Real World Evidence Evidence & Data Partnerships

Oct 14, 2014 - Oct 15, 2014, Bethesda

This year real patient data will change healthcare.

Real World Evidence: Interview with Pfizer's Marc Berger

Rochelle speaks to Marc Berger about real world evidence, the potential collaborations, challenges and new predictive tools that it might bring.



Rochelle Sampy speaks to Marc Berger, VP of Real World Data and Analytics at Pfizer. Prior to joining Pfizer about two months ago, Marc was the Executive VP and Senior Scientist in the life sciences group at OptumInsight which is part of United Health Group. Before OptumInsight, he acted as the VP of Global Health Outcomes at Eli Lilly and Company and VP of Research and Management at Merck & Co., Inc. 

“What does the rise of RWE mean for the pharmaceutical business and Pfizer?”

The pharmaceutical industry including Pfizer is focused on providing innovative medicines that will lead to better patient outcomes as well as value for the healthcare system. Real world evidence helps to inform and advance our pipeline by sharpening the value propositions for investigational compounds. Real world evidence highlights the intersection between unmet medical needs and the potential for innovation to meet these needs and provides value to patients, providers, and payers. In order to do that, we must answer three critical questions. If we get the answers right, we can significantly increase the probability of success.

The first question is: “What are the right targets [for drugs] based on our biologic understanding?” The second question is: “What are the right attributes needed for new treatments?” And the third question is: “Can we identify the right subgroups of patients who will benefit the most from new therapies?” There is an increasing demand for evidence of the advantages of new therapies, both with respect to comparative effectiveness and safety.

At Pfizer, I am trying to enhance the efficiency and effectiveness of real world data use through a number of ways. One strategy is to centralise the acquisition of real world data sets and create a Data Mart to facilitate access to this evidence to all appropriate groups across the Pfizer enterprise. I am also trying to create a venue to facilitate sharing and coordination of various efforts to use cutting edge methodology in our analysis of real world evidence.

“In relation to the patient subgroups that will benefit the most from new therapies, are there any diseases/sectors that Pfizer feels would benefit the most from RWE?”

Three areas that we believe will benefit a lot from real world evidence are oncology, autoimmune disease and pain.

“Will RWE increase the rate of collaboration between various stakeholders? Has Pfizer started any RWE collaborations?”

Absolutely. It has already started. We only need to look at the creation of the Patient-Centered Outcomes Research Institute in the US, which has a very broad set of stakeholders on its governance committee including academic institutions, pharmaceutical companies, payer, and patient groups. Another example is the Mini Sentinel [project] that is an FDA collaboration with a variety of payers to access their datasets on a real-time basis for the purpose of Pharmacovigilance. These are just two examples and they just  scratch the surface of collaborative activities. 

Pfizer has initiated several collaborations with real world evidence in mind. One is with Medco which was recently acquired by Express Scripts. This collaboration focuses on trying to effectively match patients with treatments that will benefit them the most. Another is with Humana to explore new ideas and ways to improve the quality of outcomes in the delivery of care for key populations. Most recently, we announced a collaboration with Humedica, a clinical intelligence company, to conduct data-driven collaborative research that will leverage Humedica’s expertise in gathering and normalizing de-identified healthcare data from disparate systems. The alliance is an important step in bringing together clinical informatics and life sciences expertise to better understand patient needs and the effectiveness of treatments to improve patient outcomes. Pfizer is not alone in this. I know that other pharmaceutical companies are building partnerships. We recognise that we are all in this together.

“Could you give me any examples of these other pharmaceutical companies using RWE well?”

I experienced one example of this while I was still working at Merck. At the time, Merck had just launched new oral asthma controller medication which was an alternative to using inhaled coritcosteroids in the management of the disease. In the clinical trials, the compound was effective but not quite as effective as inhaled steroids. But when we looked at its use in real world circumstances in observational studies, we found that there was an equivalent effectiveness between the oral and the inhaled treatments. This was probably due to some enhanced compliance with oral medications. That was a very important observation because payers were asking: “Why should we use a newer and more expensive medication?” The answer was that it had excellent effectiveness and better compliance; and it gave you a good value for money.

“How can researchers/ the pharmaceutical industry effectively balance RWE with other types of clinical trial research?”

Different audiences require different evidence to meet their needs. Randomized clinical trial evidence has generally been viewed as the only evidence sufficient to demonstrate to regulatory authorities that products are safe and effective enough to be used in the marketplace. Subsequently, RCT evidence is necessary to show that there are other indications that those products might be used. However, payers and patients have different needs and priorities. They want to know: “How does this treatment stack up against other treatments?”  Few standard RCTs are performed to answer this question.

Comparative information is more commonly obtained from observational studies including prospective or retrospective observational studies. Comparative information can also be obtained from indirect treatment comparisons, and models. This spectrum of evidence is currently incorporated into the evaluations of health technology assessment authorities around the world. The same spectrum of evidence is also used by pharmacy and therapeutics review committees within US managed care organizations. The demand for this evidence has increased and pharmaceutical companies ensure that it is available to appropriate health care decision makers so that they can assess the effectiveness, safety, comparative effectiveness and the value of different treatments.

“What is the future of RWE in pharmaceuticals?”

RWE is going to help us better target what drugs we should develop, how we develop them and how we bring them to the market in a precision medicine framework. We will answer questions around who will benefit the most or the least from different therapeutic alternatives; and who is most likely to get an adverse reaction [to a particular treatment]. This evidence is going to be critical in the future when we start seeing broad based adoption of risk share contracting between payers and pharmaceutical companies.

“Apart from RWE, are there any alternative data sources that aren’t currently be used but could be valuable in the future?”

There are a lot of efforts to capture real world evidence -- whether from a written medical chart or electronic medical records or lab data. There is a general belief that when this data is integrated, we are suddenly going to answer all the questions critical to manage individual patients and populations of patients. I suspect that this is not going to be the case. The process of gaining increased access to real world data is an incredible step forward and will allow us to better understand many things; it may help us to target the use of appropriate medication to specific patients or populations. However, the data recorded by health systems are far from complete. Anyone who has been a healthcare provider like I have in the past knows that not everything will be documented. Thus, there will be a major issue with missing data or sparse data that may impede our ability to answer important questions.  It may be necessary to systematically collect some information that could enhance the power of predictive models to project progression of disease or response to therapy. In the future, I suspect we will rediscover the value of health risk or health status assessments or PRO’s as inputs into predictive models.

“Are there any issues/factors that might prevent the successful development of RWE?”

The potential of the RWE can be realised only if electronic collection of health information is made available throughout the healthcare system. Challenges to rapid increases in availability include a relatively slow rate of adoption of EMRs, the lack of standardisation of data collection systems, and the lack of interoperability of electronic medical records with other electronic data sources. There are other appropriate concerns around how we guard patient confidentiality and ensuring that data is not only available to providers for quality improvement efforts, but also is made available to researchers to inform the discovery and development of new therapeutic approaches. This issue is one which is occupying the attention of government agencies, providers, patient groups and pharmaceutical companies. It is not a question of whether we will extract value from RWE but over what timeframe we will extract its full value.

“In your opinion, is there anything that patients, providers or governmental agencies can do to encourage the uptake of RWE?”

As the Institute of Medicine has suggested, we must evolve from the current state into a learning healthcare system. We need to be investigating what we are doing and what outcomes we are achieving while we are delivering care. The realisation of this vision will require integrated health information technology systems while protecting patient confidentiality. We will need to gain a societal consensus that while much information has to remain private, that other information and/or de-identified information can serve the public good. There isn’t an optimal flow of information today to drive the vision of a learning health care system forward. While there are legitimate and significant issues regarding the confidentiality of data and the purposes to which data will be used, we need to find the right balance including taking measures to ensure that confidential data is not used for commercial purposes.

“But it’s not all doom and gloom....”

These are very exciting times. I’ve been privileged to be working in this area for more than 20 years and where we are today is so much further ahead than where we were in the early 1990’s. I look forward to more progress in the future and I think we will be surprised to how fast it will come.


Marc Berger will be speaking at Real World Evidence USA 2013, for more information on his presentation or to find out what other discussions will be taking place, visit the official website.



Real World Evidence Evidence & Data Partnerships

Oct 14, 2014 - Oct 15, 2014, Bethesda

This year real patient data will change healthcare.