Tuesday, November 28, 2017

Think Enterprise: Looking beyond radiology

Amy Vreeland is a Boston-based healthcare technology consultant specializing in helping health systems and health IT startups develop and implement imaging strategies. She has been in healthcare informatics for 25 years and for the last 17 has been focused on imaging. In 2014 Vreeland founded Imaging Strategies focusing on consulting with both healthcare startups bringing new technologies for image exchange and telehealth to market, and with healthcare systems to help them look at imaging from an enterprise perspective and develop strategies and governance models to manage it. Vreeland previously co-founded Life Image, a cloud-based image exchange platform and served as Sr. VP for account management at AMICAS one of the first web-based PACS systems (it was acquired by Merge Healthcare). Last year she led a joint SIIM-HiMSS(I’m not sure HIMSS uses the lower case when spelling out the name – only in their logo) work group that produced a white paper on image exchange published in the Journal of Digital Imaging.

Why did your work group write the paper? 

I'm part of a HIMSS and SIIM workgroup that deals with the emergence of enterprise imaging, (or EI), which has radiology imaging at the core, but also includes all types of clinical imaging. It includes workflow to capture, view, and store cardiology images, GI images, photographs and evidence-based imaging, even pathology images – pretty much every type of clinical image you can think of. Because of things like smart phone cameras and low-cost ultrasound devices, imaging has been springing up in all sorts of departments in hospitals. But a lot of that growth has been off of the radar of either radiology or IT, so a lot of it isn’t really being managed by either organization.

Enterprise Imaging allows health systems to thing about imaging more broadly, and more strategically, kind of like as the imaging cousin of the EHR. It is only recently that organizations have started to look at imaging holistically, and that vendors have started to provide end-to-end imaging solutions.

The EI workgroup started to help health systems think about imaging from this new perspective. We wrote our specific paper about Image Exchange to describe use cases for all types of physicians and service lines at health systems – to help them think about image exchange holistically, and to describe emerging standards for image sharing.

Because of my experience in the imaging exchange space (with lifeIMAGE) I was a candidate to lead the workgroup and I was lucky to be able to work closely with Ken Persons from the Mayo Clinic on it, and with a number of other great folks.

Have you seen an increasing interest from organizations in developing their imaging exchange strategy?

Without a doubt. With more organizations working under value-based contracts, they are tuned into cutting out costs of unnecessary redundant imaging. They are also working harder to improve the quality of their interpretations by making sure providers have sufficient patient imaging history to do a thorough interpretation - be it having all of a patient’s relevant priors for a new breast imaging or lung screening exam. So, there are clinical and business requirements that are driving organizations to work harder to get patient imaging histories that happened outside of their facility.

And – with all of the recent buzz about machine learning for imaging – organizations are seeing even more reasons to be better able to move images around.

Are you seeing that demand grow?
Yes. As organizations do the needed diligence to develop their enterprise imaging strategies, they are finding more and more really practical use cases for image exchange. It may be the pediatric cardiologists who are delaying outside consults, or oncologists who delay getting patients reviewed at tumor boards. Organizations are finding more and more areas where clinicians are struggling to get access to outside images and to share local images with referrers, where providers are wasting time and delaying consults or care.

I think organizations quickly understand the value of image exchange, and are trying to work out how to make it happen broadly at their organizations, and how to foster wide adoption of it. Image exchange vendors started providing solutions before there was a valid and practical standard for sharing images. So, they built proprietary solutions. As time has passed, the vendors have started to modernize their solutions with a standards underbelly to them, with XDS. Though XDS may be built into their solutions, it isn’t clear to me that a lot of organizations are doing exchange using their vendor’s XDS capabilities quite yet, however.

So, organizations have to work harder than they should to get adoption of image sharing. The problem is that, because the solutions are proprietary, in order to exchange with one hospital, one vendors solutions may be required. And to share images with a different hospital, another vendor’s proprietary solution is needed. Each may have a different workflow. It is complicated for sites to keep track of these different workflows and it really inhibits adoption. So, one of the purposes of the paper was to educate the community around the emerging standards and to suggest that they evangelize the standards, and push their vendors to get better about interoperability.

In addition to adopting technology solutions for doing imaging exchange, are you seeing common challenges with adoption?

More organizations are adopting imaging technology, but not all of them are looking at it holistically. I think they often look at it (technology) as a way to deal with radiology imaging alone, when an example of a more real, pressing problem they should be working to solve is specialists having access to the appropriate outside imaging for referrals, for instance. That means cardiologists and cardiology practices should be involved, and neurology and orthopedic practices, and while I find some organizations address one or two of those, they really aren't looking at it as a quality improvement across all service lines in a way that really rewards them and provides the value that really is available. It is because it is complicated to change workflows in so many departments, and because standards aren't there it is just hard work. 

Organizations have had varying levels of success. I think one of the bigger issues preventing broader success (besides interoperability) is that the mandate from executive leadership might not be clear enough or the executives might not understand the problem well enough to be working toward making this (holistic, enterprise-focused strategy) happen.

Lots of organizations are approaching image exchange from an IT perspective rather than a clinical or business perspective. They may be thinking in terms of getting technology implemented, but not strategically enough along the lines of "how can we leverage the technology to cut costs or improve care.’ Well implemented, information exchange, and particularly image exchange can have a huge impact on the (IHI’s) Triple Aim (recently changed to Quadruple Aim). The whole provider satisfaction issue has been added to the conversation as the fourth tenet on top of the existing Triple Aim, and image exchange can play a big role there. Doctors are dealing with CDs still, which is hard to fathom in 2017, but they are. It is often a big nuisance and a big negative part of their lives, or worse yet, because they don't have the ability to seamlessly exchange information they are unable to weigh in on consults that could lead to high value patient transfers. So, there is a lot of work to be done.

Do PACS administrators have a role in advancing the idea and value of image exchange?
I think PACS administrators are in a difficult position. They may be limited to radiology and often don't have any visibility into the strategic objectives of their organizations or access to leaders outside of radiology. Yes, any motivated person in an organization who is able to educate up and manage up has a role in evangelizing the potential of any technology, but the strategic place they (PACS administrators) are in at a hospital can make it hard to succeed at that. But the more they understand and the more they want to further their careers, I think raising the issues and educating the organization about, ‘you know this could add a lot of value to some of our high-end service lines,’ is always valuable. But sometimes they don't get the right audience. That doesn't mean they shouldn't be pushing it and educating themselves about it. The more they understand all of the places such technology can add value, the more valuable they are as a team member.

The second broad issue is patient matching and the complexities around that. Who should have the major responsibility in developing strategies for improving that?

As we all know, patient matching is a challenge for all types of data exchange – not just imaging data. So EHRs and HIE solutions are working to address that, and I don’t think the imaging vendors should have to reinvent the wheel completely.

But I think we are far away from HIEs being a reality of everyday life. So, there are a few things that can really help with accurate matching of patients to outside images right now. The most important is having image exchange capabilities embedded right in the EHR, so that a lot of image transfers can happen within the context of a specific patient's record, whether it is check-in clerk uploading a CD into a patient record to normalize the patient demographics embedded in the image or someone accepting a transfer via the EHR, the frontline person who is validating the patient is then validating that the outside imaging is indeed for the right patient.

Of course, with image exchange, matching the patient is half of the equation. It is also important to match order information (which may require that an order be generated) so that exam descriptions for outside images have some consistency with locally acquired exams, so that the hanging protocols display images correctly to the radiologists. It is a much simpler problem in comparison, but still needs to be considered.

Is the rise in visible light imaging generated from mobile phones adding to the complexity?
These types of images are a really important part of the conversation. Enterprise imaging solutions provide mobile apps that let the clinician pick a patient name from an EHR patient list, take a photo, and correctly label the images. These apps also prevent the images from being stored on the mobile phone, which is obviously vital for HIPAA compliance purposes. Some of the image exchange vendors do that, and some of the enterprise-imaging vendors do that. 

Physicians are taking these photos in the name of patient care, which is a good thing, but they need the tools to do it in a way that doesn't create compliance or security risk. Some organization executives assume patient photos are being captured in a HIPAA compliant way or they know they're not and are just closing their eyes and hoping no one notices. I can’t tell you how many physicians have shown me patients photos stored on their personal phones that had patient identifiers visible, and even showed details of sexual of domestic violence, or child abuse. Photos like these legally (and ethically) need to be handled appropriately – and hospitals commonly have liability and security risks all over the place as a result of this type of use for mobile phones/photos.

I think that is an important area for PACS admins to sound the alarm, to manage up in the organization to say ‘we have a huge HIPAA compliance risk and a huge security risk here’ if that is indeed the case at their site.

Are there simple solutions for getting those images off the phone into patient record in a secure manner, or is that a problem that needs to be solved?
No many vendors provide apps that do just that. They can capture the image in a patient context and transport it to the EHR and enterprise imaging archive, which is optimal, and the image is never stored on the phone. Those solutions exist and they exist from a number of vendors and they can be an add-on technology. You don't need to buy and end-to-end enterprise imaging platform, there are faster ways to do that in more modular ways.

How is the technology of interoperability progressing?

There are improvements being made on interoperability, but some of the areas where interoperability matters are within the hospital rather than between health systems. So right now using legacy systems, radiologists commonly don't look into the EHR to get relevant patient information. They don’t do that because diagnostic workstations don’t interoperate easily with EHR systems. They can’t easily launch the EHR. They have to leave the workstation, sign into another system, search for the patient, and then search for the imaging order, and for other information that is relevant to the patient’s condition. It is just too time consuming, so too often, they don’t look. 

Physicians who order imaging exams often put poor or unclear information in (the order) as a reason for the exam, so that the reason the patient is to be imaged doesn’t help the radiologist. A lot of imaging happens where the radiologist is interpreting the exam but she doesn't know what she is supposed to be looking for. Or she doesn't understand the patient context.

As a result much of the imaging that happens and the reporting that is done is not on point, because the interpreting provider isn't looking for the right thing because they don't have adequate patient context or complaint information. This is a super important area of interoperability, being able to launch into the EHR from the PACS workstation within the patient context, in the appropriate encounter, with other clinically relevant information (like lab results) readily available.

I think that is an enormous area where vendors have made progress on interoperability, and this type of integration has become easier, possibly through simple FHIR-based calls. But organizations need to catch up with implementing such interoperability, because for a little bit of time and effort, they can quickly accelerate a radiologist's efficiency and accuracy, so that is one very important area of interoperability. 

Another (area of interoperability) is within an organization having a modern enough workstation that is able to query and look at not just a single PACS imaging archive, but able to look at any number of archives that are federated. What I mean here is that modern workstations can look at a radiology PACS archive, a cardiology archive, and an array of visible light imaging, etc. So a radiologist should be able to look at any cardiology exam that is relevant, just as a cardiologist should be able to review a radiology exam that is relevant. That type of interoperability exists now and hopefully should be adopted quickly. That would provide a huge impact on patient quality of care.

Then there is inter-organization interoperability that we’ve already discussed. I think those are all different areas that a PACS administrator should be thinking of. People use the word interoperability liberally, magically hoping that all systems talk nicely. I think it is important to break it down into those three types of interoperability that might be able to be achieved in incremental ways. FHIR-based interoperability between the workstation and the EHR, and workstation-based interoperability looking at federated imaging archives, and inter-organization interoperability for image exchange between facilities.

Are you saying doctors and organizations haven't implemented interoperability as fully as is now possible?

Integration hasn't been easy, but the technology exists to do that in a pretty lightweight way using FHIR for instance. It may not be ubiquitously available but it does exist. And some of the newer machine learning technology that let radiologists not just launch the EHR in the patient context, but also with appropriate and relevant elements of the patient record highlighted based on the type of imaging they're looking at. So, some of it exists, and some of it is emerging. But it is something PACS administrators, as they do their diligence, should be learning about. It is an enormous clinical value-add and in my mind falls into the category of lower-hanging fruit than say enterprise-wide initiatives for imaging exchange.

Is there anything more you'd like to emphasize?

Think enterprise. I think it is important for PACS admins to be looking more broadly at imaging outside of radiology. Doing what they can to educate themselves. Their role can be elevated and be far more impactful if they look outside of radiology. There is a huge job protection value for themselves, but also adding value to the enterprise overall because no one in the organization tends to own all types of imaging, and these types of silo-ed imaging has popped up all over the place with the advent of low-cost portable ultrasound for example, it is common that people don't know on an executive level that they may have liability and security risks, and PACS admins can serve their role well by learning more about that and working to educate management about it.

Are PACS admins being absorbed into IT and if so would that give them greater visibility?
Certainly I think that is more commonly happening, they might reside in a radiology area but under a bigger IT umbrella. I think they would be more effective there because they can take the whole idea of image exchange outside of just radiology and apply it from a business perspective to all service lines.

No comments:

Post a Comment