Implementing Digital Pathology in the Integrated Health Network Setting

Special contribution to tissuepathology.com by Dr. David Smith

Since its introduction in the late 1990s, whole slide imaging has steadily ushered in a paradigm shift from conventional light microscopy to digital pathology, enabling pathologists and AP labs to realize enhanced quality control and time, labor and cost savings. But managing that transition in an integrated health network (IHN) is not without its challenges. Addressing critical needs like software integration, data interface capabilities, and a general culture shift within the pathology team is essential for the successful implementation of digital pathology in an IHN setting.

Factors driving the shift to digital pathology

By simplest definition, digital pathology — also known as digital or virtual microscopy — uses advanced computer technology to convert standard glass slide images into barcoded electronic versions for the purposes of high-definition viewing, analysis and archiving. The process works by scanning glass slides, capturing the data stored within to create a digital image that can be quickly accessed, easily manipulated and edited, shared electronically across distance and stored compactly.

Traditionally, pathologists have used light microscopy to inspect and analyze tissue samples mounted on glass slides. In the nearly 20 years that have followed the introduction of whole slide imaging, implementation of digital pathology has become much more widespread because it provides so many benefits over analog methods. From a sheer volume standpoint, digital pathology offers the ability to analyze specimens much more swiftly and accurately, increasing the number of slides that can be processed in a given time frame. This streamlined workflow saves users time and labor, leading to quicker procurement of results, decreased turnaround times and, ultimately, computer-aided diagnoses that may get patients into treatment faster.

Radiology departments have already blazed the trail in the use of digital technology, but some diagnostic labs are just now warming to the concept. In some cases, their adoption has lagged because of financial or logistical factors—especially the logistical challenges of integrating new laboratory technology into established healthcare institution protocols.

Making the decision to implement digital pathology at Riverside Health

The pathology group at Riverside Health Systems works primarily in adult medicine and provides services for the hospital in five additional acute-care facilities. Like many other practices, we offer centralized histology, cytology and anatomic pathology services, employing couriers to transport specimens between locations for processing at our main medical center.

Several years ago, in conjunction with our quest to become a Breast Cancer Center of Excellence, we performed a thorough examination of our procedures for processing breast tissue specimens from pre-analysis through post-analysis. At the time, we discovered a distinct lack of standardization and some quality control issues with the way breast tissue samples were being handled and processed on the front end. What we quickly understood was the need to develop a breast cancer biomarker program using digital pathology that met or exceeded the CAP/ASCO published guidelines. Our goals included establishing a system that multiple pathologists and staff could use with reproducible results, eliminating data entry redundancies and improving turnaround times.

Originally, Riverside utilized robotic micro-imaging telescopes with analytic software to analyze cases one at a time. Later, we transitioned to automated platforms for immunohistochemistry and ISH staining (VENTANA BenchMark Ultra, Roche Diagnostics, Indianapolis), which we currently use. We also decided to make the transition to digital pathology, so that the slides that used to be read manually are now being processed through automated slide scanners and digital imaging software (iScan Coreo scanners, Virtuoso software, Roche Diagnostics). The software allows us to view all cases in the system comprehensively, and to zero in on specific files using full and thumbnail sketches with magnification, panning and gating tools for more detailed analysis.

An important part of our digital pathology solution is companion algorithm software for breast cancer biomarkers, which helps us address the standardization issue we identified in our earlier analysis specific to reproducibility of breast cancer test results. The fact that the FDA has approved the Roche algorithm for digital image analysis for estrogen receptor, progesterone receptor, HER2, Ki-67 and p53 gives us a high level of confidence in the reporting our laboratory is able to generate.

In addition to the benefits we have seen in workflow efficiency and standardization of image analysis,  the integration of the digital pathology software with our Laboratory Information System has enabled patient demographics to flow seamlessly from our existing LIS directly into the digital pathology solution. Without that integrated solution, someone in the lab would have had to reenter all of the demographics manually at a computer console, increasing the potential for human error and costly mistakes. The digital pathology data interface capabilities save us time and assure accuracy, subsequently improving overall patient safety.

Beyond these benefits, standardization of our slide analysis through digital pathology lays the groundwork for clearer communication with colleagues by creating homogenous electronic reporting that can be accessed and shared remotely in tumor board meetings or used for teaching applications.

Implementation considerations with digital pathology

While digital pathology clearly offers advantages for pathology practices, the actual implementation in labs can present some challenges. To start, there is the initial financial outlay to install the hardware and software, followed by any ongoing expenses required to maintain or eventually upgrade the platforms. Although improved results and cost savings over time can certainly justify the investment, figuring out how to pay for the implementation upfront can be a significant hurdle for smaller organizations with budgetary limitations.

Another issue is file storage. Since digital pathology mechanisms process a great volume of data in a short time, organizations need to make sure their systems have the means to store the amount of data generated. Part of that determination is deciding exactly what kinds of data will be stored and for how long. It’s a smart idea to consult your hospital’s IT staff for direction and advice when making these determinations before going very far down the road.

Managing slide quality control is another consideration with digital pathology, especially during technical preparation of slides and in-laboratory analysis at the primary (H&E) staining phase. If these steps are not managed properly, it can introduce manual errors that can lead to scan failures, slowing down the workflow and decreasing the overall effectiveness of the digital pathology process. Fortunately, this issue can often be mitigated through the use of lab platforms that automate many of these processes.

One of the goals in adopting digital pathology is often to facilitate remote consultation and sharing of information across disparate healthcare systems, but there are challenges associated with that goal. The current regulatory environment in general tends to decelerate the progress of implementing some collaborative functions. Plus, in order to share files, other consultative groups may need to be able to view images on the same platform, an undertaking that isn’t always feasible.

Finally, on a human level, the move from analog to digital pathology represents a culture shift, one that some traditional-minded pathologists may find hard to accept. There is normally resistance to change, and the thought of being required to learn a whole new system can create anxiety among the staff. There can even be a sense of fear that the introduction of digital diagnostics will eventually make the need for pathologists and even histotechnologists obsolete. This, of course, isn’t true, but lab managers should be aware that persuading reluctant staff members to embrace digital technology might require thoughtful and sensitive planning.

Making the decision for your institution

If your organization is considering going digital, the first thing to think about is software integration and data interface capability between your current lab information system and any prospective digital pathology solutions you’re considering installing. Determine a realistic budget and then carefully weigh the pros and cons of the hardware and software product choices available to you within those parameters to see which offer the best fit.

Then, set up a practical timeline to phase in the implementation, taking into account the physical transition and any necessary training procedures for your staff. And don’t forget to find out how to access related support services from your IT department and from the manufacturer before, during and after implementation. Establishing a collaborative discussion with both of them from the start will result in a smoother planning and implementation process and will enable your lab to take advantage of the full benefits that digital pathology can offer—namely time savings, streamlined workflow, standardization and sharing ability.

David M. Smith, MD, is pathologist and Laboratory Medical Director at Riverside Health System in Newport News, Virginia.