Is Digital Pathology Better, Faster, Cheaper, More Accurate? No, but it is essential for survival.

For over 20 years, I contended that when the day came our desks were loaded with 3 or 4 or 5 monitors and we were signing out cases entirely with whole slide imaging rather than conventional light microscopy, that digital pathology would be better, faster, cheaper and more accurate than analog slides alone.

Having now at least to have the option of using whole slide imaging exclusively for sign out, and on some days, doing so, for nearly 18 months, I can admit how wrong I was. It is difficult to make predictions, particularly about the future, as the great philosopher Yogi Berra was fond of saying.

Of course, a six-figure scanner costs more than a clinical grade microscope with nice lenses, tilting head and built in polarizers. Not to mention the software for the image management system, LIS integration, storage, backup and those 3 or 4 or 5 monitors, one of which costs as much as a new microscope alone.? So, it isn’t cheaper. It is multiples more. Add in depreciation, replacement costs, added personnel and space. You get the idea. Anyone who has looked at whole slide imaging for clinical practice has done this math.

The images are derived from the slides we are accustomed to using. Each of our laboratories, even if we move from one institution or group to another, has the very best H&Es. When we inspect other laboratories and review some of their cases, we think to ourselves, these are nice stains, but not as nice as mine. Ditto for consults or second opinions prior to surgery. We think these stains are good, but ours are better! It would follow then that the images from the best stained slides in the US if not the world are also the best. A good stain, a good image. A great stain, a great image. It takes time and effort and coordination and equipment and people loading and unloading slides and assigning cases to have them appear on your worklist and available to you. It’s hard to make an argument that our new way is “better” than the other way.

For the sake of argument, given all the studies, current practice and years of experience from a handful of large laboratories who have shown their results, whole slide imaging is equivalent to microscope slides. After more than 2 decades since the first scanners were conceived, we rejected the null hypothesis that whole slide imaging was inferior to the light microscope.

Faster? This reminds me off the NFL coach who during a press conference when a reporter asked him about the playoffs, the coach responded, “What's that? Ah - Playoffs? Don't talk about - playoffs? You kidding me? Playoffs? I just hope we can win a game!". For further reference on this if you don’t know the reference, search “Jim Mora Playoffs”.

Converting a 1” x 3” glass slide with tissue, stains, and coverslip (which we still have to make as of this post), to a whole slide image takes extra time. This, however, is relative and where I can make an argument that whole slide imaging is critical for survival. Nonetheless, it is not faster, having to move the slides from cover slipper to scanner and then sort again for slide delivery, if for no other reason than storage based on current requirements for laboratories.

What then is the value proposition for digital pathology? Why do you need whole slide imaging, equipment, software, people, storage and so forth?

The answer is what I called in one of my first papers on telepathology – “time to diagnosis”. This time looked at the time it took from the time the slide was placed on a robotic microscopic stage, to the time a diagnosis was rendered. The stage, and hence the images, were obtained remotely and the consulting pathologist, in these instances, would render his/her diagnosis. During our studies and validations, I would have a stopwatch and monitor “time to diagnosis” or “TTD”.

For cases from Europe or Asia, one could argue, as I did, that the TTD for a consult or second opinion was shortened by days or perhaps even a week or more. Minutes rather than dozens of hours. Patients and families spared a wait to know the diagnosis.

Of course, most of us do not practice in a healthcare system as far reaching as the military health network with requirements on multiple continents.

But, increasingly we are part of larger and larger “health care networks” that may in many ways not entirely “networked” sans for the core laboratory that centrally accessions, processes, cuts, stains and coverslips glass slides and performs immunohistochemistry, special stains, flow cytometry, perhaps some flow, molecular and FISH and so forth.

These labs can be 40, 50 or 60 miles from a given point of service where the biopsy/surgery was performed, and the pathologist is sitting. This is not new and in fact 15 years ago received my slides from a lab 40 miles away. If the courier missed the picked up, got a flat tire, ran into inclement weather or got stuck in line at the Bojangles, your slides were delayed, your TTD was delayed. Your day was delayed.

Now the lab may be 60 miles along some of the most heavily traveled roads in the area. Again, engine trouble, inclement weather, long line at the Hardees drive through and your TTD is delayed. Days are delayed. Results are delayed.

An old pathologist (not old when he said this, but old now) told me to never sign out a cancer after 3 PM. He said if you have been covering frozens and reading cases since 7 AM, that is 8 hours on your eyes and mind. Look at it, dictate it, show it around, do all those things, but don’t sign it out again until looking at it again with fresh eyes the next morning. The right answer can wait. He told me this on a Friday night about 7 PM before digging his hands into a stack of slides and signing out some malignancies. But you get my point. TTD delayed means results delayed.

Whole slide imaging, scanning on site where slides are made, helps with TTD. In fact, without whole slide imaging many cases would not be seen until the next day if one has to wait for glass. One can essentially have a continuous flow of images rather than boluses of trays. One can “see” everything before 3 PM, if not sign it out, as would be usual and customary, if the histology lab were across the hall or down the hall as many of us were accustomed to.

Many times, in the past before a weekend, long weekend or trying to get of town for a meeting or personal leave, those cases that normally appear at 3 PM or 4 PM in the afternoon after the courier fights 60 miles of traffic in bad weather have already been seen and signed out with whole slide imaging. Otherwise, cases may sit for days beyond needing them to do so. Now those slides that smell like cheeseburgers and French fries can be filed.

Digital pathology is essential for daily survival. It is not really any faster or cheaper, on its own, compared with in house histology, but for core lab models, with any distance involved, a critical tool for daily practice.

Artificial intelligence, pre-screening of slides, guided imaging, decision support tools will only enable this further.

The consideration of being concerned about signing out a new malignancy after 3 PM because slides are delayed will be a memory much like the days of having to reserve time on the electron microscope to look for microvilli, intercellular bridges, autosomes or Birbeck granules to make a diagnosis!