CancerLinQ project called part of 'learning health system'

The American Society of Clinical Oncology is again trumpeting a software development project that it says will one day use routine clinical data collection and real-time data analytics to drive members of that medical specialty toward the Institute of Medicine's goal of creating a learning healthcare system.

The society and the Conquer Cancer Foundation co-hosted a second coming out party today in Washington for CancerLinQ, an open source and proprietary software development project that ASCO CEO Dr. Allen Lichter first launched on its “maiden voyage” at a Dec. 1, 2012 Quality Care Symposium in San Diego.

The project is still in the prototype stage and will likely remain so for another year or more, according to an ASCO spokesman, but it “demonstrates, for the first time, the feasibility of a HIT-based learning health system, which the Institute of Medicine has defined as critical to the future of the nation's healthcare system,” the ASCO said in a news release.

Last fall, the IOM issued a 382-page report, Best Care at Lower Cost: The Path to Continuously Learning Health Care in America, that said “available knowledge is too rarely applied to improve the care experience, and information generated by the care experience is too rarely gathered to improve the knowledge available.” It called for better use of health IT to close those knowledge gaps.

Dr. Sandra Swain, president of ASCO, said at the presentation that only 3% of cancer patients participate in clinical trials.

“So that means, that's the data we use for our evidence-based guidelines. Our vision is to bridge that gap.”

Eventually, the system will capture data from millions of patient records, Swain said.

Having access to those records will be “like having the whole medical community available for a second opinion,” Swain said. It “redefines what a second opinion is.”

Swain and Lynn Etheredge, director of the Rapid Learning Project at Georgetown University, both served as panelists at the CancerLinQ presentation today.

Etheredge, a member of the IOM committee that developed last year's report, said in the news release that in the oncology community, “the big data transformation of healthcare is well under way. The future of our healthcare system hinges on the success of initiatives like this—and on the willingness of healthcare providers, payers and policymakers to support their implementation.”

The prototype system, created from “several open-source IT applications,” has been loaded with the de-identified data of more than 100,000 breast cancer patients treated at various cancer treatment centers across the U.S., according to the ASCO. It can “accept cancer care data in any standard, directly from the electronic health records and other sources” including genomic profiles, lab test results and physician notes. It also can generate individualized guidance on the care of any given patient with breast cancer, based on automated, machine-readable versions of ASCO's expert breast cancer guidelines.

“When implemented, the full CancerLinQ system “will also provide guidance based on the collective experiences of other similar patients,” the ASCO statement said.

Its data mining, analytics and visualization capabilities will enable clinicians to explore “an extensive database of information on the care of patients with breast cancer ... to identify real-world trends and associations and new hypotheses for research.”

CancerLinQ also will provide physicians with feedback on their performance against 10 ASCO-developed clinical quality measures, the association statement said.

The ASCO announcement follows one by IBM last month that it was developing, in collaboration with physicians at Memorial Sloan-Kettering, a clinical decision-support system using its Watson supercomputer and another by researchers at Indiana University that an artificial intelligence-based decision-support tool produced, hypothetically, improved outcomes when applied retroactively to the medical records of a panel of patients with multiple chronic conditions.