Key Takeaways:
- AI is automating the registry’s most repetitive work: Reportable-only casefinding and data extraction can now run in real time, freeing oncology data specialists to spend their time on validation, quality, and analysis rather than sorting through non-reportable reports.
- Real-time data turns the registry into a single source of truth: When case capture happens concurrently with abstraction, the same oncology data can inform screening, nurse navigation, clinical research, and tumor boards across the enterprise.
- Technology amplifies registrars; it does not replace them: Human expertise remains essential for validating AI outputs and interpreting what the data means for patients, programs, and decision-makers.
- Validation and accountability separate real solutions from hype: Registry leaders should test AI tools against their own cases and require vendors to stand behind accuracy claims.
Why the Cancer Registry Is Ready for AI
Many cancer programs still run on disconnected systems, where case identification and reporting are separated by manual handoffs. That fragmentation creates persistent backlogs, slows the data down, and leaves revenue on the table. For the modern cancer registry, the question is no longer whether to automate the repetitive parts of the workflow; it is how to do so without losing the expert judgment that makes the data trustworthy.
In a recent webinar hosted by Inspirata, Connie Boone, BS, AAS, ODS-C, Manager of Clinical Data Management at Prisma Health and a National Cancer Registrars Association (NCRA) Professional Development Board Director, joined Biff Curtis, Vice President of Sales at Inspirata, for an open discussion on where oncology data is headed. Their conversation was framed around NCRA’s 2025–2029 Strategic Management Plan and its four “future statements” for the profession.
The Bottleneck: Backlogs, Shortages, and Manual Casefinding
Healthcare has lagged most other industries in adopting AI, and for understandable reasons: lives are on the line. A safe and prescriptive place to start is with repetitive administrative workflows, and few areas carry a heavier daily burden than cancer registry operations.
That burden is intensifying. Case volumes are growing while the pool of skilled, certified registrars and ODSs is shrinking, and recent reductions in federal funding have pushed health systems to do more with less. The traditional fix, hiring more people, is no longer realistic when there simply are not enough certified specialists to hire.
When a team is stuck clearing a backlog, Boone explained, there is no time left for the higher-value work: reviewing cases for educational needs, improving quality, or analyzing trends. The registry stays in survival mode. That is the gap that automation is meant to close: not by removing people, but by removing the repetitive tasks that keep them from practicing at the top of their license.
What End-to-End Automation Actually Does
The distinction that the speakers drew most sharply was between true casefinding and a report dump. A general, non-oncology tool may simply surface every report that a casefinder still has to review manually, shifting the burden rather than removing it. Real automation identifies reportable-only cancer cases, so a registrar can quickly validate and move to abstraction.
According to Inspirata, its AI engine is built for exactly that. The company states that the engine surfaces reportable-only cancer cases, pre-populates up to 120+ NAACCR data elements for pre-abstraction, and surfaces up to 500 real-time oncology data elements to inform broader oncology operations. Inspirata has spent 23 years serving oncology registries; the company was formerly known as AIM (Artificial Intelligence in Medicine) and previously held an agreement with the NCI’s SEER program, with a footprint of more than 600 facilities. Its current client partners range from roughly 400 reportable cases a year to more than 130,000.
On the data-handling side, Inspirata says the engine is deployed on-premises, on the client’s own server, so protected health information never leaves the client’s environment and is not stored or reused. Integration happens either through export-import with the cancer data management system (CDMS) using XML, or through full API integration with an integration partner so registrars can work inside a single application rather than logging into five or more systems.
From Cost Center to Source of Truth
For decades, registry data was collected largely for reporting and compliance, and the registry itself was often viewed as little more than a cost center. The panel’s central argument was that real-time data changes that calculus. When reportable-only casefinding runs in real time, the data can be used immediately, before an abstract is complete, and shared across the enterprise.
Boone described how Prisma Health uses registry data to advance screening equity. By reviewing late-stage diagnoses by disease site and geography, her team identified that needs differ markedly across South Carolina markets: one area showed high-stage cervical cancer while another saw almost exclusively low-stage disease. That information helps target screening where it matters most. She also pointed to a multi-year Prisma initiative to shorten the breast biopsy-to-surgery window from the 60-day standard toward a 45-day target, using registry data to find and remove the barriers keeping patients from meeting it.
Curtis framed the same data through a business lens: real-time oncology data supports targeted screening and prevention (patient acquisition), faster entry into treatment pathways and clinical research (patient retention), and reduced leakage to competitors. Eliminating duplicated casefinding across nurse navigation, clinical research, and the registry, doing it once in real time for the whole enterprise, is where he sees both efficiency and downstream revenue. He used a figure of roughly $200 million in downstream revenue only as an illustration of how meaningfully accurate, timely data can move the needle, not as a measured result.
Amplifying Registrars, Not Replacing Them
Both speakers returned repeatedly to one point: the best tools do not replace registry expertise, they amplify it. AI can reduce the repetitive work of deciding whether a case is reportable, but human specialists are still required to validate what the tools surface and to interpret what the data means. As Boone put it, the expertise that earns and maintains the ODS credential is exactly what cannot be eliminated.
That mirrors NCRA’s strategic vision. Its 2025–2029 plan envisions a future where providers and policymakers actively seek registrar expertise; where registrars master new, transferable skills as manual tasks decline; where AI and technology sit at the forefront of the profession with registrars driving the change; and where registrars are respected and compensated for interpreting the cancer patient’s journey into vital oncology data.
How to Evaluate an AI Vendor Without Getting Burned
With so many vendors attaching “AI” to their messaging, the panel offered practical guidance for cutting through the noise. Curtis cautioned that some tools promise far more than they deliver, offering “half of X” today with the rest promised in five or ten years, and that general platform tools often claim oncology-specific capabilities they cannot actually fulfill.
Two safeguards stood out. First, test the engine against your own data: Inspirata described a web-based tool that lets evaluators run their own reports, for example, 50 reports that the registrar has already classified, against the engine and compare speed and accuracy directly, apples to apples. (Inspirata says 50 reports run in about 11 seconds, and cites 99% accuracy over its 23 years, figures registry leaders should verify against their own cases as part of due diligence.) Second, require accountability in writing: ask any vendor to put performance requirements into the contract. As Curtis framed it, an unwillingness to stand behind accuracy claims should be a deal-breaker.
The Bottom Line
Automated, reportable-only casefinding and real-time data extraction are reshaping what the cancer registry can be: faster, more current, and positioned as the single source of truth for accurate oncology data. But the technology is only half the story. The other half is the expert who validates the output, interprets the context, and carries that story to the people making decisions about patient care.
If you’re interested in watching the full webinar to learn more, you can find it here.
Have a scientific story like this one to tell? Samba Scientific helps life science and biotech organizations turn complex science into clear, credible content. Explore our webinar services or get in touch to start the conversation.
