DBMI Alum in the news - Article in Health Data Management titled "Data sharing critical to success of Cancer Moonshot” includes interview with Doug Fridsma

From: Health Data Management
By: Greg Slabodkin
Published: September 13 2016, 7:07am EDT

Data sharing critical to success of Cancer Moonshot


A national cancer data ecosystem must be created to make data more accessible and to accelerate medical breakthroughs to ensure that the Cancer Moonshot can realize its ambitious goal of doubling the rate of progress in preventing, diagnosing and treating the disease.

That’s among the recommendations a blue ribbon panel made last week to the National Cancer Advisory Board and accepted by National Cancer Institute Acting Director Douglas Lowy, MD.

“The bold but feasible cross-cutting initiatives in this report will improve outcomes for patients with cancer, prevent cancer and increase our understanding of cancer,” said Lowy. “NCI stands ready to accelerate cancer research in the critical areas identified by the Blue Ribbon Panel.”

The goal of the Obama administration’s Cancer Moonshot, led by Vice President Joe Biden, is to make a decade’s worth of progress against cancer in five years. Toward that end, the panel recommended 10 scientific approaches, including the creation of a “national ecosystem for sharing and analyzing cancer data so that researchers, clinicians and patients will be able to contribute data, which will facilitate efficient data analysis.”

As the panel’s report found, although “many powerful sources of data are being rapidly generated in the research and clinical communities,” they are “currently not being fully leveraged, because of the pace at which data are accruing, the lack of a coordinated effort to assemble the data in readily accessible fashion, and the inability to effectively process the data in an inter-operative fashion.”

Their scientific roadmap creates a vision for the future of cancer research and treatment in which:

  • Patients contribute their data, obtain genomic profiling information about their tumor, learn about what treatments might work best given their tumor’s genomic profile and find other relevant information, including clinical trials that may be appropriate.
  • Researchers can identify possible targets for the development of new treatments and preventive interventions, including immunotherapies, as well as learn more about how to avoid or counter drug resistance.
  • Doctors have access to information that better predicts treatment outcomes and helps control patients’ symptoms and side effects.

The American Medical Informatics Association applauded the panel’s recommendations, particularly their call for development of a national cancer data ecosystem.

“The continued digitization of clinical data, combined with the explosion of molecular information and patient-generated health data, create tremendous opportunities to make new discoveries and develop new treatments for cancer,” said AMIA President Doug Fridsma, MD.

At the same time, Fridsma noted a central tenet of the proposed data ecosystem is that patients directly contribute their data or request that a provider do so on their behalf to further scientific research.

“However, the vast oceans of data available in electronic health records remain locked inside systems incapable of transmitting more than simple care summaries,” he observed. “Without access to complete medical records, patients will not be able to fully engage with the Cancer Moonshot, precision medicine and other big data research efforts. And while we desperately need national leadership to address challenges related to infrastructure, interoperability and standards, we must first fuel this ecosystem with computable data contained in EHRs, so that free text remains free text and structured data remains structured.”

The panel acknowledges that despite the advent of EHRs “cancer patients know all too well the difficulties of accessing their own medical history, transferring their records from one provider to another, or integrating information among providers, for example between a primary healthcare provider and an oncologist.”

Their report points out that “there is no single national infrastructure to enable straightforward contribution of data from patients or the receipt of personal health data” and that “a national cancer data ecosystem will both host and encourage the development of the necessary computational tools and infrastructure to enable optimal cancer data sharing and utilization among researchers, clinicians, patients and the public.”

Nonetheless, while Fridsma concedes that the panel’s report “rightly identifies numerous barriers to our collective capacity to capitalize on these new opportunities,” he asserted that “without a digital ‘print all’ functionality—made available through all certified EHRs—we will continue to fall short of our aspirations.”

Fridsma contends that once comprehensive patient data are available for searching, indexing and parsing, developers will find ways to make the data useful for researchers, clinicians, as well as patients.

“Proof of this concept was demonstrated by Apple when they developed a way for patients to import CCDs into HealthKit—but that was only a limited subset of patient data,” he concluded. “Think of what might be possible if patients could import all of their data.”


Link to this article

Short writeup

Longer writeup on the Data Ecosystem

Longer recommendations for all the working groups

Panel members

Working group members

Post Date: 
Friday, September 16, 2016