We had the opportunity to talk to Clifford M. Chapman, MS, HTL, QIHC (ASCP) about changes in and the future of automation in the histology laboratory, as well as quality control. Clif holds the position of Technical Specialist at StrataDx (Lexington, MA), as well as Senior Consultant at Medi-Sci Consultants (Woburn, MA). Thank you for sharing your insights, Clif!
Can you tell us about your journey into histology?
I got into histology like everyone else- not planning on it. I started at the University of Milwaukee, where I took a two year course in electron microscopy. Because of that I got interested in microscopy and sectioning. I came back to Boston and worked at Tufts NE Med Center – Cancer Research, and part of that was histology. This was in 1976.
What types of labs have you worked in throughout career?
I started off in a cancer research laboratory, then went to a clinical laboratory doing renal biopsies, then back into a research laboratory doing special procedures such as methacrylate embedding and sectioning of bone specimens. From there I went into what I do now, working at a private reference laboratory, mostly skin specimens but also general surgical specimens as well. From working in the special procedures laboratory, I became technical director at various laboratories over the years.
How big of a lab is Strata?
At Strata, we have 15 cutting stations that are used around the clock – lab staff is 60 people including grossing and accessioning. This is about 30 histologists and 30 lab aids. We cut about 500,000 slides per year.
Do you ever take 5-8 slides per block? Or more?
Those would be cases where a pathologist has a differential diagnosis – let’s say, an undifferentiated tumor – and orders 3 special stains and 6 immunos. Percentage-wise at our laboratory it’s probably 5% of the total specimens. The average number of slides per block depends on the type of specimen. For dermatology specimens its roughly two slides per block, for hospital specimens it’s approximately 1.5 slides per block.
Dermatology specimens are more difficult to handle and process in the histology laboratory, due to their unique histology. A single skin specimen contains epidermis, dermis and adipose tissue; that is three very different tissue types in one paraffin block. Additionally, these specimens need specific orientation such that the final microscope slide shows the pathologist the exact areas of interest. For this reason, I worked with Dr. Izak B. Dimenstein, MD to write a book entitled “Dermatopathology Laboratory Techniques”. It has just recently been published and is available at Amazon.com. All of the intricacies of handling dermatopathology specimens are contained in the book.
What are your responsibilities as a technical director?
The main role of a technical director of a histology lab is to direct and be ultimately responsible for all slide quality. This includes all routine histology, H&E staining, special stains, and immunohistochemistry and now molecular techniques, as well.
How do you introduce new stains into your laboratory?
If we have a client that wants a new stain, then the medical director has to approve whether or not to do it, and if we move forward then it’s the technical director’s job to figure out how to get it done. This includes developing the procedure, optimizing it and validating it for use on patient specimens. We usually add a half dozen of new staining protocols per year.
What are the biggest changes you’ve observed in the histology field?
Generally speaking the biggest change is the expectation of decreased turnaround times. Thirty years ago there was not the pressure there is now to turn results around so quickly. That’s the main change that I’ve seen from a day-to-day viewpoint. Histology laboratories used to operate on one shift from 7:00 AM to 3:30PM. They would load tissue processors in the evening, process overnight, and then do it all over again the next day. Now there are multiple shifts– we currently run 24/7 with three shifts around the clock to accommodate the changed expectations on turn-around time. That’s been a driving force of change in histology laboratories.
Secondarily what’s changed is some of the automated technology. What’s had the biggest positive effect is the barcode tracking of slides and blocks and information through the histology laboratory because if you have a good system, it virtually guarantees 100% accuracy of blocks and specimens. Secondary to that or tertiary is the increased automation. The focus is to get things done more quickly while maintaining quality. That’s the final change at the bottom of the ladder. If people want all these things, decreased turn-around time, 100% accuracy, and quality, they need to consider both how to put people where they are needed and also how to automate these processes. Historically, everything has been done by hand so now we’re thinking, “What kinds of things can we automate in order to increase efficiency and minimize human error?”
As far as increasing efficiency and maintaining quality, the combination stainer/coverslipper set-ups – automated stainers hooked to coverslippers – have been rather transformative. There continues to be a large move towards integrated solutions.
The biggest automation change has been in tissue processors – everyone takes this for granted because automated tissue processors have been around since the 70s, but before then people did this by hand. You can choose closed system tissue processors (which most people use) or microwave assisted processors, which have been developed more recently.
In terms of safety to histologists, automated microtomes have been an improvement because they’ve reduced repetitive motion syndrome. At this point in time they don’t increase efficiency but they increase safety. They also increase quality. They have settings with slow rotation speeds to be sure you’re cutting and not tearing the tissues in the block.
Vendors are working towards standardizing everything. It’s all science and the idea is not to have humans handle anything.
What improvements would you like to see in the histology laboratory?
What I’d like to see – I’ve been saying it for years – I think the whole process of fixing tissues and dehydrating and embedding and then the opposite to stain – I think it’s all going away. The pressure is increasing to stop using toxic solvents. Someone will figure out some kind of medium – the magic medium – to make a water soluble block. Histology labs will put tissues in formalin and then just put them into a cassette. That cassette won’t even need to be processed – some machine will use a magic water soluble material to get a water soluble block. Then sections will be cut and affixed to slides and stained. All alcohol, xylene and paraffin go away.
This would also remove much of the danger of histology. Currently, a lot of laboratory personnel safety depends on ventilation. Also, chemicals cost a lot to purchase and then even more to remove and recycle and clean up.
I also think the microtome is going to end up being totally automated, though not in the next few years. A microtome that loads blocks on and picks a program specific to that tissue. It would measure how much to face off of each paraffin block, and then take sections, heat them, and then the slides go right into a staining rack. It’s entirely possible that once the grossing tech loads up a cassette, it might not need to be touched again until the slides are delivered to the pathologist.
Is histology an art or a science?
Right now histology is the last art form in medicine. By that I mean that when you go for your health care now, everything that gets done to you is automated. A machine does it. People used to do a complete blood count (CBC) by hand. Have you visited a clinical chemistry lab? It’s a giant space the size of half a football field. Inside the space is 100 machines all interconnected with little conveyor belts. A tube of blood is put on a conveyor belt with a barcode for a test and then a little robot pulls out the amount of blood, performs the test, then based on the results either stops there or sends it to another machine. This is the ultimate automation. Everything in medicine is done this way.
Histology is the last hands-on aspect of medicine. Because it’s so hands-on you call it an art form, but it is not going to be allowed to stay this way. When humans touch things they can make mistakes. With the emphasis on smaller and smaller pieces of tissue, there can’t be any mistakes – it’s got to be zero. Right now in histology labs you can’t guarantee zero mistakes. An acceptable suboptimal slide rate for us is anything less than 1.0%. We actually run at around 0.1%. Let’s say we average 3000 slides a day – out of that number, three are suboptimal. That doesn’t mean they can’t be signed out – and that’s excellent. But it’s still not zero. That is the focus. That is the goal. Even to this day, even now there are news stories, “took the left kidney out instead of the right!” How could this happen? Because humans make mistakes.
As a technical director I see quality control problems and I ask – why did this happen? Then I ask: how could it have been prevented? Histology is a technique used to make a diagnosis. We want to do anything that we can do to automate that and take it out of the hands of the humans. A few things are left by hand now but those are going to go away.
Histology is science and it should be treated as such. If you look at it historically, people thirty – plus years ago in the field got into it randomly. It primarily wasn’t science-based, and there was a lot of leeway in the procedures and you could still get similar results. Now, if all those histology blocks used for molecular testing aren’t processed properly, that information is not available for diagnosis.
Where are quality control and standardization headed?
Quality control is making everyone’s slides look the same, for example slides made in Michigan look the same as those made in Boston. This is already being pushed and documented by the College of American Pathologists (CAP) with their Laboratory Improvement Program. Labs have to grade slides and record all the information regarding processes and machines that were used in making the slides. They have a giant database to correlate which machines produced the best slides. For CAP it’s in their best interest to have the quality standardized because it helps pathologists do their jobs better.
How are reimbursement structures changing how things are done in histology labs?
Some background on CPT coding: The idea is that when you run a lab and run a test you submit a CPT code to whomever is paying you, e.g., Medicare or Medicaid or private insurers. The billing code is for one block and one slide. The “meat and potatoes” of histology slides is CPT code 88305. This includes all technical components of making 1-3 slides. During 2012 that reimbursement was, let’s say $100 – on Jan 1, 2013 it was cut by 52%. What does this mean? In 2013 if you did exactly the same amount of work in 2012 – you get half as much money. Now what do you do? You can’t change the costs of rent, heat, etc. The only unfixed cost is personnel. If payroll is $1M in 2012, you need to cut that in half. The logical thing to do is get rid of your techs, and a lot of large labs did this.
At StrataDx we went the opposite way. We didn’t cut staff because, as a result service would go down, turnaround time would increase, and ultimately, patient care would decrease. We kept everyone on and kept moving forward. When your revenue gets cut by half you can either cut staff down or you can do twice as much work. In 2012 you did 100,000 at 88305 – now do 200,000 – but you need more people and more machines – a giant balancing act. A few big laboratories on the east coast filed for reorganization – this is a serious concern. The people who don’t think histology will ever be fully automated need a wake-up call. If they are comfortable in their jobs and not working hard they better start paying attention. I make all 60 laboratory employees know about this since Strata is a private company, so we share in the profits. This puts an extra impetus on everyone to be working as hard as we can.
Thank you so much for a really interesting conversation, Clif!
If you liked this interview, check out previous Histology Thought Leaders interviews.
Stay in touch by signing up for our newsletter, contacting us, or following us on Twitter, Facebook, LinkedIn, or Google+.