Friday May 15, 2020

This interactive discussion featured Harry Glorikian, Chair of AI in Biopharma Summit and author of MoneyBall Medicine joined by Ülo Palm, Thomas Bock, and Lixia Wang, who were open to discussion and debate about topics including:

• The definition of RWE
• Examples of companies that are effectively working with biopharma to ensure that RWD is thoughtfully collected and analyzed to turn it into a powerful tool
• The ability of RWE to predict future pandemics
• Challenges to adoption
• Privacy concerns
• The potential over reliance on p-values
• RCT’s (random clinical trials) limitations on drug efficacy in big populations
• Working with regulators

Below you’ll find selected talking points, information shared and the recording to watch the full webinar.

A productive community has several key components: a shared purpose, a willingness to assist others in the group, a gathering space to learn from each other, and an ability to move forward as a group toward a common goal. On Friday, May 15, 2020 the AI Applications in Biopharma Summit community came together in a virtual space to engage in a discussion about how best to leverage RWE and advanced analytics for drug development in a pandemic and post-pandemic.
Each speaker generously gave their time in preparing for the webinar and shared their thought leadership by providing specific examples of effective uses of RWE, in combination with AI, to manage the quantity of data available during a pandemic to meet the need for speed to develop therapeutics.

Key takeaways:

Note: These are summaries of what speakers discussed and not verbatim comments.

Defining RWE and How it Can Be Used

Ülo: RWE is derived from RWD – data collected in a natural setting. This data is becoming more important to payers. They want this data because they are not always sure that the results of randomized clinical trials are always applicable to the broader population.
Lixia: Using RWD is not a new concept and has been used to evaluate the patient journey and to formulate health economic models and social impact. With digital health becoming a reality, RWD has exploded and has now become big data. This is now the priority for everyone as evidenced by the framework established by the FDA that asks pharma companies to use RWE as a supplement, or in a situation where a RCT is not feasible, RWE should be used to support the drug development decision-making process.
Thomas: COVID-19 is putting RWD front and center because of the amount of data out there and the need for speed to develop therapeutics. If RWE is thoughtfully collected and analyzed, it can be extremely powerful.

Companies Mentioned

BlueDot’s outbreak risk software safeguards lives by mitigating exposure to infectious diseases that threaten human health, security, and prosperity.

Use case example: predictive AI paired with RWE in a Virtual Clinical Trial. Biovista conducted in effect a Virtual Clinical Trial with the FDA and a large EHR system in 2015 validating its hypothesis with RWE potentially saving $8B/year in avoiding hypothyroid patients getting diabetes from taking statins, published in ADA’s Diabetes Care.

Epistemic AI
With Explainable AI, Epistemic AI is working on making it effortless to connect knowledge and formulate hypotheses.

Used during COVID-19 when patients in their 30s and 40s, who did not have pre-existing conditions, were arriving at Mt. Sinai Hospital with strokes and heart attacks. The critical and immediate need was to determine if this was an observation at just one hospital or if it was more widespread. TriNetX has access to the RWD (through EHRs) of over 100m patients and thousands of COVID-19 patients. They were able to review this data and could see that these symptoms were happening all over the place, not just at Mt. Sinai, and this was a widespread risk even for healthy people in their 30s and 40s.  

The Ability of RWE to Predict Future Pandemics

Ülo: RWE can help to predict future pandemics because it can be used to better identify patterns by using RWE and AI in real time. This can trigger alerts that something unusual has happened and potentially help to guide public health professionals to identify areas to focus on for testing.  

How is AI Being Used With RWE?

Thomas: You need an integrated technology platform to make RWE effective and then you need to apply AI. RWE can provide information across a variety of factors and the ideal would be using real time RWD. This would enable decisions to be made based on a larger group of patient data and at a faster speed than has been possible before. Ultimately, the goal is to allow for more robust conclusions in real time.  

Challenges to Adoption

Lixia: RWD is unstructured data and therefore it is more difficult to manage. Also, bias can be unintentionally introduced into the data. Using AI with RWE can reduce “noise” in the data to get to conclusions faster and can be used in conjunction with randomized clinical trials.  

Over Reliance on P-Values

Lixia: P-values have been used too much and abused. It is not effective to just use p-values to make decisions. P-values should be used as a reference in conjunction with a random clinical trial hypothesis and is useful for this purpose. RWE will give you a clearer picture because it uses the totality of the evidence.

Data Privacy

Ülo: Technology today allows for an adequate amount of privacy. However, there is no barrier that prevents every possible criminal activity so that needs to be acknowledged. Thomas: You need to put in place what you can to provide privacy and security but you also need to reduce the impact of people having less privacy related to their shared data for medical purposes to impact the greater good of improving patient treatments and outcomes.  

Working with Regulators

Ülo: Regulators have been very open to focusing on RWE and providing guidance based on the need to learn from a larger number of patients and to learn quickly. However, the regulators are under pressure from some old guard thought leaders who are too wedded to RCTs. General thinking needs to be updated with knowledge in science and technology and disease states that are current. RWE is an opportunity to pay for something that works vs. something that doesn’t and industry needs to work together using the FDA Framework to move the use of RWE to be more mainstream.

Watch the full webinar:

We encourage you to listen to the recording to hear directly from each of the speakers and we’d love to hear your thoughts and comments regarding this content.

Thank you to our community

The successful and productive gathering of the AI in Biopharma community would not have been possible without the passionate group of attendees who actively participated in the webinar and we appreciate their contributions and questions. The chat room was a lively forum for interaction between attendees and several participants posted relevant articles or other information to share with the group. We would also like to thank the webinar moderator, Harry Glorikian, Author, MoneyBall Medicine and General Partner, New Ventures Funds for his thoughtful questions and skill at keeping the conversation moving.

Register your interest to be notified of upcoming webinars

/* */