Women in Pathology Informatics - A Conversation with CAP Staffer Mary Kennedy

• Council on Informatics and Pathology Innovation

Lisa Tomcko:

Welcome to the latest edition of the College of American Pathologists CAPcast Series. I'm Lisa Tomcko, content specialist at the CAP. March is Women's History Month, and in this episode, we're celebrating the illustrious career of CAP staffer Mary Kennedy. Mary joined the CAP in 1995 following the start of her career in cytotechnology and laboratory management. Her initial work was in support of the gynecologic cytology proficiency testing or PAP program, non-GYN program and PIP program, and included serving as primary staff to the cytopathology and surgical pathology committees.

Mary later joined the CAP SNOMED team in 1999 and became oncology services manager with SNOMED Terminology Solutions. In 2010, she left the CAP to serve as pathology informatics program manager for the National Cancer Institute's Cancer Human Biobank Program before rejoining the CAP the following year as Director of Clinical Informatics Initiatives. In that role, which she still holds, Mary supports the Digital and Computational Pathology Committee, Informatics Committee, and the Artificial Intelligence Committee, as well as the newly formed Council on Informatics and Pathology Innovation, which is the CAP's first new council in 18 years.

Among her many impressive accomplishments at the College of American Pathologists, Mary secured grant funding from the Centers for Disease Control and Prevention to support the CAP Cancer Protocols. For this standout achievement and her numerous other contributions, Mary received the CAP Outstanding Staff Award in 2021. She also received the Association for Pathology Informatics Lifetime Achievement Award in 2022. Mary, this seemed like a good time to sit down and speak with you about your career at the CAP and your experiences in pathology informatics. So welcome.

Mary Kennedy:

Thank you, Lisa.

Lisa Tomcko:

And just to get right into the questions, you made your career at the intersection of diagnostic testing and technology. What impact do you think technology and specifically informatics has had on pathology and laboratory medicine?

Mary Kennedy:

I think it's important to first define what we mean by informatics, since it really is a very broad topic. But, first, let me begin by dispelling the notion that informatics requires a computer science degree or coding experience. Informatics harnesses the power and possibility of digital technology to transform data and information into knowledge for patient care. So there obviously is a strong focus on the use of computing to analyze the data, but the field also incorporates the human resources and the devices as well as the designs that are required to optimize that information.

Many people will say health informatics is the intersection of people, technology, and data used to improve the safety and quality of patient care. And that certainly has been my experience over the years. Remember, computers are tools, but it takes a lot more than computers to get something right.

Lisa Tomcko:

That's interesting. So informatics is more than just computers and technology, although that is a big part of it.

Mary Kennedy:

So, of course, without technology, a lot of drug development wouldn't have been possible. Vaccinations such as COVID would not have been possible without informatics capability of the human genome project and the molecular analysis. In terms of pathology and laboratory medicine, informatics today is ubiquitous. It's part of the daily workflow in a lab. So if you think of cytology, genomics, flow cytometry, molecular, and the instruments used in both the clinical and anatomic pathology, you can begin to understand the impact of informatics. Our understanding of mutations based on big data wouldn't have been possible without informatics. Even laboratory information systems or LISs not only store that data, but do provide the analysis of the data across multiple disciplines.

Lisa Tomcko:

Wow. It's amazing to think of the impact that informatics has had within healthcare at large, but then also specifically within the field of pathology. What was it like to be in that healthcare informatics space as a woman? Did you encounter any resistance?

Mary Kennedy:

I wouldn't say that there has been resistance, but there's not always been recognition either. A lot of informatics work is what we call invisible work that is really accomplished behind the scenes, so that the end user of the data and the knowledge that they've been given has very little understanding of what goes into providing that knowledge. I've heard from other women that this type of invisible work behind the scenes work is not always recognized by male counterparts, yet a male performing that same time of work will be recognized. I have noticed in the past that when engaged with a group of men and women discussing informatics such as an informatics conference, men tended to address men during the conference. But as we see more women entering the field, that tendency is lessening. The field is still dominated by men, although there is a growing number of women who are now invested in pathology informatics.

Lisa Tomcko:

Good to know that some of those may be traditional dynamics are hopefully changing a little bit as more women enter the field. And let's get around to the Cancer Protocols, which are a huge resource for our members. They're consistently the most downloaded files on cap.org, and they're also something you've been heavily involved with. Can you talk to us about what cancer reporting was like before the advent of the protocols?

Mary Kennedy:

Before the Cancer Protocols were developed, most surgical pathology cancer reporting was done in a narrative style and format, which was dictated by the preference either of the institution or of the pathologist. And even pathologists within the same institution had different formats too. So this narrative format was not interoperable across systems and sometimes not even within the same institution because of that lack of standardization and structure of the report. So the Cancer Committee, through their development of the Cancer Protocols, allowed cancer reporting to become standardized in a structured format that can then be consumed by an LIS.

This greatly benefited physicians who were treating patients as well as cross-institutional understanding as it reduced a lot of the confusion on the diagnosis and where to find the diagnosis on a report. The standardized list format in the Cancer Protocols has reduced the time for a pathologist to fully describe a case and ensures that all the necessary data elements are now captured in that final report.

Lisa Tomcko:

Would you say pre-Cancer Protocols and standardization, was it a little bit kind of pathologists having to rebuild the wheel every time they're having to put together a report?

Mary Kennedy:

I would say that when there was a narrative report, and I worked in a lab when everybody was still in the narrative report phase, even pathologists within the institution I was in would report out differently. There was some kind of a structure, but basically as long as you're putting something in narrative, it's very difficult to take the data elements out of a narrative and standardize them so that they can be compared to other similar types of reporting. So having that list of data elements that you need to put and trying to structure it and standardize it in a very, very formatted way, made it very easy for oncologists and surgeons to read the new kind of structured report. They knew where to look for everything.

Lisa Tomcko:

Definitely. That makes sense. And at what point in your career did you get involved with the Cancer Protocols?

Mary Kennedy:

After several years at the CAP in the CSA area, I moved into informatics at the time that the CAP was focused on the development of SNOMED CT. So the CDC was very interested in our development of the cancer protocol data elements and viewed it as an opportunity to standardize cancer reporting to cancer registries. So, as a result, I was asked to manage the first CDC grant awarded to the CAP to SNOMED encode the cancer protocol data elements. It was the start of our continued collaboration with the CDC over the years, which still continues today, as well as our involvement with the National Cancer Registry organizations.

As most people probably know, cancer reporting to cancer registries is a requirement. So the creation of a standardized reporting format allowed cancer reporting to become much more efficient. It also reduced the amount of time that the government needed to report out on cancers and their outcomes.

Lisa Tomcko:

So very beneficial for pretty much everyone it sounds like. And when you say that you managed the first CDC grant, what did that management entail?

Mary Kennedy:

The management of the CDC grant, because it was a collaborative effort with the CDC, we were able to work hand in hand with not only the CDC but with the cancer registrars. And then also because we owned SNOMED CT at the time, we were able to in-house code those Cancer Protocols using SNOMED codes, and then those SNOMED codes could be put directly into an LIS system. So if you go to retrieve that data from an LIS, it was much easier than trying to pick out and review every narrative that came in. This way, you had very standard data elements that you could lump together using a SNOMED code then and actually retrieve and analyze the data.

Lisa Tomcko:

Got it. And since 1995, you've spent the bulk of your career at the CAP, say for a brief stint working for the National Cancer Institute's Cancer Human Biobank Program in Washington, DC. What made you want to take that job with the government?

Mary Kennedy:

Well, I was offered an opportunity by a female member of the CAP who was leading that CA Health program at the NCI and to work with her and her team on developing some robust biorepository standards of operation that could then be scaled up to the national level. I actually spent a very exciting year at the NCI learning about a different side of informatics from the viewpoint of the government and the ways to optimize standards from the government's perspective. I knew that what I had learned at the NCI could be leveraged at the CAP as we begin to investigate new technologies, especially with advances in data science that will be needed for machine learning and artificial intelligence.

Lisa Tomcko:

How did informatics from a government perspective differ from your informatics experiences up to that point?

Mary Kennedy:

I wouldn't say I had a too narrow of a view, but we focused so much on pathology at the CAP that it was very interesting to see informatics from a much bigger perspective, especially in terms of a public health perspective, which is what the government is very focused on. So I think just seeing a bigger perspective was something that was very valuable to bring back to the CAP.

Lisa Tomcko:

I'm sure. And what were some specific takeaways, if you recall, from your year with the Cancer Human Biobank that you kind of knew right away that you would want to take in and implement?

Mary Kennedy:

I think, again, that goes back to the standardization of reporting because we could not have done or could not have built a large scale biorepository for the country if we didn't have standardized reporting. So I focused a lot on the standardization of some of the data elements as well as some of the coding systems that could be used to help implement those into... Well, it wasn't an LIS in that case, but into their system and map them, as we say, to a different type of ontology so that the two were equal.

Lisa Tomcko:

Got it. That makes sense. And I'm sure it was really helpful to see large scale like that. And then coming back to your present role with the CAP, now as Director of Clinical Informatics Initiatives, you support the newly formed Artificial Intelligence Committee as part of that role. Can you tell us what you're excited for regarding AI in pathology? Anything you think is on the horizon?

Mary Kennedy:

That's an interesting question. I think artificial intelligence will play a significant role in pathology in the next five to 10 years. And when I say AI will be important, what I mean is that pathologists who use AI tools will elevate the standard of patient care. I don't think AI will replace pathologists, but it will become part of the workflow. We often use the term augmented intelligence when talking about AI in pathology and the AMA actually defined augmented intelligence as a conceptualization of artificial intelligence that focuses on AI's assisted role, emphasizing that its design enhances human intelligence rather than replaces it.

And I think that defines it very succinctly and is something that we at the CAP believe in. AI tools will serve as an adjunct to the human experience. And pathologists using AI will receive the highest level of diagnostic expertise regardless of their level of experience. I think AI will allow inexperienced pathologists to operate at a level beyond their experience and practice and could be especially helpful for pathologists practicing in smaller hospitals who really don't see a wide range of case types.

Lisa Tomcko:

So it's kind of helping pathologists to level up and be able to diagnose more cases rather than a situation of AI supplanting pathologists.

Mary Kennedy:

Right. And I also think it's going to serve as a teaching mechanism, especially for some of those who are newer in practice, and especially once we have AI tools that have been FDA approved for primary diagnosis or primary review, because then it will kind of serve as quality control. So if a pathologist wants to look at an AI first, they might be directed by that tool to look at these cells or slides first. And I think that will be very beneficial in the long run.

Lisa Tomcko:

Wow. Yeah, that sounds fascinating and very helpful. And on a similar note, what trends are you seeing these days in the field of pathology informatics?

Mary Kennedy:

I think this is a very fast-growing field in pathology and hope that more women will consider becoming involved in informatics. We've definitely seen an increase in the demand for pathologists with informatics knowledge or experience. And actually some larger institutions now have dedicated pathology informatics departments. Healthcare organizations are also realizing the potential and power of data analysis, which certainly includes lab data. And that's an important aspect for everyone to remember is the importance of lab data in developing any kind of artificial intelligence, machine learning or informatics special studies in general.

There's an understanding that informatics can accomplish what humans alone can't because of the large amounts of data or what people refer to as big data. This is especially true I think in whole slide imaging that will likely be the precursor for many artificial intelligent tools for AP.

Lisa Tomcko:

Mm-hmm. Whole slide imaging definitely seems like something more and more practices are doing. And what advice do you have for women getting into the informatics side of pathology?

Mary Kennedy:

I would advise women to take an active role in the informatics areas. Don't let your prior lack of experience make you think you're unqualified. From a practical perspective, networking is a very important skill, and as you build connections in the informatics field, you also build your skills. So as if you reach out to pathologists who are also interested in this area, they can help mentor you through this. I've spoken with a number of CAP members about encouraging more pathologists, especially women, to enter this particular field. These members have said, and it's probably a little tongue in cheek, that pathology is a specialty of rule followers, while informatics is about developing better rules through quality improvement. Informatics will always be a part of a lifelong process as it continues to evolve and advance because the tools will continue to evolve and advance.

I personally think that informatics cultivates the natural-born leaders in pathology because of the need for continuous change management skills and the willingness to be open to improving how things are currently being done. So my advice is that while woman face more challenges than men in being accepted as a leader in this area, stay firm in your beliefs and opinions. Women should not be afraid of entering this field. You can't affect change if you think you must always conform. And don't be intimidated in expressing your ideas. In the end, very few individuals are willing to champion new ideas. I think that informatics will continue to be a cornerstone of how medicine is practiced. We are no longer in the information age. We are now in the knowledge age. So go boldly into informatics. I think you will enjoy it.

Lisa Tomcko:

Love that. That's great advice to women getting into the field of informatics and specifically pathology informatics. Well, Mary, thanks so much for joining us today. You've made a lasting impact at the CAP and on the field of pathology informatics. Thank you for sharing your insights.

Mary Kennedy:

Thank you, Lisa.

Lisa Tomcko:

Thanks, Mary. And for all those listening, check out cap.org and the MyCAP app for more exciting Women's History Month content from the CAP.

Download Transcript