Interview With Laura Myer of the American Joint Committee on Cancer
To learn more about the AJCC Content Interoperability project from the perspective of the AJCC we interviewed Laura Meyer, the AJCC 8th Edition Project Manager and Managing Editor, AJCC Cancer Staging Manual.
What were the circumstances that led to AJCC considering a move to structured authoring, content management, and digital publishing?
The AJCC Cancer Staging Manual basically provides the source code for determining a cancer patient’s survival prognosis. Staging provides a common language for doctors to effectively communicate about a patient’s cancer and collaborate on the best courses of treatment.
AJCC began publishing the Cancer Staging Manual in 1977. Staging has become much more complex in the last 40 years with the discovery of new types of cancer and new ways in which a person’s individual genetic make-up can determine their prognosis.
With this increase in complexity, AJCC has quadrupled its content base from 20 to over 80 chapters, and the number of contributors from approximately 100 to 420 worldwide. We needed to centralize our content base and streamline the process by which authors reviewed and edited content.
Most importantly, we needed a single source to share electronically with organizations that used our content for data collection. In the past, we would give them a print book and expect them to do the rest of the work to build their systems. Each organization had their own software developers (sometimes in different regions) who interpreted the rules and definitions differently. Additionally, some groups were slow to integrate a new manual into their collection systems because of the substantial investment to make the updates. This resulted in an infrastructure gap that often caused inconsistent data during a “transition year” after the publication of a new manual. AJCC relies on that data to inform updates to future cancer staging manuals.
Can you describe the process of updating and delivering the manual prior to this process?
Every 5 years, AJCC would convene a number of SME (subject matter experts) task forces (generally based on organ systems) to review the content from the latest edition, review literature published since the last publication, and run data collected since the last publication. After a series of face-to-face meetings and conference calls, the task forces would determine what needed to be changed based on the available literature, data, and clinical judgment. One or more members of the task forces would edit their respective chapters, which would then be submitted to a major scientific publisher for copyediting, print production and sales.
What led you to easyDITA and Scriptorium?
In 2013 we embarked on an exploratory information mapping exercise. We wanted to see if our content could be structured in such a way that it was less narrative and more tabular and instructive. We learned from this activity that the nuances of cancer staging do not lend themselves well to step-by-step guides. Narrative content needed to be retained for the clinician audience, but the defining elements of cancer stage could be in an easier to understand tabular format. We struck this balance by developing a chapter template that would be consistent across all chapters, applying narrative and tabular structure at the appropriate points.
We were referred to Scriptorium by a business consulting firm that was working with us at the time. That firm’s goal was to help us find the most appropriate vendor to develop the information architecture for our content. Scriptorium recommended a number of component content management system developers, and after a long round of interviews and demonstrations, easyDITA was awarded the contract. Both Scriptorium and easyDITA were a good fit for our organization. We were putting our toes in the structured authoring/content management water, and these companies patiently eased us into it.
Was there internal and/or external resistance to making this move?
Of course, but that’s to be expected. AJCC is a part of a much larger organization, the American College of Surgeons. We spent a non-significant amount of time proving the business case to College’s Executive Director. He has authored a number of papers, periodicals, journals, commercially published textbooks and self-published textbooks. The College also publishes a number of textbooks through both channels. Why did AJCC have to do this now? However, once he understood the benefits of structured authoring and content management, he worked tirelessly to help us find the products and pathways that would fit into our budget. Moreover, he wanted to find “right fit” vendors who could potentially work with other areas of the College on future projects. AJCC is the test case.
What were the biggest challenges, from AJCC’s POV, during the implementation?
We finalized our contracts with Scriptorium and easyDITA just as our authors were submitting their chapter manuscripts for editorial review. Our project kicked off in January 2016, and our publisher deadline was April 15, 2016. As such, I couldn’t dedicate a lot of time at the beginning to finalize the information architecture. We based the architecture off the chapter templates, but our authors did go rogue in a few areas. There were elements of the architecture that we couldn’t really predict until the chapters were final, as the authors were still validating their proposals. It wasn’t until around May/June that I could do a full audit of our content and work with the Scriptorium and easyDITA teams to modify the XML structure to include new information. Additional adjustments were required in July, August and September when we really started moving content into the system. The staff at both Scriptorium and easyDITA were very patient with me during this time.
Also, because we were so behind on entering content, there wasn’t a lot of opportunity for more extensive training. There are features of easyDITA software that I know would have made my life easier, but I just didn’t have time to learn it. That’s not easyDITA’s fault; it just worked out that way. I would have liked more time to learn the system before using it, but we were in a self-imposed rush. We are planning a formal training in spring 2017.
Now that it is in use, what are the principal benefits for AJCC, your authors, and your users?
Our authors haven’t used it yet, and we don’t have other users on the software besides myself. As I mentioned before, we haven’t been able to leverage all the features just yet, but when we do I am sure we can start transitioning to managing content solely within the CCMS. As it stands, I have to update errata in the master Word files (as a failsafe), staging forms, the publisher’s proof (for reprints), and the CCMS. The CCMS is linked to our API, and the API end-users probably reap the most benefit from the system now; their content is more up-to-date than the print manuals. We plan to build more outputs that will help us move away from having to maintain separate documents.
How important is the ability to access the manual’s information within existing Health Information systems?
It’s very important. As I mentioned earlier, these HIS groups would hand printed manuals to their developers and tell them to build software. With structured content and delivery of our XML through the API, software developers will parse our content directly from the source with minimal intervention. It is important that users in the clinical setting are able to pull up the staging manual on screen while they complete charts, so they don’t have to search for a book or rely on memory (or guess). If a clinician can record accurate information in real time, data collectors will not have to spend so much time combing reports for key words.
How would you sum up the principal organizational benefit(s) involved for your organization?
- Our content is our product. Our staging systems are the foundation of treatment decisions for cancer patients. Accuracy of content and accuracy of interpretation is critical.
- Single source content that is correct, every time it is pushed from the CCMS
- No searching hidden or lost files; no working in different files at the same time.
- Streamlined content management with checkpoints for draft, review, and publish. Governance structure for same.
- Hopefully will streamline SME work on future projects.
- Hopefully will reduce burden on small staff.
In hindsight, are there things you would have done differently?
I would have started the information architecture and implementation of the CCMS after the authors completed the manual. We would have known the entire scope of the content (templates + rogue unknowns) and more thoughtfully structured the XML. Not that the XML structure is bad! I just think it would have saved a lot of time on the part of Scriptorium and easyDITA staff, and myself. I am certain we went beyond contract hours, and I am eternally grateful for their patience as I sorted through our final content.
We began the search for structured authoring and CCMS experts 3 years before our authors commenced work on their chapters, with the expectation that a system would be in place before they began writing. By the time we made the business case and got the contracts together, the authors were already submitting manuscripts. I wish we would have waited just 5 months before kicking off. We could have spent more time on training and developing outputs to obviate the “failsafe” master files I’ve been keeping. The lesson? Know your stakeholders!
More on cancer staging: http://sxc.cancerstaging.org/references-tools/Pages/What-is-Cancer-Staging.aspx