As the American Society of Clinical Oncology (ASCO) readies its request for proposal (RFP) to take the next step in creating its “big data” system—CancerLinQ™ (Learning Intelligence Network for Quality)—incoming ASCO President Peter P. Yu, MD, provided ChemotherapyAdvisor.com with a glimpse of what the next few months will bring.
The move into the full custom build follows the successful unveiling of the CancerLinQ prototype last March which was showcased at the ASCO Annual Meeting in Chicago in June 2013. Completed in just 8 months, the prototype included more than 170,000 de-identified medical records of patients with breast cancer provided by oncology practices throughout the United States.
“ASCO believes that this is a transformative step; that it can be of tremendous value to our members and our patients and to society at large,” said Dr. Yu, director of Cancer Research at Palo Alto Medical Foundation, in Palo Alto, CA. The primary purpose of CancerLinQ, a project coordinated by ASCO’s Institute for Quality, will be to improve data quality to improve outcomes. This, in turn, will “inform our guidelines process.” He pointed out that “guidelines are not meant to be prescriptive. Part of the purpose of the rapid learning system is saying that guidelines don’t always work, and finding that out so we can improve them.”
Dr. Yu said that the prototype demonstrated proof-of-concept on several levels: it was possible to bring real-time patient data electronically into a center data repository from different electronic health records (EHRs), interpret the data in a common language, mine the data to perform analytics, and provide feedback, such as physician performance against 10 quality measures from ASCO’s Quality Oncology Practice Initiative.
Peter Yu, MD
ASCO was “not intending to discover something brand new at this point,” rather, was seeking whether it was possible to provide correct factual information. The answer was very positively “yes,” he said. “Now, we need to actually build a test model or development model and eventually that will become the production model.”
Last July, ASCO issued a request for information (RFI) to determine whether it would be technically feasible to create a more robust CancerLinQ based on the prototype. Dr. Yu said that, despite the 2-week timeline to reply, one dozen companies, including Oracle, IBM, and Thompson-Reuters, provided favorable responses, which were reviewed by a committee outside ASCO that included those knowledgeable on informatics in general, medical informatics, medical oncology, and former officers of the US Food and Drug Administration (FDA).
This independent review confirmed that, although “no one’s done this before, your vision makes sense,” he said they were told. ASCO’s board of directors is using the RFI responses to create the RFP. A partner will be selected from the responses for the actual build of CancerLinQ.
In conjunction with this build, “we also are creating a whole new set of ASCO committees and subcommittees to look at the overall project, as well as data and data governance,” Dr. Yu said. “The data governance committee will be looking at data quality: how good is data coming out of EHRs? Can we actually use it?” Privacy, data access, and technology assessment will also be addressed, as will patients’ and physicians’ perspectives.