Real World Data Europe

Apr 28, 2014 - Apr 29, 2014, London, England

Demonstrate the true effectiveness of your drugs to satisfy payers, HTAs and improve patient outcomes

Real World Data: Promises and Challenges

Big data, big promises, but getting the complete picture needs multiple sources of information. Real World Data (RWD), randomized clinical trials (RCTs), and small-scale, focus-group-based surveys are not members of opposite teams, but pieces of the same puzzle.



Although the debate about the use of RWD is heated, it seems that the controversy is missing the point. RWD will not replace information gathered in more traditional ways, and a savvy way of combining multiple sources of data is needed to give a more rounded patient insight.

“This is the challenge with big data: You get information on what is consumed, when, and how often, and from that you can easily figure out that something is not done the way it should be, but the drivers remain unknown,” admitted Oliver Mast, Head of Global Market Access, Roche Diagnostics GmbH, in his account of how Roche turned information from a variety of sources into a successful therapy solution.

Multiple Insights

To find out more, Roche conducted a small-scale, detailed survey, which revealed that noncompliance was due to the level of disruption caused by having to control their blood glucose levels.

To understand what’s going on in the lives of their patients, Roche have adopted different survey approaches. One of them was asking all patients starting their diabetes therapy detailed questions about their behaviours. Another way was buying big data, e.g. in China the company gained access to 10,000 individuals, which allowed them to analyze in detail currently available treatments.

What did they discover? Roche support insulin regimen, which requires glucose monitoring, where a patient needs to test their blood sugar level three times a day for optimum results. Using RWD, Roche found that patients often neglected the prescribed measurements. On its own, however, that information wasn’t useful. To find out more, Roche conducted a small-scale, detailed survey, which revealed that noncompliance was due to the level of disruption caused by having to control their blood glucose levels. Equipped with those insights, Roche was able to integrate and optimize technology in a way that made the process easier. “Every measure used to take about 40 steps, now it takes nine. It’s a big difference,” Mast said. Still, though, the company didn’t feel like they knew it all. The next step was to verify whether the intervention had a real impact. “We then did a clinical trial, which has proven that this technical innovation and process improvement does also result in better outcomes,” Mast explained.

Big challenges

In this case, information from a large-scale study was supplemented with a small-scale survey before designing a successful intervention was possible. But a fragmented picture of the overall situation is not the only challenge one has to face when working with big data. “When you analyze real-life data, you need to do a thorough analysis on what the potential sources of bias are,” Mast explained. “You always need to ask yourself where does this data come from, who’s taking part in the survey, who could be missed potentially,” Mast stressed, adding that even though the data has some limitations, so does information gathered in RCTs, where samples are often too small to accurately represent patient population. “Additionally, RCTs happen under very specific conditions, with particular attention to the patient and a huge amount of documentation, but it has little to do with life outside of the lab,” Mast elaborated.

Big skepticism

Insights into the real-world dynamics might make reimbursement more likely in the future. In times of tight medical budgets, selling something is dependent on demonstrating that there is a robust problem your product can solve, and to do that the given phenomenon needs to be identified and measured. “Data allows us to qualitatively and quantitatively describe problems, e.g. 70% of patients don’t meet their therapy goals, 50% of patients don’t test as often as they should, 30% of patients don’t follow the recommendation of their doctors. This is data that we can only get from real-life sources, from day-to-day therapy. The more demanding the therapy is, the bigger the need to support it with adequate research that reflects those challenges”, says Mast.

At the same time, using RWD enables creating predictions more accurately than those based on RCTs. Mast elaborates: “Data can be input into health economic models, which estimate the impact of a specific therapy on specific payer budget. Obviously when you use real data for a disease model to project real life economic implications it might be more reliable than when you use clinical trial data generated in an environment that does not have much to do with a real life situation.”

Unfortunately, as of right now, payers are skeptical about the value of RWD, and, in the health technology assessment process, RCTs remain unchallenged where there is need to prove benefit. “I spent many years in pharma working on compliance issues and I remember one very important example where after a year only 50% of patients remained on the therapy they had started, even though the RCT showed that the 4-year drop-out rate was 10%,” Mast recounted, admitting that even though the importance of RWD seems obvious, payers (and the industry) have to agree on what the value of it is, and how to deal with challenges that come with it.

“There’s a heavy debate about the relevance of RWD. On the one hand, there are some ideological considerations. As long as you insist on the research paradigm of control, it’s more difficult to accept observational data. I’m a true believer in that we need RWD to complement RCTs”, concludes Mast.

Comprehensive insight into the life of patients requires thorough research that cannot be limited to just one source of information. If novel therapies are to be successful, they need to be examined from all angles, including finding out their clinical efficacy identified in RCTs, real-life adherence to treatment from RWD, and barriers to health from small-scale panels.


Oliver Mast will be speaking at Real World Date Europe in April. For more information on his presentation, click here. 



Real World Data Europe

Apr 28, 2014 - Apr 29, 2014, London, England

Demonstrate the true effectiveness of your drugs to satisfy payers, HTAs and improve patient outcomes