NEW YORK (GenomeWeb) – At the recent annual meeting of the Association for Molecular Pathology, the organization announced that it was creating a dedicated information subdivision to address the role of the scientific discipline in molecular pathology.
The new subdivision is intended to provide a "professional home" for scientists working in bioinformatics, clinical informatics, and other informatics as it relates to molecular and genomic pathology, AMP President Elaine Lyon said in a statement.
Alexis Carter has been tapped to chair AMP's new informatics subdivision. In the past, she helped form and chair AMP's Informatics Interest Group. Commenting on the launch of the new division, Carter noted that it "allows members increased organizational support to continue to service the clinical laboratory community." She also said that that community can expect to see "enhanced educational offerings, volunteer opportunities, and new projects" from AMP throughout the year and at the next annual meeting.
GenomeWeb caught up with Carter, who is also the director of pathology informatics at Emory University, to chat in more depth about AMP's newest subdivision, and the opportunities and challenges of using information in the molecular pathology domain. What follows is an edited version of the conversation.
From perusing past AMP schedules including this year's, I can see that there have always been informatics presentations and activities at the conference. Why launch a separate subdivision?
The Association for Molecular Pathology since its inception 20 years ago has had four sub-divisions — hematopathology, infectious diseases, genetics, and solid tumors. This is the first time that they have actually formed a new sub-division within the organization. Subdivision status within AMP gives the particular interest group the ability to represent their area of expertise more formally with the governing board of the organization. In addition, we have formal representatives sitting on AMP's nominating committee, their programming committee, their training and education committee, and the clinical practice committee.
What does having a more formal representation translate to?
It gives them voting rights. It also gives them a forum to bring up specific issues to make sure that informatics is not left out of initiatives that they are dealing with. Informatics is different than all of the other subdivisions in the sense that [it's] kind of a foundational underpinning of the practice that all the other subdivisions engage in. We are all using information management and technology and all of these other things, and there is a science behind how to do that in such a way that you are going to be fast and accurate. I think it's helpful to get that into the community in a formal way.
Is there something specific happening now in the informatics space that makes having a unique forum within AMP a more pressing need?
The leadership of AMP as well as multiple members have had an interest in learning more about informatics. About eight years ago, they tried to start an interest group, but at that point, we hadn't reached critical mass and there weren't enough people who realized the importance of it. Over the last 18 months or so, there has been a significant shift in molecular and genomic pathology but also in pathology and medicine as a whole, in recognizing the importance of informatics and informatics practice to the field.
In your opinion, how important is informatics to the growth of the molecular pathology domain?
Informatics is not the same as information technology. It's about having very sound practices as to how you collect information, store it, analyze it, report it, retrieve it later; and making [it] such that users accurately understand the information that you are trying to convey to them. From that perspective, it is becoming abundantly clear [that] informatics is critically important to everything. As we start moving more and more of our care to computer systems, it's really important to make sure how that information is getting in, how it's coming out; [and that] how we are looking at it and analyzing it and interpreting it is accurate. People make mistakes when information is not handled in an appropriate way.
Following up on that, what would you say are some of the major challenges or barriers facing informatics in terms of both tool development and adoption?
I think there continues to be a barrier in terms of people understanding what informatics is. Within the molecular community, there's a little bit of an education process that we are working on in terms of trying to get people to understand that bioinformatics, for example, [has] a small degree of overlap with clinical informatics but they are actually very different in a lot of ways. When you have people who don't have a good handle on what something is, it's hard for them to be willing to give up money to develop an infrastructure around it.
From a practical perspective, the biggest challenge that I see now in genomics is that we have massively parallel genomic sequencers and bioinformatics pipelines that are being developed at 90 miles an hour. [But] you still have somebody manually typing the results into a report in a laboratory information system because there are no instruments that I'm aware of that actually have electronic communication between those systems. So we have a lot of these incredibly expensive, technologically advanced instruments that literally can't communicate with anything outside their own box.
Health Level 7 (HL7) is the internationally recognized standard for communicating health data between different databases and between instruments and laboratory information management systems and electronic health records, [and] there is an HL7 genomics working group which is trying to develop standards specifically for working with genomic data. But the problem is that most of these molecular analyzers have been developed for research where there is no need to have those communications that exist.
You mentioned a need for community education earlier on. How might the informatics subdivision take that task on?
It is our hope within the informatics subdivision that we are going to start generating some of the best practice informatics guidelines. There are several important service efforts going on. I mentioned HL7. The Centers for Disease Control [and Prevention] also has a standard that they are trying to develop for a clinical grade variant call file. Other issues that have been brought to the table, [such as] how do I handle my data in the era of cloud computing given that there are all these regulations?
Speaking of standards and best practices, are you looking to tap into community-wide efforts along those lines such as the Global Alliance for Genomics and Health?
That was actively discussed by several members in our last meeting. We are going to have people who are going to be looking into what that group is doing and potentially seeing if there are areas [in which] we might be able to collaborate.
What else did you talk about at this year's informatics special interest group meeting? [Editor's note: the meeting was titled 'Bridging the Gaps between Clinical Informatics and Bioinformatics']
We had several people give presentations on how they are trying to marry up the pipeline with the clinical importance of the data that's being generated. Some people focused on how are we getting the data in, how are we making sure it stays accurate even though we are having to manually transcribe it, what solution are we coming up with to try to deal with these gaps that exist as the specimen comes in and then goes through the pipeline process and then comes out in the report. It spanned everything from the clinical grade variant call file that we were discussing to setting up your lab and how are you going to manage dealing with all the infrastructure that you have to have on board.
You mentioned electronic medical record systems earlier. What are some of the issues in that arm of the molecular pathology arena?
There is a large project called eMERGE to look at getting genomic data into EHRs. I think the challenge there is going to be figuring out what is going to get stored in the EHR. I fall into the camp of 'storing FASTQ files and raw sequence data in an EHR is not terribly helpful.' But clinically significant variants or unknown variants need to be in the EHR because as we learn more about what variants mean, it is helpful to have that information on record, in case a patient's status needs to change. [Also], because these variants can have huge therapeutic implications for patients … having [them] in an EHR in a discrete manner will enable us to start writing drug-genome check rules which don't currently exist in any EHR that I'm aware of.
Right now, there are no standards for what goes into an EHR. The HL7 genomics working group is trying to address some of this.
How will having a new subdivision affect future AMP meetings? Can we expect to see more talks, posters, and other activities around informatics?
I think there will be more informatics-targeted presentations and talks. The other advantage of the subdivision is we will also have our own area for people to submit abstracts. Previously, those were all submitted under technical topics which really kind of encompasses a very broad range. We do have some work to do with that because informatics is often combined with whatever area of genomics, medicine, or pathology that you are practicing in. So if it's an infectious diseases informatics [application], we have some work to do to figure out should it go into an infectious disease category or should it go into the informatics bucket as far as the abstract presentations go. I think one of the things we are going to try to do is not just say this is strictly an informatics talk, we want to have talks that really bring informatics home to people and put it into a realm that they can understand.
So the next AMP meeting is roughly 12 months away. What do you hope to accomplish between now and then?
We have to get appropriate people assigned to all of the individual committees of AMP where we need to have representation. I will be representing the division at the board level. We need to help develop the content for next year's program, we have to define the ground rules for how abstracts are going to be submitted. AMP has a process for getting white papers and clinical practice guidelines through the organization, so we are going to be working with the AMP leadership to do that. Also, [the] informatics subdivision needs to communicate with other [AMP] subdivisions and figure out how we can support their efforts as well.