June 22, 2016
By Alex Keown, BioSpace.com Breaking News Staff
ATLANTA – As big data continues to become a larger part of the pharma and biotech industry’s arsenal in developing new therapies, companies should take heed of three data risks facing the industry.
Gary Palgon, vice president of healthcare and life sciences solutions at data management company Liaison, said currently pharma and life sciences companies spend approximately $100 billion annually on big data strategies. While there is a tremendous amount of money being spent on big data, Palgon said “value creation” has been stunted “due to roadblocks in applying integrated analyses for accurate insight due to skewed or crucial missed information—halting both patient and industry financial advancements.” The more active data a pharma or biotech company has, means that company has more insight into developing drugs or devices, Palgon said in an exclusive interview with BioSpace.
The three data risks, or perhaps more appropriately, three data hurdles, that face the industry are security, data integrity and data interpretation.
“There is a risk with all the patient source data, therefore how companies address those challenges become bigger,” Palgon said. “At the end of the day, we’re taking advantage of historical pieces of data… Most conversations about big data goes to technology, but the truth of it is we’re all in an infancy stage learning how to aggregate the data and understanding what we’re going to do.”
Compliance includes security and privacy concerns, which requires stringent oversight of the data.
When it comes to data integrity, Palgon said traditionally there has been a lot of scrutiny over data included in clinical trials or other studies. However, now there is a tremendous amount of data coming from multiple sources, which means pharma companies have to protect the data in order to properly validate it. The storage of data must include its raw form so the data lineage can be audited should it be necessary, he said.
Multiple companies are incorporating big data into their product research and development, including tech giants Google and Apple . Google’s life sciences division, Verily, is undertaking a billion-dollar study called Baseline, part of its precision medicine initiative to “understand what it means to be healthy, down to the molecular and cellular level.” There are some questions about the feasibility of the study, but it is a good indicator of how big data can be used to drive wellness.
One company looking to harness big data that could serve as a model for Palgon’s warning, is the Palo Alto, Calif.-based blood testing company Theranos.
“That’s a good example. Theranos came out with a new device to transform the way testing was done. In the end, it wasn’t as validated as they thought, which leaked over into data integrity issues,” Palgon said, referring to the fact the company was forced to invalidate two years’ worth of data earlier this year.
When it came to analyzing the data, Palgon said Theranos made missteps, which has hurt the perception and bottom line of the company.
“Theranos would be a hard lesson to learn for the industry. We all need to keep the learning in check. We can’t jump to conclusions and we have to follow a well-thought out process or people will literally die,” Palgon said.
Proper interpretation of big data will require education of data scientists in order to make proper use of the data, Palgon said.
“Most data scientists spend 80 percent of the time trying to cleanse the analytics. We need to flip the model so they can spend 80 percent of the time analyzing the data. If scientists will gain insight into taking advantage of the data lake, we need to make sure they’re utilizing and seeing all the data sources and interpret it the right way,” he said.
As the reality of big data becomes more engrained in the pharmaceutical and biotech industries, it will serve as a hiring driver. Palgon said the industry will need an “entire ecosystem of people to manage the big data innovations.
Among those roles required are software developers for the databases and skilled employees who can interpret the amounts of data being generated and, of course, those who understand data security. Palgon said other employees in demand will be those who know how to implement the data.
