Big Pharma can’t handle Big Data on its own according to Priya Sapra of SHYFT Analytics.

Big Data, unstructured data – even Commercial 3.0. Whatever you call it, the hitherto undreamed-of scale of the information now available to pharma has the potential to change the way the industry does business.

In particular, there has been a lot of buzz around real-world evidence (RWE) for the last couple of years. RWE helps life sciences organizations understand how physicians use treatments in real life, and allows better comparison of actual drug effectiveness. It also helps with putting together clinical trial design and trial feasibility analyses, defining populations of interest and, in theory, speeding up the whole clinical operations process. In the R&D space, RWE helps determine which molecule to pursue, while in new product planning, the sheer availability of such data on millions of patients can help pharma understand unmet needs and which indications to pursue.

Observational research studies can prepare the groundwork for a new product from conception to launch, giving manufacturers the ability to understand and target specific patients, identifying exactly who the ultimate end consumers are. In a payer world which is thinking very differently about value now to the way it was a decade ago, there is demand for more evidence of cost-effectiveness well past initial reimbursement. This could also be a factor in incentivizing pharma sales reps in a manner which reflects patient outcomes.

Revitalising analytics

This is quite a list. “RWE has revitalized data analytics across the life sciences industry,” says Priya Sapra, Chief Product Officer of SHYFT Analytics.

So much for what is available – when it comes to translating this mass of data (including RWE) into actionable strategic decisions, there is a catch. “Big Pharma can’t handle it on its own,” Sapra insists. “They can procure the information but they can’t transform it into something usable. This is a different competency: companies are going to be looking at who is the best partner for them and they are asking for help now.” Companies see the clinical and commercial possibilities of data “but they want an integrated view of how to leverage it”.

Pharma has not yet been able to drive better patient outcomes via its use of RWE – despite the fact that this must surely be the goal. “We’re going to have to see some consolidation of data and therapy-specific analysis to see which data is best – if there is such a thing,” she says. “We also need to provide solutions with more streamlined data: pharma needs solutions, not services.”

Bigger pharma companies are making some moves in this area, bringing together cross-functional internal teams which take in representatives from, say, HEOR, medical affairs, analytics and safety to look at how to evaluate what is out there.

But Sapra thinks some smaller and mid-cap pharma firms and biopharma outfits may be overwhelmed. For a start, cost is an issue when accessing data – and then there is the question of what the best dataset to use might be, when you are on a relatively limited budget: electronic health records versus insurance claims versus patient registry information, for instance.

‘Best’ dataset?

In fact, Sapra, who has a Master’s degree in Epidemiology and Biostatistics from Boston University, believes that the idea of a ‘best’ dataset is misleading. “You need the juxtaposition of multiple sources to make sense of the whole puzzle,” she says. Yet for pharma to do this – in most cases at least – collaboration is going to be vital, she thinks.

Sapra sees three fundamentals for RWE success: data strategy, comparative analytics and technology. “Data strategy must be robust and vigorous,” she begins. “Companies must know how to take different data and have it speak a common language. All RWE providers currently have their own structures, their own language.”

She compares it to having information in French, Spanish and Italian which must be put into English. In effect, this involves gathering it into an industry-standard data platform which allows it to be examined in a consistent fashion, comparing like with like. “You need to aggregate it, translate it and place it into this platform,” she says.

Pharma has not been quick to uptake. “In some areas, I’m not seeing any data strategy at all,” she continues. “This means there is no way of doing comparative analytics across data sources, comparing apples with apples. The industry understands that these type of analyses need to be done. The mind-shift is occurring, but it is primarily talk at the moment.”

Disparate sources

She insists that companies “need to be comfortable with” taking data from disparate sources and making something coherent from them. It goes back to her point about there being no single ‘best’ data source, simply a series of perspectives which need to be married up to get a full picture.

“The truth lies somewhere in between,” she says. “This is the acumen– the critical part of the solution. By aligning all this data we can look at subsets of patients and align treatment algorithms with patient outcomes and annual cost of care.”

The proper, integrated use of RWE could start to make a real commercial difference to companies looking to gain an edge over rivals. It might allow payers and providers to identify, for instance, where the use of a more expensive medicine up front can reduce overall costs. “Pharmaceutical companies are now allowed to say: ‘When my product is used in this subset of patients, it is a better fit because they are less likely to be hospitalized or suffer other complications and will have a better overall quality of life’.”

This is, finally, where technology comes in: selecting what data is important and presenting it in a way that creates a visual platform in order to deliver insight from that data. “RWE gives us a true understanding of the patient journey, bridging the knowledge gap,” Sapra continues. “We can see from diagnosis to treatment to outcome and cost of care.”

In this way, data-driven analytics and insights can help bridge the knowledge gap around value. “By getting more – and better – information in a more timely manner, we can understand which product is allowing the best outcomes for patients at the best value for payers – driving market access decisions.”

Potential savings

She believes SHYFT Analytics’ platform can help pharma companies make annual savings of $6million. To explain this eye-popping figure, Sapra outlines a scenario in which a pharma company is planning some observational research: the clinical researcher pinpoints that they want to investigate a particular cohort of patients and asks an analyst to identify the individuals who will fit the profile. The analyst finds a subset – but based on the size of the population, the researcher now realizes they need a larger group of patients with less restrictive criteria. So it goes on, the project batted back and forth in this iterative format: they may eventually get the information to publish a paper or develop a trial design, but Sapra argues this is an expensive, laborious process taking many weeks.

“Our product puts all the power into the hands of the clinical researcher, who can build and iterate on the cohort they want, define the variables they need, and conduct the analyses they require, such as survival analysis, regression analysis, correlation analysis and so on, enabling them to conduct an end-to-end analysis,” she suggests. “Six weeks becomes six days.” Hence, she continues, the $6 million savings. “And that figure is before you look at the revenue that can be generated.”

Another of the major issues around Big Data is storage: quite apart from not always having the know-how, companies simply do not have the space required for large-scale data warehousing.

Cloud storage

Sapra is a great believer in the power of the cloud to help pharma negotiate the challenges presented by the sheer weight of data available. “It is more secure and it’s a better way to store the depth and breadth of data we are talking about,” she says. “The cloud answers a lot of those problems of data scale.”

In the future data itself is, she thinks, likely to become more commoditized – i.e. cheaper. Yet given the sheer volume of data available, surely the threat of data overload is real? “Good question,” she says. “’Yes’ is the simple answer. Just looking at terabytes and terabytes of data is not going to solve anything. That’s why you need partnerships with organizations who can take that information and make it into something intuitive and useful.”

So data itself is certainly not the issue – there is always loads of that about, and it has never been more accessible. “But people  really have no idea what to do with it,”laughs Sapra. “That’s fine, it’s not their job. They need to evaluate the results and take action – and we are here to facilitate that.”

She concludes: “And what’s so great is that finally, we can understand outcomes for patients align them with the cost of care – it’s a win-win: it changes the way treatments are developed, are prescribed and paid for – allows for much more optimal personalized medicine and a complete understanding of the patient journey.”