ASCO Provides Early Look at CancerLinQ Big Data Initiative

Allen S. Lichter, MD

At the 2015 ASCO Annual Meeting, the much heralded CancerLinQ big data system for helping oncologists more clearly understand treatment patterns and options was offered for demonstration in advance of its rollout.

In a press conference, Allen S. Lichter, MD, chief executive officer of ASCO, said that once fully up and running, CancerLinQ would provide data that was “logical, compelling, and searchable,” and would, eventually, become an indispensable tool, enabling doctors to improve quality and understand how they are performing in comparison with others across the health industry. “Physicians will say ‘I cannot practice without this,’” he predicted.

CancerLinQ has four chief capabilities: continual performance tracking for comparison with clinical quality measures; trend evaluation based on anonymous patient data; patient cohort identification based on shared characteristics; and individual patient timeline construction based on treatments, side effects, and outcomes.

ASCO aims to have 15 practices using CancerLinQ before the end of the year. Twelve practices across the country have already signed up to share data through the system. They range in size from six practitioners to NCI comprehensive centers. With 15 practices, CancerLinQ is expected to have around a half million patient records, Lichter said.

“CancerLinQ will soon become a reality for trailblazing oncology practices, and their patients will see its impact right away,” ASCO President Peter Paul Yu, MD, said in a statement.

“Our vision for CancerLinQ is ambitious—it is for every patient’s experience to contribute to the most compassionate, effective, and sustainable cancer care possible,” Yu said. “This first version offers a glimpse at the potential for big data to make cancer care better and more seamless. But it’s just a taste of the future, in which every patient’s care will be guided by up-to-the-minute science and data-driven insights.”

Giving examples of system capability, Lichter said CancerLinQ can provide “longitudinal” perspectives on individual patient cases, enabling doctors, for example, to very quickly find out how a patient within their own practice was treated months earlier, something that might otherwise have involved “a great deal of time flipping through charts, a page at a time.”

Similarly, patient status can quickly be plotted within a map of care guidelines so that doctors can know whether or not a patient falls within the “window of opportunity” for certain treatments. Practices will also be able to drill down into their own experience with patients and determine whether their findings are at odds with what other practices have encountered, Lichter said.

“Most quality programs are tests,” Lichter said. “What we’re trying to do is say everybody ought to get 100%—not do a ‘gotcha’ moment but do a ‘help ya’ moment. If you can move this patient along you will achieve 100% compliance.”

CancerLinQ will also provide physicians with “real-time” comparisons of their performance against that of other doctors, Lichter said. In addition, physicians can query the database to gain insight into treatments that may or may not work, Lichter said, citing the hypothetical case of a 70-year-old patient with neutropenia, the danger that represents, and the dilemma over the type of chemotherapy to use.

A prototype of CancerLinQ was established in 2013, when ASCO announced it had begun creating a depository of breast cancer information involving a variety of data from different sources. Too much information on cancer patients was locked away in unconnected servers and paper files, depriving physicians of valuable insight into the nature of cancer and its treatment, ASCO said at the time.

CancerLinQ was also touted as a prototype for the big data projects that are entering the mainstream of medicine. Data from 100,000 breast cancer patients who were treated at cancer centers around the United States were gathered. The CancerLinQ system was designed to accept data in multiple formats on a real-time basis, overcoming the hurdle of inconsistent health data standards. The data that was collected ranged from genomic profiles to lab tests and even physicians’ notes.

Another feature of the system was its capability to provide feedback for physicians’ performance based on 10 quality measures from ASCO’s Quality Oncology Practice Initiative (QOPI), a tool for care assessment and quality improvement.

Participating practices include MedStar Washington Hospital Center, Washington Cancer Institute, Washington, DC; Marin Cancer Care, California; Montgomery Cancer Center, Alabama; New England Cancer Specialists, Maine; INOVA Medical Group, Virginia; Medical Oncology Hematology Consultants, Delaware; Southwest Centers for Cancer Care, Massachusetts; Zangmeister Center, Ohio; Michiana Hematology Oncology, Indiana; Space Coast Cancer Center, Florida; Cancer Treatment Centers of America, Arizona; and Catholic Health Initiatives, Colorado.

The CancerLinQ platform will be powered by big data technology called HANA, offered by SAP, a global software developer. The HANA software can manipulate the data that are collected in a number of ways. All data are stored in main memory, which enhances the speed to analyze the data and can provide the real-time insight that a busy oncologist needs when he or she is seeing a patient. In addition, HANA can collect and adjust data that are collected from multiple sources (ie, different community oncology practices).

Over time, further development and input will come from physicians, patients and experts in relevant disciplines that include quality improvement, health outcomes, epidemiology, and health IT. ASCO will continue to maintain control over the data, services, and products, including decision support tools and analyses. SAP will provide engineering and technical support to continue to enrich CancerLinQ’s versatility.

<<<

View more from the 2015 ASCO Annual Meeting