Adopting Digital Pathology on a Budget: Getting Started, Knowing What’s Feasible, and Funding Your DP from Overlooked Sources

Expert Presenters

Keith Kaplan, MD
Chief Medical Officer
Corista
Concord, MA

Keith J. Kaplan, MD, is CMO of Corista, a digital clinical pathology workflow platform that aims to surpass traditional pathology workflow in speed, accuracy, and ease of use. He is a member of the College of American Pathologists, American Society of Clinical Pathology, and American Society of Cytopathology and an executive board member of the American Pathology Foundation. Before Corista, Dr. Kaplan was an Associate Professor of Pathology at Mayo Medical School and was Pathology Informatics Chair as CIO of Pathology Practice at Carolinas Health Systems in Charlotte, NC. He is a graduate of Michigan State University and Northwestern University’s Feinberg School of Medicine, completed his residency training in anatomic and clinical pathology at Walter Reed Army Medical Center, Washington, DC, and is board certified in anatomic and clinical pathology.

Andrew Evans, MD
Medical Director of Laboratory Medicine
Mackenzie Health (MH)
Richmond Hill, ON

Dr. Andrew Evans is the Medical Director of Laboratory Medicine at Mackenzie Health (MH), a large community hospital in North Toronto that is currently transitioning to complete reporting by digital pathology. Before moving to MH in 2020, he was the director of digital pathology at University Health Network (UHN) from the inception of the program in 2004. UHN successfully implemented whole slide imaging (WSI) for frozen sections, consultation, and primary diagnosis at various partner sites. For the past 10 years, he has been extensively involved in developing guidelines and best practice documents for digital pathology through his work with the College of American Pathologists (CAP). He currently chairs the CAP Digital and Computational Pathology Committee and WSI Validation Guideline expert panel.

William N. DeSalvo III, BS HTL(ASCP)
Anatomic Pathology System Production Manager, Sonora Quest Laboratories and President, Collaborative Advantage Consulting LLC
Phoenix, AZ

William N. DeSalvo III has 40+ years of experience in the anatomic pathology laboratory industry. Working as a histotechnologist, clinical histology laboratory consultant, and product/marketing manager for histology and immunochemistry, DeSalvo brings a well-rounded background to help bring organizations to the next level of digital pathology. He earned a Bachelor of Science in Chemistry (minor in Biology) from Southeast Missouri State University, developed a Quality Management System for anatomic pathology laboratories, and received Six Sigma and Lean methodologies training. Over the past 15 years, he has used his education and expertise to provide educational presentations and publish various articles on process improvement, standardization, and automation.

Lisa-Jean Clifford
COO and Chief Strategy Officer
Gestalt Diagnostics
Spokane, WA

Lisa-Jean Clifford is COO and Chief Strategy Officer at Gestalt Diagnostics, a company focused on digital pathology solutions and services. She has held marketing and business development roles at leading healthcare solution vendors, including McKesson and IDX (GE Healthcare), International Data Group (IDG), and eBusiness Technologies. Clifford’s expertise includes strategy development and execution, general business administration and operations, marketing, business development, and product management. She contributes her industry expertise in publications including Advance, Medical Laboratory Observer, Clinical Lab Products, CAP Today, Health Data Management, Health Management Technology, and Forbes Magazine. Clifford also authored a book on XML and has presented educational and thought leadership sessions at healthcare industry conferences. Clifford holds Bachelor of Science degrees in Business Management and Marketing from Johnson & Wales University.

Webinar Transcript

Andrew Evans (00:06:52):

Okay. Thanks very much, Keith. Then I’d like to thank also the organizers, the dark intelligence group for inviting me to participate today. Next slide please. So what I’m going to do is to try and give a very practical overview of the dollars and cents of implementing digital pathology and from my own personal experience at doing so in a large academic center in Toronto versus a community hospital where I’m located now. So the large academic center where I practiced for close to 20 years as the university health network, where we use digital pathology for a number of niche applications and where I am practicing now the institution before I came, had made a decision to make a strategic investment, to go completely digital. And I’ll go into some details on that as well. So I think my, my take home message at the end of this is that with respect to digital pathology on a budget is that digital pathology does not have to be an all or nothing proposition that you can use it for just selected applications as we did at UHN or you can go for the whole enchilada and try and digitize everything like we’ve done here at McKenzie health.

So this is the this slide summarizes really everything that we did with respect to digital at UHN. And this was in initial vision of the program medical director at the time Dr. Sylvia Asa to begin with implementing slide or a digital pathology to cover frozen sections at our Toronto Western location. And I’ll explain more about the, the geography of UHN and the next slide. Then after an early win there, we transitioned into transplants using the technology to cover transplant biopsies that are frequently came after hours, particularly for liver and kidney. We then transitioned into doing forming partnerships. So digital was a big, big enabler for supporting partnerships which can further help the fray or spread out the cost of implementation when you have several different institutions participating.

Andrew Evans (00:09:07):

And so we use that for frozen section support in a couple of locations that I’ll describe this paved the way for a multi-jurisdictional telethon network involving provinces of Manitoba and Newfoundland and Labrador, and with UHN and the center hub. We used it to support some consultation contracts with hospitals in the middle east, and then finally for a primary sign out in, in a limited role with a regional cancer center with which we had partnered back in 2012. So excellent, and is the lab medicine program, there operates on a subspecialty pathology model. And as I mentioned, there are multiple sites. And so to provide sub-specialty coverage across multiple sites, then you’re confronted with either moving slides, moving pathologists are going digital. Next line point was back in 2000 2004 when we kicked things off.

Andrew Evans (00:10:08):

And that was with neuropathology frozen sections at our Toronto Western site. So in 2006, the pathology department at UHN was fully consolidated at the red star and the upper right corner of the map at Toronto general. And at the other end of the double-headed arrow is Toronto Western site where which is home to the crumbled neurosciences center. And it’s where all of the neurosurgery is done at UHN. And there was a steady flow of frozen sections, not an overwhelming number, roughly anywhere from two to about 10 or 15 per week that would require a pathologist to run back and forth to read those frozen sections. So again, this was this theme of moving slides, moving pathologists are going digital, and we opted to go digital as per the, you know, the vision of Dr. Hasan next line.

Andrew Evans (00:11:04):

And Keith will remember this very well after we met at the AP triple line meeting in Pittsburgh in 2003. He was kind enough to pay us a visit in Toronto to tell us what he was doing with respect to the use of a trestle robotic microscope at the Walter Reed in in Washington and provide some advice on how we can get started. So we actually started with a, like a robotic microscope, which was a capture store and forward system which worked actually worked extremely well. There was no issue with image quality. The problem was it was slow. So we used that for about two years. And I know by slow, it could take up to 10 minutes to review an individual slide, next slide.

Andrew Evans (00:11:48):

So by the time we got to two, 2006, we implemented an, a perio scanner and went to whole slide imaging with a very quick validation study. And it was immediately apparent that this would better replicate the microscope experience compared to the robotic microscope we were using before. And so in my time at UHN, we had looked at over 6,000 frozen sections. Most of them coming from neurosurgery, and then we can see the parameters here in terms of performance discrepancy rates with finals, deferral rates, and total turnaround times where we’re very much on par with what you get in an analog glass slide workflow. So we had no issues there. Okay.

Andrew Evans (00:12:32):

So in terms of the dollars to get started with that initial application it was roughly 150,000 Canadian dollars, and that was to purchase a demo scanner. So we did get a break on the, on the device there, and then the other costs up front were purchasing, installing the scanner training of staff procuring, dedicated it support. We found very quickly that, you know, that this application was not a help desk type of situation where you’re calling somebody who’s not familiar with what you’re doing. If you need technical support we needed a dedicated it support. And then of course, the ubiquitous service contracts and upgrades of roughly 15,000 per year. So that was the cost to get going. Next slide, the sense of things it’s difficult to assign costs. So we often hear about soft costs in in in digital pathology and being counters and institutions don’t necessarily appreciate soft costs, or even like to discuss them.

Andrew Evans 00:13:30):

But the efficiency and quality benefits for this application were clear. It just obviated the need for a single pathologist to travel back and forth, which was inefficient disruptive to their regular workflow and offered no possibility for consultation on difficult cases. So that’s really the sense of this. Next slide, early after the early wind of the Toronto Western application we then switched to providing frozen section support to a hospital that’s roughly 425 miles north of Toronto by driving 350 miles by air in Timmins, Ontario, and the value proposition for digital there is that one week out of every month, there would be no pathologist on site to provide frozen section support to the surgeons there. And then the other site is Kingston general hospital. It’s affiliated with Queens university in Kingston. So next slide, please, as I mentioned, the in terms of the dollars for Timmins and district hospitals.

Andrew Evans (00:14:31):

So there was a part of a bulk purchase UHN and the partner sites procured seven five slide perio scanners. So then the cost per unit came down. So there was some efficiency realized there. And we set up two scanners because of the distance involved. You needed some type of a backup system if one system went down, so there was some necessary redundancy there. And then of course, service contracts and upgrades. And the sense of this, as I mentioned, was that the Timmins and district hospital surgeons would now get continuous frozen section support when the one pathologist on-site pathologist there was off at another outfit, other sites so this kept continuity of care going. And then we were actually able to leverage those scanners for other purposes as shown on this slide. And we’ve since actually that partnership is still ongoing and they are having marked on some primary sign-out now for routine cases, particularly GI cases, next line, moving to the project with Kingston.

Andrew Evans (00:15:37):

This is an academic pathology department, and it was a very specific niche application there, which was neuropath frozen sections. They had a volume of roughly one to five per week with only one neuropathologist. And when this person was away the other pathologist felt the need for some backup coverage. And so the enter UHN and the digital pathology systems. So we were given remote access to their EPR, to their diagnostic imaging limited consultant credentials, had everything ready to go to provide support just as we did for frozen sections at our Toronto Western site next line.

Andrew Evans (00:16:15):

Of the dollars and cents here, this was a definite win for UHN. There was no upfront cost for, for the organization. Kingston had their own scanner, they took care of their service contracts and their upgrade process, et cetera. In the sense of this was that they Kingston general pathologists or and neurosurgeons had some UHN support for neuropath frozen sections. And the UHN pathologist actually made some money out of it because we build them on a case-by-case basis. So next line, then we moved to a primary diagnosis and what that was with involving the Lakeridge health center in in Oshawa. And that’s roughly 40 miles to the east of Toronto general which is located in the downtown of Toronto and as noted, it’s a regional cancer center. And it was Mo when UHN took over the medical leadership of the lab at Lakeridge health, we were transitioning into a subspecialty reporting model. And that meant that roughly 300 slides per day, up to 300 slides per day from particularly from GCU liver, endocrine and head and neck cases had to be transported to UHN. So again, move pathologists, move slides, or look at digital next slide, please.

Andrew Evans (00:17:30):

So when we did that, in terms of the dollars, this is Lakeridge health scanner so they said spend $250,000 Canadian purchase install, a trained staff, dedicated it support, et cetera. The usual service contracts there were some special applications made for storage. So I’m extra costs that were incurred there slide scanning, personnel, et cetera. And the interesting thing there is that that project we actually leveraged at administrative assistance in the in the pathology office who were trained on how to scan slides, run the scanner. And they worked in close collaboration with the lab manager and histo technologists to spot problems with respect to histology and image quality. So, so the annual cost, did you balance the cost of implementation against the cost of transporting those slides by courier? It was roughly $35,000 for a standing order for a daily courier run up 1:00 PM.

Andrew Evans (00:18:31):

If glass was shipped, there were ad hoc couriers that were, were invoked for rush cases. There was the cost of clerical staff packing and scanning and scanning out slides at both ends, both at lake region, Toronto general side. And this was an interesting one. The lab manager at lake health told me that they spent roughly a thousand dollars a year on slight folders that left lake Ridge health and never returned. So if you tally all that up at a minimum, it was close to $80,000 that was spent to ship the slides back and forth. And I should point out that the couriers that we’re running on a daily basis also shipped other things. So these weren’t specific for digital pathology and the scanner and image and archiving costs, we felt could be recouped in two to three years by not having to ship all the glass.

Andrew Evans (00:19:18):

So we felt that it just didn’t make sense to ship the glass when we had WSI as a validated alternative next, in terms of the workflow perspectives, if you can take the traditional approach of lab medicine and look at pre-analytic analytic and post analytic components the lab and admin staff at Lakeridge health liked it because it was faster, cheaper, and easier and avoided having to pack up slides and meet courier deadlines, et cetera. The post analytics side, the clinic staff, they liked it because the follow-up appointments were kept turnaround times could be could be could be tighter when you didn’t have to ship glass slides. Now with pathologists, there’s no happy face there because there was there was no consensus as to across the group. So pathologists who were, you know, early adopters and ready adopters of the technology said, find this works for me, others felt it slowed them down, disrupted their schedule. And basically what’s in it for me. Why are we doing this? So next slide.

Andrew Evans (00:20:22):

So, okay. And then we come to Mackenzie health where I am currently located a large community hospital, and we currently digitize every single slide that the lab generates using two Phillips ultra fast scanners. Next slide. The health is a, now a two site hospital. When I arrived here, back in, in August of 2020 we were everything was operated through the Richmond hill site. And I’ll show you in a map and a couple of slides where these two hospitals are located and everything was run through the Richmond hill site, but the idea was that it would transition to a two site model. So the corridor Luci Vaughan hospital, which is where I’m sitting now and that’s the hospital on the bottom, right? And so over the next three to five years, our volumes will double surgical volumes will double and we can and a strategic decision was made to go digital to support this to site model.

Andrew Evans (00:21:27):

So next slide, please, here’s the, this slide summarizes the discussion that was held internally and this took place before I joined the group. So it was essentially balancing a strategic investment versus a return on investment approach. So the return on investment approach, that traditional thing of do, I recoup all my investment within three to five years, how do I generate revenue, et cetera. The leadership at McKenzie health decided that this was the wrong way to look at it, that there really was no traditional business case for digital pathology and their view and the, they looked at it, for example, looking at a CT scanner and an MRI machine is that there’s no return on the purchase of those of that equipment as opposed to a strategic investment, recognizing that the future of pathology would be digital retaining workforce through flexibility increasing quality to patients enhancing collaboration between pathologist, particularly between pathologists in different institutions by reaching subspecialty experts.

Andrew Evans (00:22:31):

And of course the emerging concept of machine learning and artificial intelligence. So organizations should expect based on this, this exercise at McKenzie health to budget, approximately four to $5 million for a mid range digital pathology solution. And I shouldn’t find out that McKenzie health had the luxury of being able to leverage a long-term diagnostic imaging contract with Phillips to piggyback on me and the implementation of the digital pathology solution that we’re using. Now, next slide, please, in terms of the journey discussions initially started back in 2018, with respect to concepts of digital pathology discussions with internal stakeholders, the system was procured and set up in 2019 and initial validation process was done. And then by 2020 last August, it was ready to go alive where pathologists had we’d get glass slides and digital slides together and would work at their own pace to transition towards digital next slide, please.

Andrew Evans (00:23:37):

And I think it’s and quite remarkable. So since August of 2020, now, seven of seven McKenzie health use whole slide imaging for primary diagnosis on most to all of their cases for six out of seven of us no glass slides aren’t delivered and that kicked in as of may 10 2021. So just very recently, and that with the proviso, that glass slides are always available for specific cases on demand. So nobody is discouraged from using glass slides and we still very much use microscopes. And we’re now starting to collect data just from the LIS on why pathologist defer to glass and how often they do and what types of cases are involved. I should point out a huge driver for adoption of digital at McKenzie health had been the implementation of home workstations, which was very valuable during the pandemic. And we were able to do this develop to implement validated verified performance verified home workstations for just over $2,500 Canadian per workstation. So I’m happy to speak during the questions about my views on the lack of a need of medical grade, $12,000 monitors to use digital use for digital pathology. And then inally, we’ve implemented a video microscope for frozen sections that are Richmond hill site recognizing all of the following lists are now either at home or working from home or at the Vaughn site. Next slide, please.

This is a map which shows where the two are two sites. This is the Richmond hill site, the original site from Mackenzie health. And this is the new site in Vaughan where I’m located, it’s roughly six miles apart. And then we’ve been starting to work these other triangles here. We’ve been working with a group of pathologists in north Toronto, east Toronto, Midtown, and downtown, both academic and community to try and establish a digital pathology network to really leverage the tool that we’ve got here at McKenzie. Next slide please. So this has been dubbed the greater Toronto area, digital pathology network it’s currently in its early planning stages of present. The, the must haves for all the participants was a scanner agnostic solution needed to be flexible in terms of its applications so that not everybody was going to be scanning every single slide that would, could be used for a variety of different applications.

Andrew Evans (00:26:06):

And it was very important that autonomy for each institution was preserved. So generally where we’re at, I think Keith alluded to it. So discussions about digital pathology being unproven and very risky et cetera, I have largely dropped away. The participants are now pretty much on the same page that this would be an excellent idea, a really good idea, but the choke point now is how to pay for it. So in terms of local implementations, that it will likely come down to each, each individual institution using capital budgets and foundation funding. And then the network connectivity piece, we’d be able to approach a government innovation table and come forward with some in kind funding to try and make piece this all together. But there’s a lot of work ahead there. And I think this is actually, this is my last slide.

Andrew Evans (00:26:58):

It’s a nice point to leave off, off on to introduce our next speaker. Lisa Jean, who will talk about the procurement process and provide some very valuable information for us. So in summary, I think hopefully I’ve left you with the idea that digital pathology does not have to be an all or nothing proposition, and that you don’t have to digitize every single slide in your institution to use the technology. You can pick it and use it for niche applications, as we did at UHN establishing some early wins and securing funding for success of applications using partner sites to lessen the funding burden. I think the big challenge at UHN is that, you know, being such a large organization, there are many, many competing priorities. And so it made complete digitalization, very difficult there. And I know they’re still looking at a way to, to try and do that. Whereas McKenzie health, they decided to make the strategic investment for complete digital pathology. And we’ve been able to transition in the, just inside really eight, eight months to more or less complete digital reporting. And I think my big takeaway here is at a small institution, it’s much, much easier to get this done. So without all staff, I hope this was useful and I’ll pass it on to Lisa-Jean.

Lisa-Jean Clifford (00:28:19):

Thank you, Dr. Evans. I have put together several slides in my section of the presentation that you can feel free to go back to and revisit and use as guides or information that might be helpful to you in deciding how you’re going to fund and adopt digital pathology on a budget. So I won’t be going through all of my slides in extreme detail. This is one in particular that I am listing here, all of the hardware components and software components for you to consider as you’re looking to adopt your full blown digital pathology solution. So again, we’re going to look at how you can start small or find sources of funding on a budget to be able to implement digital pathology and a stepping stone approach, so that you’re able to start walking down that instead of going all the way to B end, this will give you a foundation to be able to add all of these components in the future when you get to that whole week grail or the full deployment.

Lisa-Jean Clifford (00:29:34):

So when we look at the different components, again, focusing on hardware, there are certainly things that you want to consider when you consider both your budget and the features and functionality that you’re looking to be able to acquire. So you’re not looking to make an investment in the short term because of budget constraints that you’re not able to leverage longer term. And some of those things that you’ll want to consider are the number of slides to be scanned. So you’re looking at your volumes and you’re looking at the speed of the scanners. You’re also looking at the quality of the image and the QC process that the scanner provides. So is the QC process manual because you’re investing in a lesser expensive model, or is it an automated process where you’re looking to make a little bit more of an investment in a higher quality or a higher end standard to be able to leverage that as part of your foundation. And again, what you’re doing is you’re picking and choosing, weighing the features and functionality that are available. Short-Term on a smaller budget versus building that foundation for what can be a great foundation and, and stepping stone for expansion in the future. So again, looking at the number of slides to be scanned, you’re really looking at your hourly daily and weekly rates at this point, any further than that probably isn’t as impactful or as important in your initial decisions. Next slide please.

Lisa-Jean Clifford (00:31:09):

And then the same holds true for the digital pathology software solutions. So this is really key four points to consider when looking to adopt digital pathology, software solution interoperability and universal capabilities. And when I say universal, what I really mean is vendor agnostic. These are the two most important components in selecting digital pathology solution because these two things really provide the stepping stones and that foundation for you to be able to truly take advantage of the advantages and capabilities that digital pathology really offers you, not only in the short term, but also longer term. So by selecting a digital pathology solution provider, who has the capability of being interoperable you’re really opening up the use case, not only internally for your organization, but really for expanding the digital reach beyond your geography and beyond your parking walls. So the same thing with universal capabilities or being vendor agnostic, again, being able to integrate with different scanner vendors, with supporting the different image file formats being able to do multiple different ones simultaneously, and the same thing for your it and laboratory applications, being able to integrate with multiple different IOT applications that you’ve already invested in and deployed within your organization will truly help streamline that digital pathology process and the workflow for the pathologists.

Lisa-Jean Clifford (00:32:57):

But what you’re also doing is eliminating barriers to adoption for again, really being able to take advantage of the true value of digital. And that’s expanding beyond your core laboratory, whether you have remote locations and remote or outreach customers, whether you’re looking to do consultations again, if you have customers that you’re looking to bring them onboard or remote locations, we’ve already invested in different scanners or in LIS is EMR, et cetera, you don’t want to have to go to them and tell them that they have to rip and replace, or that they have to reinvest in different hardware, different software applications in order to adopt digital again, removing those barriers and being truly and profitable is key. A couple of other things to really focus on is a true image management system. So early lessons learned from the radiology space is being able to have a true PACS solution and image management solution that supports intelligent and automated workflow routing really, really impacts the successful adoption and use and specifically when by the pathologists.

Lisa-Jean Clifford (00:34:14):

I know Dr. Evan’s mentioned not everybody is happy with every pathologist has a positive experience, but this goes a long way to not slowing them down into actually making them work kind of ship. Which of course is, is the best route for everybody. And then two other areas to really focus on obviously artificial intelligence and image analysis. So most organizations are going to be looking to deploy these. If you’re looking to start on a budget and take small steps and justify your costs, you may not be looking at that initially, but keep in mind that that will be the holy grail when you get to the end of the adoption phase. So you want to make sure that whatever you’re selecting today has the ability to integrate those algorithms and those image analysis applications in a way that again, supports that streamlined truly in tropical environment.

Lisa-Jean Clifford (00:35:15):

Another place where you can really have an impact on your costs is whether you choose a hosted or non hosted or on-premise solution as well. So for smaller or mid-sized organizations or larger organizations adopting a digital pathology process and implementing it internally can be a heavy lift. There’s a lot of equipment that’s necessarily on the it side from servers, chief firepower to fans with, to consider there’s the storage. There’s the staff to implement and support and manage all of those operating systems, the upgrades, the download times, the hot fail over the security, et cetera. If you’re looking to really take that whole and not have expense the out of pocket expense and true dollars, but then also the it staff expense and click that on a vendor. That’s another really good way to, to get started down this path is to put all of that expense, both human and real dollars on a vendor and in adoption hosted solution. Next slide, please.

Lisa-Jean Clifford (00:36:36):

And then looking at some of the selection process. So when you’re looking to select the digital pathology applications or solutions, again, you want to consider things like your volumes and your speed. So here you’re less concerned about the number of slides. You’re more concerned about the number of cases, and you obviously wanted to make an investment in a solution that again, you’ll be able to build upon, and that it will be able to scale with you as you grow in your adoption. So some of the questions to ask and the things to look for up front are, are there additional costs as I look to implement additional features or more cases or more case types, or I look to go from say, 20% digital to a hundred percent digital, what are the incremental costs and will this solution scale again, looking at things like speed.

Lisa-Jean Clifford (00:37:30):

So the quality of the image, how fast did the images in the cases load for the pathologist? Again, we not only don’t want to slow it down their process as you’re adopting a new digital, but you also want to make sure they don’t get that CSUN effect that some systems actually do provide with the pixelated than doing it, the images. And then you also want to look at things like flexibility from the vendor’s perspective. So when you’re on a budget and we’re looking to have X amount of dollars you’re going to have an understanding and you want to ask the question, is this coming out of the capital budget? Is this coming out of an operational budget? Is it some combination thereof and will a vendor work with me on my payment terms? Will they maybe flexible will they have be sensitive to my budget constraints instead of constantly trying to upsell me or to not provide me all of the information we need so that there’s hidden costs. Are you getting nickel and dime later on? And then lastly, on this side, he wants to make sure that you pick a vendor who is able and willing to work with other vendors as well, whether it’s the hardware scanner vendors or the other it application vendors, laboratory application vendors. So not only are they able to, but are they willing to play nicely with others? Next slide please.

Lisa-Jean Clifford (00:39:01):

So when I look at adopting digital pathology, obviously there’s the holy grail. There’s where all the way down the road and we’ve implemented, and that’s the best scenario, but in taking steps, when you’re considering budget and adoption you know, there’s the good, better, best approach. So a good start where you’re looking to justify or offset the costs. You’re actually able to start with the external or remote pathologists, and this provides an instant hard dollar savings. So as Dr. Evans mentioned, there’s the actual cost of shipping the glass, whether it’s couriers, whether it’s FedEx, whether it’s the packaging, there’s also the staff involved to actually do the packaging, the logging, the checking looking for missing glass, there’s great breakage and so forth. So as an organization, if you’re looking to justify the cost and you’re looking to figure out what your exact ROI is, you need to go to accounting.

Lisa-Jean Clifford (00:40:04):

If you don’t have a handle on what these actual costs are and pull these dollar amounts and look at them. I know of one laboratory that realized him paying $135,000 a year just to FedEx, but they can count their carriers, their packaging, their supplies, their staff, et cetera. And they were able to build a case to justify paying for their start down the digital adoption path with a complete break, even in year one. And again, this was a start, this wasn’t their full born digital deployment. And then again as Dr. Evans mentioned, we’ve also seen success with starting with frozens. So again, not having to have a pathologist travel to a remote site or have them waste that time, that windshield time not being in the office or a new primary main laboratory. Being able to start with person sections is another quick, easy, hard dollar room then that you can point to better starting with TC clients.

Lisa-Jean Clifford (00:41:10):

So again, this is an instant hard dollar savings. So you not only do you have a faster turnaround times, but you’re able to get the cases to your CTC clients, same debt starting small. So again, being able to identify a specific case type or a subset of pathologists, it’s a smaller investment upfront in terms of the number of scanners and the digital pathology solution or licensing, depending on pricing model smaller investment upfront. And you’re realizing your gains, you’re realizing your ROI as you’re adopting. Another thing that I can suggest is offsetting the cost of your standard by offering scanning as a service. I know of a laboratory that actually offers this and they call it somewhere, there’s somebody new, you said their scanner, and that is helping to pay for an offset, their cost. It, you can look at it as an additional source of revenue, or you’re specifically applying it to this budget.

Lisa-Jean Clifford (00:42:12):

And then better is being able to identify vendors again, who are willing to work with you and be flexible on the payment terms or the way that you’re looking at paying for your adoption, whether again, it’s capital, whether it’s operational but whatever the accounting department is really dictating for you in terms of your adoption, but being able to do things where you’re able to adopt a full-blown solution upfront and pay for it over time, as you scale would really be ideal, because then you not only got your foundation, you’ve got all of the components and pieces for you to truly realize and start gaining ROI sooner rather than later. So things like a lease or a reagent rental model doing a sliding scale, if you pay per case some vendors it’s a per case fee. So obviously you’re only paying for what you’re using instead of paying for the whole system upfront and not realizing the usage or the game until much further down next slide, please.

Lisa-Jean Clifford (00:43:26):

Okay. so to RFP or not to RFP, and, and this may be something that is up to the laboratory, it may be out of the laboratory hands, depending on your organization and their requirements. Sometimes it’s based on a certain dollar amount you’re spending over a certain dollar amount. You have to go through that RFP process. And in some cases some laboratories will choose to do this because it’s new, it’s an education process, and it gives you the opportunity to clearly define your minds the right questions to ask the way to analyze and compare and contrast the different vendors who are out there in the market, providing these solutions, because there are you know, a good handful on all sides. I’m not going to get into the details. Again, these are things that you can come back to, but the next two slides do this slide.

Lisa-Jean Clifford (00:44:23):

And the next one you focus on this. The thing I really wanna point out though, is that when you look at building an RFP, really look at defining the relevant questions for the right area. And what I mean by that is you have hardware vendors, you build scanners and scammers and hardware on what they do. It’s their expertise. It’s where they’ve invested as a company, it’s where they make their money. So it’s the most important to them, for them to focus on focus on the questions and the vendors that are specific to providing the scanners that will really support your initiatives for digital, the same thing on the software side, the digital pathology software solution vendors. It’s what they do. It’s what they’ve invested. It, it’s where their expertise and their knowledge lies. I truly believe in adopting a best of breed approach so that you can pick and choose the components that are the right components for your laboratory. However, if you do decide that you’re going to go down a single path, you know, choose a vendor that provides both the scanners and the digital understand you are going to be giving up some flexibility and functionality if you go that route. But again, separate out those questions so that you’re getting the, the detailed answers and understand the functionality that you’re getting or giving up in that software side of the solution. Next slide, please.

Lisa-Jean Clifford (00:46:01):

Identifying vendors to invite again, you can, you can look at this, there’s multiple ways to identify which vendors are the right ones for you to hone in on and focus to invite, to participate in the R eight. But again, I want to focus a little bit more on requesting bits and this again goes back to supporting your budget and understanding which vendors might be flexible with you in supporting the features and functionality that are within your budget today, and understanding where you’re looking to get to in 2, 3, 4, 5 years with your adoption and how your budget is able to be justified and scaled to support that full deployment. And so again, really understanding when you’re looking at the bids so that you’re able to compare apples to apples. Next slide, please.

Lisa-Jean Clifford (00:46:55):

Okay. Now here, where I’m saying, understand your ROI. I do mean your ROI, each facility, and each organization is going to have different goals. They’re going to have different things that they’re looking at, whether it’s hard dollar ROI, whether it’s those soft dollars that we all are aware of in terms of efficiency, gains streamlining patient safety, et cetera or it’s some combination of the two, but again, understanding what your ROI metrics are, understanding what those actual dollars are. So spending some time to do research internally and understand what you’re actually spending on different areas like shipping in courier costs like packaging, labeling staff understanding what your current turnaround times are, both in-house. And if you have remote locations or regional facilities and, or outreach customers such as TC NPC customers are you looking to grow your organization? And if you are where those dollars come from and what are your goals for achieving additional revenue growth?

Lisa-Jean Clifford (00:48:07):

Again, these are all things that are supported by and bolstered by digital pathology and AI, but really understanding where your dollars are, will help justify those costs. Same thing on the SOC side. So increasing your volumes being more efficient allows you to increase your volume, your current volume with the same amount of staff, or if you have a reduction in staff, because we all know that there’s going to be, and there are shortages of pathologists and there’s increase in cancer rates being able to maintain your current volumes with less staff case routing and real time to triage cases, using image analysis and artificial intelligence algorithms really streamlines that workflow as well, and then higher accuracy. So improved accuracy, isn’t just better patient safety and outcomes. It can also provide legal defensibility. And again, you know, the pathologists specifically in an organization might not be concerned with that. The C-suite is and then faster turnaround times. So not only do you get a faster and are you able to streamline your business for faster turnaround times, but you also have a better client satisfaction. And again, that helps keep the competitors away from your door and you’re able to capture those metrics and you’re able to leverage those customers, those happy customers as references, taking business customers, and maybe even take some from several of your competitors. Next slide please.

Lisa-Jean Clifford (00:49:52):

So some quick wins. Yeah. And again, this is something that you can certainly look at and read in detail, but to help offset the investment of digital pathology again, being able to support remote locations without the need of a pathologist, having to take the time to travel. I’ll say this over and over picket packaging, shipping carriers, et cetera. But then there’s also the open and instant access for collaboration and sharing. And when colleagues for consults and for education in teaching many digital pathology solutions will have education modules and research modules as well. Streamlining the efficiency based on reductions and paperwork for pathologists, the need to work in multiple systems, searching for case information, et cetera, back to legal defensibility. So I know of one laboratory in particular, who said that they were able to eliminate one lawsuit that they were able to track, which paid for their entire digital pathology deployment.

Lisa-Jean Clifford (00:50:57):

And you’re able to actually, how are you able to do legal defensibility? Well, there’s really two ways. One is with digital pathology, most viewers or most digital pathology solutions actually track well, what the pathologist texts we looked at on a whole slide image. So you’re able to prove what the pathologist has seen and that they viewed the entire image. And then you’re able to show the whole slide image two. So that, and by doing that a couple of years later, or several years later there’s no degradation in the tissue sample or the glass, for example. And then another one is with artificial intelligence. So I remember early on maybe 15 years ago when we started talking about digital pathology and working with it on the LIS side, I had a pathologist say, you knew this stuck with me. It’s not the misdiagnosis that keeps me awake at night.

Lisa-Jean Clifford (00:51:54):

It’s the missed diagnosis that keeps me awake at night and AI and artificial intelligence algorithms can help with that by redirecting a pathologist to an area of interest that they may have Glastonbury not necessarily the same. And then again, there’s all the increases in revenue. And again, offsetting, you know, scanning slides as a service, so the Southern wedding, your scanner but then there’s also creating an image thing. And this is something that as a vendor, they started looking to build their algorithms and train those algorithms. This is a potential huge revenue source for a laboratory that does decent case volumes. You have the whole slide images with annotations and diagnoses associated with those images. There are vendors who will pay for that information and access to that information, obviously anonymized again, being able to grow your business with TC and PC work, but, and as Dr. Evans said you know, being able to replace locum tenens without the need to travel. So again, being able to take in additional business or have a pathologist or chief who just does that for other pathologists throughout the country, maybe completion or who are outset, next slide please. Okay.

Lisa-Jean Clifford (00:53:17):

So two final slides that I’d like to leave you with, and one is really the definition of success or identifying how you can be successful. And that really goes to management planning. And I do mean from the top down and whether this is part of your medical director, or this is from the laboratory director, whoever is responsible for your digital pathology initiative. You really need to determine what your objectives are and how you’re measuring your success. What are the criteria you’re measuring for success against those objectives? And this, this is key to being able to point to yes, we were successful. Yes, we had a great disappointment. Yes, we’ve achieved our goals. I do recommend that we, how do we phased approach for deployment? And this is from a budgetary standpoint, but also from a successful standpoint. So when you’re looking at the road for, you know, the end of the road is the holy grail.

Lisa-Jean Clifford (00:54:15):

You want steps and milestones that are measurable, that you can attach timeframes to that are achievable steps to point to that success. And as you’re moving down the road, and you’re also able that way to have ROI gains as you’re moving along. So you can point to, you know, we spent $200,000, we recouped 135,000. And we can point to that in this forum, for example again, an interoperability and transferability, I feel like I said of interoperability in my sleep, but that is really key to being able to leverage all of the benefits and realize the gains for digital pathology and AI adoption and deployment board approval. And buy-in get it up front and keep them involved. So they’re invested in the process and that they’re excited for your wins and we support your setbacks. I, I believe very strongly in that and then selecting the right vendors as partners.

Lisa-Jean Clifford (00:55:17):

So some vendors, you know, you sign on the dotted line and they’ve made the sale and you’re off. You really want a vendor who is also in this to help drive digital adoption and understand and work with you on the benefits. Because again, your criteria, your goals, your objectives are going to be different than your neighbors. So you want a vendor who’s truly going to partner with you to ensure that they understand your business objectives, and they want you to be successful and achieve them because that helps drive adoption for everybody in successful adoption. Next slide, please.

Lisa-Jean Clifford (00:55:56):

And common pitfalls. Don’t bite off more than you can choose. So if your initiative or your initial goal is too big, there’s too many moving pieces. There’s too many people involved. It’s going to become the never ending project. Again, I really recommend a phased approach and there are multiple ways that you can select what that phased approach is going to be. But I do recommend a gradual adoption setting, clear goals and timeframes, and then revisiting notes. So if goal number one, isn’t achievable in the, you know, your 20% mark timeframe because something happened with the, when I’m missing delivery for a piece of hardware, for example, moving goal number three, while you’re working on goals. Number one, so constantly Vincent repeat, rinse, and repeat. And then I would like to just leave you very quickly with some additional sources of funding within your organizations.

Lisa-Jean Clifford (00:57:00):

So again, being able to do an image repository, being able to do some subletting of your scanner not having your pathologists have to travel. And then Dr. Evans mentioned, this is very smart, not over the notes. A lot of the scanner vendors have demo systems that you can negotiate and reduce price for, for purchasing. And in some cases, those demo systems have only been used for a month or two by one client. So there are definitely ways that you can, you can do digital adoption on reduced budget and grow as you scale. I hope this was useful information for you. And with that, I would love to introduce Bill DeSalvo. Who’s going to talk about some of the technical and practical actions.

Bill DeSalvo (00:57:59):

Thank you, Lisa, Jean. Appreciate it. So today with all the information that you’ve had from, from Andrew and Lisa Jane we’re going to switch gears a little bit. I’m going to talk more about what does this do to our histology department cause that’s where a lot of changes happen, not only with the pathologist, so on this first slide, if you’d go out and Google histology, workflow, pathologist, workflow value, look at what kind of results you get, what kind of hits that still doesn’t help you. You still have to come up with your workflow and your, you have a unique lab.

Bill DeSalvo (00:58:40):

So each lab, whether you’re a single site, multiple sites, you have to look at your workflow and figure out what is going to affect the end result of your digital pathology. This is the simple answer that I try to tell most labs is this is exactly where lean technology comes back in. I truly believe that to be real successful with this add on technology, into the technical process, you’ve got to lean out, you’ve got to work in small batches and you have to move it through as efficient as you can. Everything is going to be affected from the of the samples and your gross dissection to your fixation putting some standardization. And, and that variation is it has gotta be removed as much as possible. Unfortunately, we don’t have any standards for all the laboratories. And, but I do believe most of the clinical laboratories have worked towards narrowing their variations. The other thing that helps us in the technical process is the ability to, once we bring in the digital pathology to enhance our QC QA it’s very hard to do that in a manual process, but you’re going to also connect your quality in your histology process right into the whole DP or digital pathology project. Next slide please.

Bill DeSalvo (01:00:15):

So it’s all about the workflow. Number one, we’ve heard about gaining efficiencies without scanning, without doing distribution. That’s not necessarily going to happen at a large scale in the technical side or in the histology department. You, you do have to add a little bit more case triage while you’re embedding cutting staining. If you are a traditional clinical lab, you only have a smaller percentage of those diagnostic biopsies, if you’re a full GI lab or prostate that less triaged. But when we consider the traditional clinical laboratory that has a hospital clientele, you’re going to have to do some organizing and moving the workflow through because the types of cases that the pathologist are going to need to get quickest results are, are going to be your diagnostic biopsies. So triaging, you’re going to have to start right from the beginning and trying to figure out how you’re going to keep those moving through the fastest in your process.

Bill DeSalvo (01:01:25):

The other thing is, is that we’re used to doing case management. You know, we’re, we’re either cutting by a case where definitely it may not be standing by cases at this point, but definitely when you get to distribution and you to do that, that loading of the flats and packing everything is by case DP. If you set it up, right, you have an open platform. DP now manages that taste assembly of delivery. So it kind of moves you to, how do we look at this process in a slide by slide case? And what it does is that it now pushes those cases out the software, then assembles them as you get that done and in a nice way, and in a very efficient way, the pathologist is actually managing more of that, that case review. But you do have to still do negotiate that and make sure that happens cutoff times for delivery for things leaving your laboratory are going to change.

Bill DeSalvo (01:02:24):

Hopefully the you’ll see that they’re going to extend, but it could be difficult if you are only a first shift, one shift pathology department to maintain that same day turnaround once it gets into the lab and out. So you do have to look at that and that’s where the lean is going to help you. And then how you’re going to interface. That’s, that’s a, a big issue. There still are laboratories that are not barcoded. If you’re thinking about doing digital pathology, if you’re not barcoded there’s no way to move a volume of work through your three library. Next slide please.

Bill DeSalvo (01:03:09):

So what type of, how many, so when you start talking about scanners we’ve heard all about what the speed of the scanners. But you also have to look at capability capacity since you’re in a clinical laboratory. Not only are you going to calculate the speed of the scanner and how many slides, that’s the first simple calculation you do, you have to also put in there the time needed to load the scanner and then remove the scan slides out of that, because that really goes into your scanning time. And then you have to look at for your case mix your number of pathologists, and you’re in the volume that’s moving through. You will have peak times how many slides in every two minutes do I need to get out? Do I need two standards? Do I need three? Do I need 10 scanners that can help through those peak times and get those delivered out?

Bill DeSalvo (01:04:07):

That’s a very important calculation that you’re going to have to do. And then with the third, what I call the third generation scanners, there are multiple generation scanners out there being used for digital pathology, with the third generation scanners. You want to see how much technical intervention has to be applied to the scanning process. The older standards, if you have to then review thumbnails and refocus and do all of that, you can’t, you can’t consider moving a high volume number of slides. Luckily these third generation scanners, they’re all either have built in QC or they’re auto focusing through the whole process. This takes away the absolute need that somebody needs a QC yet. And what’s going to be important for you in that, in that evaluation of the system and in your, in your clinical lab, you need to do proof of concept with your pathologist, that today, if they’re receiving your slides and, and they’re reviewing and signing out that’s how you want to do your proof of concept.

Bill DeSalvo (01:05:20):

You want to be able to scan the slide, let it go right to the image management software and right to the pathologist. That’s where the lean comes back in. If you’ve got a good stable, controlled precision process, you should be able to do that, but all the slides now will continue to reside in the histology lab. So you also have to look about if you’re sending these off site, when you send them out for review, you also have to look at what is going to be your physical storage capacity and where you’re going to do that. Where are you going to group it up.

Bill DeSalvo (01:06:02):

So typically there we go. Thank you. So typically in today’s laboratory, these are the steps and this is very simplified, but most of your, you know, you may or may not have a barcoding tracking system. You may or may not be connected to your LIS and his all the way through the process. If you really think about it, most of your communication is all via phone and email, back and forth with pathologists. There’s a lot of time consuming and you can go back over this slide to look at some of the things that you do need to concentrate on, but this is a typical today’s clinical laboratory. It takes a lot of people to manage, to track down, to, to just keep the process going. And it’s not always very connected. So it is all case. I would have to say most of my experiences, I say that it’s case oriented.

Bill DeSalvo (01:07:05):

And if we go to the next slide, when you move to that digital workflow, it has to be all connected. It’s just, it’s just too time consuming that when the slide gets to the standing portion of your technical process, that someone’s typing in some sort of identifier, whether it’s succession number, name, or whatever, they just can’t do that. So you’re completely connected. And I truly believe that as, as you look at this and once you connect up the, all of your pieces and parts, and that’s what you first have to understand, what are all the pieces and parts that I do need to connect out. But once you do that I’m very, very confident that what you’re going to do is you’re going to start moving once you get the case delivered to you, because you’re not going to change that the specimen is going to come to you, specimen or specimens, but at right after grossing, you’re kind of moving through a single block, single slide, because you now have the power of that information management system, which comes with the digital pathology platform to then manage that distribution and where those slides go.

Bill DeSalvo (01:08:21):

The nice thing is, is that if you have a multi location where you have pathologist, and Andrew mentioned that where you have different pathologists, all of you will have had the situation where pathologists have schedule it location X, and you find out after you’ve done your distribution, that their location, Y once you move to digitized workflow, the digitalization of the slide, digital pathology that sits out on the server. So if a pathologist moves to another location, they still have access to that. And you’re not going to have to move it back and forth. You do to you take some true consideration of what is going to be your savings, because if you have X amount of people working the process now, and they’re doing all that distribution most likely they’re going to be responsible for the digital scanning process that putting on taking off and managing that somehow there still is some management to it, but it can be very automated. And then the other change will be that once you move into this connected system, this workflow, the pathologist will be less and less calling on you. There’ll be interacting with all the multiple softwares to LIS the scanning software going to that, that, that workflow management page that they’re going to be working from. And then everything else flows along digitally. Also, whether it’s a tumor board, whether it’s a consultation within the group, outside the group, there, all of this moves to that more efficient workflow. Excellent, please.

Bill DeSalvo (01:10:09):

Yeah. So what do we require multiple scales? We’ve got to learn in histology lab, just exactly what all this means. The biggest thing that I didn’t hear in the other two talks is that if you’re thinking about going digital is your, is your it infrastructure in the, in the institution or your multi-site institution ready for it? May not be that is can be a process stopper can have some dollars behind that, but you need to be considering that before you moved to the demonstration and then the information workflow everybody kind of hit on it, but are, are your slides or your images? And these are large images. They’re at least one, some of our to three gigs where, how are they going to travel? And you have to solve that before you try to move into doing some sort of demonstration.

Bill DeSalvo (01:11:12):

But the other thing I think you need to know is as you understand that technology flow bandwidth you’ll find that you want to know the basics of the digital pathology platform. And that’s how I put it together. You’ve got one or more scanners. That’s one part of the platform. You have an information mini server information management system in the middle. That’s connected to those standards that will then connect to the pathologist or other software, whether it’s a case management software, whether it is AI, whether it’s deep learning, there’s that many server. And then you have, what are your viewing experience going to be? There are multiple out there now, there still is only one system that is fully FDA approved for all three portions of that. And that takes us a step back is that that decision has to be made with your pathologist.

Bill DeSalvo (01:12:12):

Are they going to go with an FDA approved platform here in the U S which is very important when you’re looking at risk assessment or are they comfortable with, and can be the, almost the same amount of risk if we do it correctly, go with a laboratory developed test and validate. You have to validate both systems anyway, anyway, so you have to, you have to be looking at all of these things and how they’re going to affect your workflow within the, the, the technical histology lab. I remember basically digital pathology is another add on to your process. It’s coming right after scanning. And you have to make sure that you can manage that quickly. And consistently the worst thing that can happen is that the, you know, once you move to a full WSI and that’s a number that moves around a lot too, if it’s just the H news, let’s just talk about the HDS.

Bill DeSalvo (01:13:11):

If you’re having scanning or connected connecting difficulties it’s somewhat out of your hands. It’s either the vendor for the scanner, or it’s your, it support their it support. It’s not as easy as, you know, just cut me another slide and send it and we’ll work it out later. So you have to understand how this is going to affect you. The standard technology is, is very important to you, but one of the things I wanted to talk about with the scanner technology as well, we saw some of the early versions of those scanners. They were multiple minutes to get you a thumb’s tail. Then you had to have some level of technical expertise reviewing those slides, re positioning and re selecting vocal points before they could be rescanned again, and then sent out. So what you want is something that can take away that technical expertise at the end.

Bill DeSalvo (01:14:10):

So most of the third generation scanners out there either are continuous focusing, or they have, as I think Lisa Jean brought up there is some built in QC. I understand what that’s going to do to the process. They may scan in a minute or 45 seconds, but how long is that internal QC going to take? Then the key number for your workflow is how much faster do you need to be than two minutes? Remember, the histology workflow is a very manual task oriented workflow. If you have a scanner that can go 30 seconds or 45 seconds, you may or may not realize any change in your workflow because that manual workflow in histology really can’t keep increasing to that level. There’s going to have to be at some point, if we want instantaneous and quick through, there’s going to have to be some point to some additional automation in the histology workflow.

Bill DeSalvo (01:15:08):

And then I think a lot of people kind of just glass over or are forget about it is that soon as those scanners come into your histology, your QC and your QA increase, you now can more effectively and re in real time control the intensity and the color of your stains, whether it’s H and D special stain, I’d see, by using the digital scanner and having a better communication back and forth from the pathologist to letting them know what’s happening with the stain. I have found that when we’re in a slide workflow, pathologists do have a very wide level of acceptance and, and can work through sometimes large changes in the way the stain happens. As you move to digital pathology, you’re going to find that that acceptance gets a little lower and there is a real need that you create that control and precision for all of those slides.

Bill DeSalvo (01:16:13):

The other thing I’m going to suggest is that most labs today use as their indicator of how well their process is working today is the H and D slide that has, that is not our best slide to do that. I suggested as soon as you bring in digital pathology, you start looking at, I see, as your indicator for how well the process is working, and that is checking your pre-analytics. There are a lot more sensitive and they can show smaller changes than the HD can. And then the last part, as you’re looking at, you’re going to have to start understanding that AI software what’s it going to do? What’s it, how it’s going to affect your technical process? What does it mean to us? And how’s that going to work? You really have to come up with some basics in information technology for you to discharge manages process. But I think you’ll, you’ll see that a quickly move from a batch, a larger batch process, which is all case oriented. As you move through to slide management, how, you know, what, what is our process for getting the blocks, getting those slides off of stains and onto the scanner, because now you have technology software management software behind you, helping you with that, that distribution. Yeah. Please.

Bill DeSalvo (01:17:44):

I’ll just hit on a little bit of, we’ve seen what people have done for budgeting. I will say every site is unique, you know you have to first find out and you’re going to be talking with all of your internal people. Whether, again, whoever, whatever that co group meet is that that’s got to be your it people, because there has to be most cases you’re going to have some infrastructure upgrades. It has to be, everybody is usually involved in it. And then you have to find out what is your process capital purchase? You own it. Some people have asked me, you know, do I really want to own the standards? Well, with the third generation, I think they fit in great for the, the type of information that we want to digitize. And it’s usable with all of the software that’s out there.

Bill DeSalvo (01:18:37):

So the scanners are not going, I don’t see a big change in the scanners over the next several years. So owning is still an option for you, cost per slide. Again, I think Lisa Jean mentioned that that can be negotiable. You will have to negotiate it because the problem is, is when you’re in a clinical laboratory, when we go WSI. So every slide will be counted as a slide scanned and a slide stored. And I think that’s one of the big places where there’s a lot of discussion, again, in, in a purely clinical laboratory, what will be the image retention, and that’s a continuing cost. So that has to be discussed upfront. And then there’s also a couple of companies out there that do platform scanner rentals. What they’re doing is, you know, providing you an open source end up, but you still have to do some, some very hard and deep digging into what that open source is.

Bill DeSalvo (01:19:39):

You may have offenders such as your histology information system that isn’t willing to open up with a bi-directional interface so that you can connect all of this. All of those pieces and parts really have to be addressed. And I hope agreed upon before you try to move to digital pathology, hard dollars, soft dollars, Andrew and Lisa Jean had hit on that. You know, it in the clinical lab, it, there can be an ROI. There can be an acceptable ROI, but I think mostly you’re looking at how to justify this with a strategic implementation. You know, how are you fitting in with the innovation, the patient safety, the correctness of your diagnosis, what we can do for the, the oncologist, things like that. So it, it, you may get into a place where it is all about a strategic and innovative placement instead of just an ROI.

Bill DeSalvo (01:20:42):

And again, I, I have to hit on, again, it infantry infrastructure upgrades, most of our histology labs in the U S just because you have access to the internet, doesn’t mean you’re ready to send into one gig image across a hall across to another building and everything else. So that is where you have to have a hard sit down with your it department. And the other thing you need to be aware of is that we’re very used to putting in an instrument today, Val quickly validating and using it. DP is not a digital pathology is not a plug and play for us. It has to be into fully integrated in to all of the systems and that that’s going to be new to us here in the clinical technical workflow in that almost everything we have is a plug and play put in. We’re allowed a lot of variation of how we’re going to use it, but it’s not always completely integrated into all of the it systems within your, your institution or your, your site. Next slide, please.

Bill DeSalvo (01:21:48):

So you’re, it’s all about what are you gonna do about implementing the change? I know Lisa Jean hit on, you know, there’s, there’s niche applications you know, if you’re going to do digital you’re, you’re still going to have to go through the, the same cost for your it, whether you’re going to do niche or whether you’re going to move through a phase project into whole slide imaging. But the first thing you have to do is, is really come to the conclusion of what you’re scanning. You want it to accomplish no, as adjusted, I want to replace class lives and get it out to the pathologist. Am I looking to do so, what am I looking to do with that scan digital image? And that can be a very involved conversation, but it involves your, it, your, your management and your pathologist about what we’re going to do and what we’re trying to accomplish with the scanning. [inaudible]

Bill DeSalvo (01:22:46):

Well, as you’re going through the process, digital pathology, you’re going to be in multiple workflows. Now, if you’re only doing a few computational pathology stain, such as ERPR, you can manage two workflows. You’re not going to be able to manage multiple workflows for for indefinite time where the majority of the slides, or either for the whole slide imaging or they’re glass, that’s very, very costly and very, very difficult to do. So you go to the situational DP as I call it, but you’re still trying to justify the whole project of what is your end goal. But I think the, the first win that we can all use in the clinical lab is that non-primary diagnosis is you have many options, QC QA, and especially your recuts special stains. I see when those get ordered, if you have a little adjustment in your workflows, those can come out and save you hours, half days or a day before they get into the pathologist, letting them letting the pathologist transition first with the recut specials.

Bill DeSalvo (01:24:02):

And ICS, I think is a really good idea that to take you through your phases, and then, you know, you have to define for everybody in the system, what the realistic time is for your turnaround or you’re going to extend the technical workflow process within your laboratory. And then, like I said, it, this definitely will, as we get to full adoption enhance ability and I think just solidify the pathologist as the doctor’s doctor. We will have I think the most volatile part in all of this is what new software’s coming at us. What do AI, what new deep learning software, but the key for us in the laboratory is a completely open system that has this, these new peripherals come in. We’re able to add those and just integrate them into the process that you’ve set up next slide. Yes.

Bill DeSalvo (01:25:06):

So you’re still doing the same things your, your meeting, the patient need the customer need, or we’ll be able to better put the pathologist into the, make them from a general pathologist to the specialist within, within seconds, within a few keystrokes, they can do that. We’re going to increase the accuracy of the diagnosis. It’s not a, I think Liza Jane said she had an experience where misdiagnosis or the missed diagnosis. I think we’re going to get into an age as more and more laboratories adopt is to what is the most correct diagnosis and looking at that over and under, they are not missed. They’re not missed, but we can increase the accuracy. Then there’s going to be a lot of data mining off of those slides that I’ve most Lisa Jean absolutely brought this up is that there’s value. Once you move into digital pathology and you and you, that bank of slides is what that value is, is not only to your own institution and your own patient population is how you can partner as you go out. And then this definitely enhances the ability of all pathologists. It will be a transition for them. But you know, if you understand that at any one point in time, in most laboratories, especially in the laboratories I’ve worked with or worked in there’s as many as four different generations of pathologists. So this will not, I think everybody has it. This will not be a quick transition, but it can be a successful transition. Next slide please.

Bill DeSalvo (01:26:46):

So with that, I’ll turn it back over to Dr. Keith Kaplan and we’ll see where he goes,

Keith Kaplan

Thank you to everyone. Thank you to Andrew, Lisa-Jean, and Bill for some great presentations very informative. I think we accomplished the goal of the webinar. We are approaching the bottom of the hour. We want to be cognizant of the time for the audience, as well as the participants. I don’t know if we have any questions that came in, I guess, maybe to put a ribbon around this, and we’ve all been in the lab long enough now to have watched this in some capacity. But the evolution of digital pathology to me seems very analogous in retrospect, I’ve graduated medical school 25 years ago tomorrow. And, and so I’ve watched this now and in that capacity long enough, the evolution of digital pathology seems very analogous to me to immunohistic chemistry, implementation of immunohistochemistry, particularly automated immunohistic chemistry.

Keith Kaplan (01:28:29):

When I started as a medical student and certainly as a resident, I mean, it’s chemistry was a manual process, labor intensive time, intensive, very little standards, not regulated, but not completely unregulated. Kind of the wild wild west. Everybody was coming up with their own cocktails and their own dry ice recipes and their own microwave recipes. And eventually that started to become standardized and uniform. And all of us know that the immunohistochemistry stainers the people that repair those machines, unlike the Maytag people are not the loneliest people in the world. They’re very, very busy. And I, and I think it’s also somewhat analogous to a lesser degree perhaps, cause there was certainly more, I think, direction from a regulatory perspective on the front end, someone analogous to liquid-based cytology. We were early adopters back in 2000 or excuse me, yeah, 2000.

Keith Kaplan (01:29:33):

But I know of labs by, 2010, 2012, we didn’t necessarily have high volume, high throughput, automated pre-screening machines were still reluctant concerns about job losses. Cytotechs psychopaths costs, bandwidth capacity, space storage, et cetera. So I, you know, having watched that, I’ve tried to convert some of those lessons learned into it’s a different digital pathology implementations over the year. Let me start with building and Lisa-Jean, and Andrew to wrap it up. And in 60 seconds or less Bill, what would you say? You know, how do you, how can you be a change agent when it comes to, you know, this kind of disruptive really, I would say disruptive workflow in the laboratory with the stakeholders. How do you spur that change? Lisa-Jean from a, from a vendor perspective, having watched these different technologies disrupt the lab, how, from your perspective, being a change agent,

Lisa-Jean Clifford (01:31:28):

I think really having an internal champion and making sure that you have a team who is excited about and looking forward to, and understands the advantages and benefits that are going to impact your organization. And then again, building a clear message for your organization around what those goals and objectives are and showing your successes along the way as you’re achieving those goals and accomplishments. And, and again, having an internal champion who whether it’s your medical director or a subgroup of pathologists who are truly excited about using the new technologies and can talk to their peers about their excitement and the benefits and the value that they see and are receiving

Keith Kaplan (01:32:22):

Andrew. There’s a couple of there’s a directed question for you. And then one for Bill let me combine these two questions so that we can have the participants get answers to these what are the advantages or disadvantages of whole slide imaging versus remote robotic stage for frozen section? Okay, sure. With respect to what are the qualifications of the personnel used in the gross room or in the gross examination and frozen section for remote frozen? Sure.

Andrew Evans (01:32:59):

Okay. And to answer the second one first, so at least in Canada, we have certified pathologist assistants who do the day-to-day grossing of the vast majority of specimens. So in other words, they are fully certified to perform delegated medical acts. And then for remote specimen review, we can use, there are a variety of different options using webcams et cetera. We just simply go through MS teams and have an iPad with a high resolution camera. And that’s how we we review specimen gross specimens remotely. And they’d be processed by a pharmacist assistant with respect to the advantages disadvantages of whole slide imaging versus robotic microscopy for frozen sections. I think starting with the ladder, the robotic approach, particularly real-time video, which is what we’re using now at McKenzie health, we just use this Aqura vision tech device and we use it in real-time video microscopy mode.

Andrew Evans (01:34:00):

And that allows you to do multiplanar focusing just as you would with a microscope. So recognizing frozen section quality is not always the best and you get thick and thin sections and et cetera. So and so that seems to work. It seems to work very well and that flexibility is, is important, whole slide imaging. It does work well, but it places a much greater reliance on the quality of histology in my experience. So if you have lots of big tissue folds, the way most of these scanners work, you know, you could end up having a section that’s poorly focused and you end up having to recut the section or rescan it, and which will delay the interoperative turnaround time. So if you’re trying to achieve CHP turnaround time benchmarks of 20 minutes for a single block frozen sections, you can sometimes run a run up against some challenges.

Andrew Evans (01:34:55):

It’s also been my experience that while it fails very infrequently. So at UHN and the, in my experience with the ditch path the whole slide imaging frozen section coverage, it was points 0.3% of our cases. So it was in a handful of cases per year, where there would be some challenge with the scanner and a pathologist had to run over to the Toronto Western site to review a frozen section. And those things often happened with some, some kind of unexpected calibration problem, a light bulb failure in the scanner. We have not yet had that happen with the hybrid scanner. So I think there are some possibilities for breakdown, but it’s, as long as you have really good quality histology both whole slide imaging and, and video microscopy, robotic microscopy will work very well.

Keith Kaplan (01:35:49):

Very good. This is always an interesting question. So I’ll put you on the spot here. Do you think digital pathology decreases turnaround time in the scenario where the pathologists are on site?

Bill DeSalvo (01:36:10):

Okay. You can put me on the spot. Depends if, if a pathologist is in the habit of walking over and before it’s put into all the processes and, and grabbing a slide, absolutely. That can happen. But if you set up your workflow properly in whether it’s a one shift to shift three shifts, let’s just talk about a one shift. If you set up that work for properly when the pathologist comes in at whatever time they want to come in one or multiple this process will have images available for them to see which in my mind keeps them from walking across the hall to the histology department. It is not just, it’s not that hard. And I, I I’m at a loss for words, you should be able once you have this workflow established with the right number of scanners and how you’re going to manage it, you, we should all be able to stay even, or ahead of the pathologists as they, as they need slides.

Bill DeSalvo (01:37:13):

Most often my experience is if you’ve got pathologists coming over to the histology lab two things they just want to get out of their office, or there’s some sort of delay in the histology lab. So this again, just reinforces, we, we need to continue to embrace lean and set it up. And that’s why I, I mentioned that you’re moving really to a slide management. If you’re triaging your diagnostic biopsies, first, some of those could only be two or one, two or three slides to a case. You can turn out multiple of those cases and keep your pathologist being working and being efficient before you get to those 10, 20, 30 block cases resections. So there, the workflow we’ll be able to stay ahead of the pathologist. I truly believe

Keith Kaplan (01:38:03):

I have a, I have a follow-up to that, but in the interest of time, I want to make sure that we address this comment as well. And Andrew, I think you promised this in your opening remarks, if you can address your monitor resolution comment, how I don’t need a, a $25,000 medical grade monitor for

Andrew Evans (01:38:21):

Sure. I mean, this is, this is something that I think has become a bit entrenched because of the FDA approval process. So for the Phillips system, it’s a closed system that is, is submitted for the, in this case for the pre-market approval trial as a class three medical device. And then that’s the, that whole closed system is what’s approved for, for diagnostic use. So it’s become entrenched in there. And that Barco monitors in particular, the very expensive medical grade monitors we’re part of that system going forward, I think, I would say from personal experience, we have the Barco monitor here as part of our Phillips system, and that’s what we use in the office. But when we were looking for, you know, considering budget for the home workstation application, did we really need that monitor for the homework station?

Andrew Evans (01:39:11):

And the answer was no. So what we did, we went through 300 cases, prospectively divided by across seven pathologists and reviewed at home using a, just an off the shelf. Hewlett Packard monitor that taxes in et cetera, came in around $500. Reviewed those cases, came into the office. The next day, reviewed the reviewed the images on the Barco monitor and glass slides. There was no difference in diagnostic performance based on the monitor. And that’s, that’s, that’s our experience. That’s just our empirical evidence. Another paper, a very interesting one that I would cite is a paper from Memorial Sloan Kettering that described a validation study for remote sign-out during the pandemic published in modern path about a year ago and what they did there, there was no specific institution issued workstation that was given for home signup that pathologist there just simply used what they had. And so that could range from high-end gaming computers, with ultra fast graphics cards to laptops, and likewise in their validation study, their internal validation study. There was no difference in diagnostic performance based on the display or monitor used. So I think it’s, you’re well and truly able to use any monitor that you, that you really want to use, and that fits your budget provided you validated.

Keith Kaplan (01:40:41):

Very good. Thank you. There’s a, there’s a question It’s a, it’s an extensive question actually. I think there’s a lot of reasons. I’m going to ask everybody to chime in on this if you’re doing. So the question is given digital pathology candidates have been around for many years. Why do you think the uptake has been so good? Slow is cost really the only limiting factor is the reason. Yeah. To believe the, is there a reason now to believe this adoption will accelerate? If so, why now? And let me kick this one off. I think that there were several issues when post slide scanners started to when I became interested in this in the late nineties, when whole slide scanners started to become to a degree where they could reliably reproducibly scan slides over time. So we could have discussions now about two minutes times don’t matter, or one minute times don’t matter, cause it will overwhelm capacity, et cetera, but years ago that wasn’t always the case.

Keith Kaplan (01:41:49):

And so I think to some degree, the industry as a whole, as a community, we put the cart before the horse a little bit, and we, I think we over-focused on the technology rather than the applications. In addition to that, I think that, and this has been pointed out numerous there’s times. Unlike radiology and they’re transition, of course the films were already digital, but still being printed. But when it came time to use packs, we didn’t have additional CPT codes to account for slide scanning costs by scanning services and reimbursement for those services. So there wasn’t a killer app quote unquote, in terms of ROI, it wasn’t, and this was, this has been mentioned. It wasn’t analogous to plugging in a culture counter and having return on investment from day one. This was going to be a good to have that would have some tangibles and intangibles, but they wouldn’t necessarily be our dollars that you could measure that could say this one we’ll pay for the scanner in 27 months or 22 months, depending on volume.

Keith Kaplan (01:43:01):

Yeah. Et cetera. I think what’s happened. Since then, those days certainly is a lot of disruption in the market. There’s been a lot more laboratory consolidation. There’s many hospital systems that have bought up other hospital systems and farm core laboratories. So I think increasingly pathologists are farther and farther away from where their histology is produced or histology laboratory, the supporting many different sites. So I think there’s applications now, you know, and certainly I think the regulatory issues for the better part of a decade and a half probably slowed adoption as well, but I think that’s changing. And of course, lastly, I would say the ability and, and I remember your beliefs and Lauren [inaudible] and Darren trainer about this years ago, their trainer said digital pathology is the future pathology and always will be. And the last time he gave that quote, he then talked about his fully digital department of pathology at Leeds.

Keith Kaplan (01:44:06):

So the, I think one of the, one of the value propositions that be that’s being recognized now, and both of these have already been addressed, but they remain the same. One is some form of teleconsultation remote read. And the other one was historically image analysis in terms of a hard, tangible ROI in terms of a Delta and reimbursement. Now we just moved the words around and instead of image analysis, we have artificial intelligence, we have IAA and AI and people talking about this to 20 years ago, having these vast digital data sets, which can be mined for information for either clinical decision support or workflow enhancement or efficiency given fewer pathologists aging population than with chronic diseases, et cetera. So those are my 2 cents maybe 5 cents. Let me open it up to the floor and see if anybody agrees, disagrees any other reasons why the adoption has been slowed. And are we at an inflection point now or are we going to continue to see a similar trajectory? It looks like Lisa-Jean has  the box.

Lisa-Jean Clifford (01:45:36):

So I, I would echo and say ditto to what Dr. Kaplan says. However I do think that there are a couple of other reasons and, and some of it has been wait, so waiting to see what other laboratories are doing, waiting to see what the fallout from adoption or the benefits from adoption in true clinical practice actually showed in terms of success rates or efficiency gains and actual ROI. So I think that the cost has certainly benefactor, but I think it’s been more of a wait and see type of attitude. And what we’ve actually seen when you talk about adoption is yes, it’s definitely accelerating. So I have been talking about digital pathology and working with digital pathology vendors for 15 years. And I’d say in the last two to two and a half years, we’re really starting to see the adoption and attraction took place.

Lisa-Jean Clifford (01:46:39):

And a lot of that has been because of the benefits and the outcome. But I also think a lot of that has to do so obviously cannot adopt or deploy AI without digital pathology. And so I think a lot of laboratories and facilities and pathologists are realizing that there’s an additional technology that’s going to be coming down the road and they’re being left behind. If they’re not adopting the steps that they need in order to take advantage of what we used to say, our future technologies that are available today and understanding that that’s just going to continue to grow. This is actually the future of pathology.

Keith Kaplan (01:47:23):

Andrew, the next 18 years looked like the past 18 years.

Andrew Evans (01:47:29):

I agree with exactly with what both you and Lisa Jean had said. I don’t really have that much to add. Just for my own, as I mentioned during my presentation there that in the, for the greater Toronto digital pathology network that we’re trying to get established here I don’t think we could have had these conversations about 10, 12 years ago. I’m not sure that these other institutions would have come to the table, but now they’re all interested. The, the stumbling blocks largely seems to be how to finance it and then how to generate some revenue to keep it going. So that, that’s the biggest shift that I’ve noticed. And I think that these things will be overcome. I know, I think the pandemic has undeniably created a lot more interest in particularly in remote sign out. I mean, look no farther than the waiver that was issued by health and human services and CMS to allow remote signup without the need for a CLIA license for your home. So I think, you know, those, those types of things you know, out of this disaster, I think it will create I think will cause people to think about digital pathology in a new light.

Keith Kaplan: Bill I’ll let you have the final word on this.

Bill DeSalvo (01:48:41):

Oh, well, thank you. Well, pathology always moves slow because we have multiple generations, but I think that the, for me, the discussion started changing when we had the first FDA approved system. And as you’ve mentioned, the null hypothesis that the two were equal, I can remember all the way back to 2 0 6. We continually argued glass digital glass, digital what’s best that kind of has gone away. The other thing is I take a little bit different look because I’m on the technical side. I think the understanding that the AI that we’ve been promised is actually useful and we have a patient population and precision medicine that they already believe we’re we’re functioning in completely. And this digital will allow us to do that. There are great gains every place that I go and talk to them, they’re already getting a lot of push from their oncologists about how do we do this?

Speaker 5 (01:49:44):

And then you’ve seen several of the vendors go in into, so companion diagnostics are going to come out more and more companion diagnostics and complimentary. That is since we’ve been using that term today, the holy grail, that is your revenue gain with digital pathology, as we get closely connected to precision medicine with those companion diagnostics, then you have to make the decision. Am I going to participate on that level, or am I going to stay as a conventional laboratory and deal with regular specimens, pushing them through with an H and E I think that’s more of the conversation that I’m hearing from different labs is it’s almost, we can’t be left out of, this is kind of how I look at it at this point and I’ve been hearing. So I think that’s where it’s going. I think that’s, what’s really making it considerable in the U S

Keith Kaplan (01:50:44):

Very good, I think. Well, set. I think that’s a, that’s a good note to wind up on I, my thanks again, to all the participants for taking a couple of hours and my sincere thanks to all of the speakers for their time and preparation. And two is our report wishing you all a happy, safe holiday weekend. Yup.

Liz Carey (01:51:08):

With that, that’s going to be a wrap on this extended dark daily webinar. Thanks everyone for joining us. And at this time we’re signing off from dark daily. Have a great afternoon. Okay. Bye-bye thanks.