A physician at the Dana Farber Cancer Institute in Boston is building a database to advance research into treatment for metastatic breast cancer.
His idea is to go beyond anonymous data and include patient narratives about their experience during treatment to identify patients who had extraordinary responses to treatments. Finding these “lucky” patients whose tumors had unusual mutations that created positive responses to specific drugs will lead to more focused research. He is reaching out via social media, online forums and advocacy groups to recruit patients who will give permission for him to incorporate their medical information as well as their personal experiences into the database to allow researcher follow up if needed.
I was fascinated to learn about the doctor’s approach that recognized that the answer to new treatments may lie in the stories of a few patients who reacted in unexpected ways to treatment.
There were also a number of other points raised in the New York Times article, “New Cancer Treatments Lie Hidden Under Mountains of Paperwork,” describing Dr. Nikhil Wagle’s journey that caught my eye.
The reporter noted that private companies do collect patient information to build large databases for research purposes and described the efforts of one company that has collected 2.2 million records of cancer patients.
I was shocked to learn that the technology company hired 900 nurses and certified tumor registrars to code the data from the 2.2 million medical records and put it into a usable form rather than find a technology solution to the problem. The company representative notes that the need for these people is due to the amount of unstructured information in a medical record.
Databases developed by private companies are not helpful for Dr. Wagle because the information is de-identified, but he is also finding the amount of unstructured information and the variety of formats for medical records to be a challenge as his staff deals with faxed records along with those sent via snail mail.
Extraction of data from medical records will be easier when – or if – there are universal standards that apply to all healthcare records and all providers. In the meantime, there is technology that can aggregate and harmonize data from disparate sources.
Platforms that aggregate information – structured and unstructured – from previously siloed sources in all formats harmonize the information to make it accessible to researchers. The use of natural language processing and scalable platforms that can grow in capabilities as needed provide the support that researchers need for a robust database.
Technology can handle complex, high-volume data needs for mega-trials as well as lower volume needs for focused research. Relying on aggregation technology to put data into a usable format frees clinicians, researchers and critical personnel to focus on finding the best treatment for patients rather handling coding or data cleaning tasks.
Unfortunately, access to data is a major barrier for many health and life sciences researchers. The need to collect and normalize data from different sources and in different formats will only continue to increase as mobile devices, personal health trackers and new software applications collect data.
Tapping into these sources of information to identify new treatments for cancer and other diseases is essential for continued innovation.
Dr. Ned Sharpless, director of the National Cancer Institute was quoted in the same article, saying, “Data are trapped. This is a huge problem.”
What about your organization? Are you ready to take the first step to solve the problem?