge, Cheryl
Sure. I do believe that’s true.

0:18:16.590 –> 0:18:34.840
Petersilge, Cheryl
And I I see a couple things going on. So #1 clinical decision support. So we’ve had the PAMA legislation that’s been on the books for many, many years and I’m not sure if it’s ever actually going to come to fruition, but really.

0:18:36.130 –> 0:19:6.770
Petersilge, Cheryl
It doesn’t matter the form in which it comes, but can we help guide our providers to when imaging is appropriate and when imaging isn’t appropriate and have them accept that guidance and we don’t have that yet. We’ve tried doing its things like hats scams for pulmonary embolism. There are different clinical criteria, different laboratory tests that if the patients don’t meet a certain threshold.

0:19:6.880 –> 0:19:29.740
Petersilge, Cheryl
The chances of a pulmonary embolism are very, very low and can you use that data to deter somebody from performing a CT for pulmonary embolism when the pretest probability is extremely low. So I think that’s one component that can be very helpful and clinical decision support is.

0:19:35.980 –> 0:19:36.370
Jordan Cooper
Mm-hmm.

0:19:30.390 –> 0:19:38.520
Petersilge, Cheryl
It’s a big part of the electronic medical record, so it really happens more at the EMR side than in the imaging ecosystem.

0:19:40.0 –> 0:19:45.980
Petersilge, Cheryl
The other component of excess imaging, which you didn’t really talk about is redundant imaging.

0:19:46.480 –> 0:19:46.700
Jordan Cooper
Umm.

0:19:46.990 –> 0:19:53.270
Petersilge, Cheryl
And so that’s where an imaging ecosystem with a very robust.

0:19:54.810 –> 0:20:27.380
Petersilge, Cheryl
Image sharing among different organizations caring for the patients can really make a difference. So the patient had a CAT scan yesterday afternoon for abdominal pain, and now they’re coming to see you when they’re abdominal pain hasn’t changed. Do they really need repeat imaging or do you just need to be able to see the images they had taken yesterday? And if those images were acquired outside of your system, do you have an easy way to get those images while the patient is still sitting in your office?

0:20:27.780 –> 0:20:43.820
Petersilge, Cheryl
And their the statistics I see there are about 25% of of redundant imaging can be eliminated by actually having access to the outside images. It’s probably higher than that.

0:20:45.550 –> 0:20:45.980
Jordan Cooper
So.

0:20:45.500 –> 0:20:51.920
Petersilge, Cheryl
So that’s another component of excess imaging where I really think enterprise imaging comes into play.

0:20:52.870 –> 0:21:6.210
Jordan Cooper
I am very glad that you articulated that additional criteria for potentially unnecessary imaging and as we approach the end of this podcast episode, I’d like to ask you a final question.

0:21:7.370 –> 0:21:11.70
Jordan Cooper
On this very topic, which is Cheryl, you know.

0:21:11.730 –> 0:21:33.390
Jordan Cooper
There are. There are. What is the? How should a CIO listening to this podcast episode proceed? What needs to happen in enterprise imaging in order to facilitate sharing of different images? What what steps need to be taken to promote interoperability of these images in order to reduce redundant imaging?

0:21:48.660 –> 0:21:48.990
Jordan Cooper
Mm-hmm.

0:21:59.290 –> 0:21:59.660
Jordan Cooper
Mm-hmm.

0:21:34.150 –> 0:22:7.640
Petersilge, Cheryl
So there I think you need to look at your image exchange system and your image exchange strategy. So a lot of times organizations will tell me we use XY and Z vendor for our image exchange, but it’s only to these three organizations and then the other times patients need images, they have to have them on CD’s. So it’s understanding who you’re sharing images with and under what circumstances are you sharing those images.

0:22:8.110 –> 0:22:13.370
Petersilge, Cheryl
Is your vendor meeting all of your different use cases?

0:22:13.880 –> 0:22:14.180
Jordan Cooper
Umm.

0:22:20.680 –> 0:22:21.30
Jordan Cooper
Umm.

0:22:14.850 –> 0:22:25.180
Petersilge, Cheryl
And then my dream and I haven’t yet seen it come to fruition, but I’ve seen organizations move in that direction is to have.

0:22:27.40 –> 0:22:29.520
Petersilge, Cheryl
Every organization gets to have one.

0:22:30.750 –> 0:22:40.300
Petersilge, Cheryl
Image exchange vendor and those vendors all communicate with each other, so we’re not managing 10 different gateways within the organization.

0:22:41.380 –> 0:22:47.270
Petersilge, Cheryl
Every organization has to have an electronic image exchange mechanism.

0:22:48.230 –> 0:22:48.970
Petersilge, Cheryl
And.

0:22:50.260 –> 0:23:15.570
Petersilge, Cheryl
That one vendor that you have communicates with your centralized archives, so you have one point of contact for all of your images. You know, I’ve asked to ophthalmology departments if somebody needs to take their images somewhere else. I don’t know. I think we have to call the Secretary and you know, hopefully the right ones on and they make a CD and then we burn it and we mail it out. Well, if everything goes to your V and a.

0:23:16.180 –> 0:23:16.510
Jordan Cooper
Umm.

0:23:16.350 –> 0:23:20.140
Petersilge, Cheryl
You’ve gotta process. It’s easily replicated.

0:23:27.360 –> 0:23:27.730
Jordan Cooper
Umm.

0:23:21.360 –> 0:23:28.10
Petersilge, Cheryl
And it eliminates all that intermediate physical media so.

0:23:28.650 –> 0:23:34.910
Petersilge, Cheryl
Just like you have CHIME that manages all the release of information from your EMR.

0:23:35.190 –> 0:23:35.530
Jordan Cooper
Umm.

0:23:36.630 –> 0:23:44.270
Petersilge, Cheryl
1 centralized place, maybe CHIME that manages the release of all of your imaging data.

0:23:45.80 –> 0:23:46.610
Petersilge, Cheryl
It should be just as easy.

0:23:47.590 –> 0:23:57.290
Jordan Cooper
Thank you. Well, Cheryl, we’ve covered a lot of ground. We’ve talked about good governance, selecting a vendor partner we’ve discussed.

0:23:58.530 –> 0:24:28.920
Jordan Cooper
The standardization of naming files about addressing underprivileged and about promoting interoperability and overcoming barriers to sharing images, so I really do appreciate you joining us today for our listeners has been Cheryl Petersilge, the founder and CEO of Vidagos Enterprise imaging strategy consulting and former medical director of integrated content and enterprise Imaging at Cleveland Clinic. Cheryl, thank you very much for joining us today.

0:24:29.950 –> 0:24:31.950
Petersilge, Cheryl
Thank you for having me, Jordan.