Big Data An Opportunity or a Distraction? Signal or Noise? Maya R. Said, Sc.D. SVP & Global Head, Oncology Policy & Market Access, Novartis 3rd International Systems Biomedicine Symposium Luxembourg, 28 October 2015 The opinions expressed in this presentation and on the following slides are solely those of the presenter and not necessarily those of Novartis Pharmaceuticals Corporation ( NPC ). NPC does not guarantee the accuracy or reliability of the information provided
Big Pharma s interest in Big Data is on the rise... 2 3 rd International Systems Biomedicine Symposium M. Said 28 October 2015 Luxembourg
...and traditional Big Data players are entering non-traditional pharma/biotech space 3 3 rd International Systems Biomedicine Symposium M. Said 28 October 2015 Luxembourg
But is it really an Opportunity or a Distraction? 4 3 rd International Systems Biomedicine Symposium M. Said 28 October 2015 Luxembourg
In 60 years, R&D productivity was cut by a 100... Source: Scannell et al. Diagnosing the decline in pharmaceutical R&D efficiency. Nature Reviews Drug Discovery 11, 191-200 (March 2012). 5 3 rd International Systems Biomedicine Symposium M. Said 28 October 2015 Luxembourg
...despite a dramatic decrease in the cost of sequencing As the cost of sequencing has decreased, the volume of genomes sequenced has grown exponentially 6 3 rd International Systems Biomedicine Symposium M. Said 28 October 2015 Luxembourg
Reduced productivity has been driven partly by the increase of the informatics burden... Contribution Of Different factors To The Overall Cost Of A Sequencing Project Over Time Sanger Seq. NGS Future (2020) Source: The real cost of sequencing: higher than you think! Genome Biology, 2011 7 3 rd International Systems Biomedicine Symposium M. Said 28 October 2015 Luxembourg
...as well as the dramatic increase of the search space without adapting the search algorithm Drug discovery Preclinical Phase I Phase II Phase III & submission Launch O O O O Advances in basic science leading to more unvalidated targets More focus on molecular biology vs translational medicine Incentives to continue advancing projects leading to late and costly failure 8 3 rd International Systems Biomedicine Symposium M. Said 28 October 2015 Luxembourg
More data has disproportionately contributed to more noise Source: The Boston Consulting Group 9 3 rd International Systems Biomedicine Symposium M. Said 28 October 2015 Luxembourg
What will make it an opportunity: our ability to extract the signal from the noise 10 3 rd International Systems Biomedicine Symposium M. Said 28 October 2015 Luxembourg
Many past successes to learn from 11 3 rd International Systems Biomedicine Symposium M. Said 28 October 2015 Luxembourg
How much do we really need to know? Voltaire (1694-1778) "Doctors are men who prescribe medicines of which they know little, to cure diseases of which they know less, in human beings of which they know nothing" Archie Cochrane (1909-1988) Between measurement based on RCTs and benefit in the community there is a gulf which has been much under-estimated. 12 3 rd International Systems Biomedicine Symposium M. Said 28 October 2015 Luxembourg
What should we focus on? Molecular & Clinical Data Physiciological & Behavioral data Increased productivity, better outcomes for patients Literature, publications, Knowledge Real world evidence 13 3 rd International Systems Biomedicine Symposium M. Said 28 October 2015 Luxembourg
How do we integrate across various data types? Needs Care delivery Market Clinical Trials Real world data EMRs Disease management EHRs CROs Evidence Hypotheses OMICS data Knowledge Research 14 3 rd International Systems Biomedicine Symposium M. Said 28 October 2015 Luxembourg
How do we make sense of a complex and rapidly evolving environment of tools and analytics? SEL End-user interfaces/ workflow etc. Proprietary UI Source HIS Systems System benchmarking a. improvement Biomolecular data In-silico science, Data analytics and Conv. DX/ IM Collaboration N/A Infrastructure Data sources Provider of GUI Decision support in context of care Discovery of optimal treatment paths Omici a Omicia Archimedes unstructure d data History, symptom s, Swedish outcome, HC registry cost Data management ECT EXAM ED PLES Patient clients unstructure d data Watson Integrators of scientific knowledge base Archimedes Watson Patient data integrators Swedish HC registry Storage & comp. Server/DB softw. Server/Network tech. Data transfer Biomolecular data Conventional diagnostic data Hist., symptoms, outcome, cost Existing clinical knowledge 15 3rd International Systems Biomedicine Symposium M. Said 28 October 2015 Luxembourg Source: The Boston Consulting Group
A number of hurdles need to be addressed... 1 Data Acceptance 2 Alignment on Outcomes 3 4 5 Access to Data Analytical Methods Data standards 6 Privacy 7 Transparency 16 3 rd International Systems Biomedicine Symposium M. Said 28 October 2015 Luxembourg
...necessitating collaborations among all stakeholders across the ecosystem Academia Hospitals Patients Pharma Biotech Providers Tech Payers... Regulators Governments 17 3 rd International Systems Biomedicine Symposium M. Said 28 October 2015 Luxembourg
Progress is being made......starting to align on what data is needed for what purpose Evidence required Now Past Budget impact Post-marketing commitments (safety etc.) Unmet need/ disease burden Adherence Head to head safety, comparative effectiveness Effects of switching on outcomes Patient recruitment Utilization/ prescribing patterns Long-term safety, clinical outcomes Differentiation in sub-populations Target populations Usage Difference Differentiate with or vs. protected formulation Understand standard of care Trial design Development Submission Launch Pricing Review New Competition New Indication / Formulation Competitor goes generic Source: Adapted from McKinsey PMP practice perspectives on RWD 18 3 rd International Systems Biomedicine Symposium M. Said 28 October 2015 Luxembourg
Progress is being made......major recent advancements in data collection and processing tools EXAMPLES Single (Simple) Biomarker Detection Only Does John have a mutation in Gene X? EMR/ Clinical Data Only Has John smoked for >30 years? Multiple Biomarker & Pathway Detection Does John have a mutation in Gene X and does he have high expression levels of proteins Y? Integrated Treatment Decision Making Has John smoked for >30 years, and does he have a mutation in Gene X, as well as high expression levels of proteins Y? Source: EY Analysis, Company Websites, Analyst Reports 19 3 rd International Systems Biomedicine Symposium M. Said 28 October 2015 Luxembourg = biological data = phenotypic data = both types of data
Progress is being made......engaging all stakeholders to solve a common problem Example: The IMI2 Big Data for Better Outcomes programme Goal Support the evolution towards outcomes-focused and sustainable healthcare systems, exploiting the opportunities offered by big and deep data sources Themes/Enablers 1 2 3 4 Design sets of standard outcomes and demonstrate value Increase access to high quality outcomes data Use data to improve value of HC delivery Increase patient engagement through digital solutions Sets of target outcomes Clinical endpoints Alignment of HC stakeholders on the value of those outcomes.. 20 Mapping of sources, methods and tolls for collection and harmonization Governance and technical standards... Drivers of outcomes variation Best clinical practices Methodologies to predict outcomes... Patient Reported Outcomes opportunities Profiling patients behaviors Tools to increase patient engagement...
Big data can have a major impact on pharma across the value chain... Key opportunity areas across the value chain Discovery Research Clinical Development Regulatory & Access Manufacturing Sales & Marketing Real-world evidence Identifying new drug targets and validating MoAs Predicting properties of drug molecules Using innovation networks to enable science Developing optimal clinical protocols Optimizing clinical trial enrolment Profiling patients and monitoring trials via analytics Analyzing clinical data for regulatory submission Integrating realworld evidence with RCT data Mining clinical data to demonstrate value products to payers Forecasting demand Identifying bottlenecks and optimizing equipment efficiency Predicting and preventing equipment failure Predicting prescribing behavior Developing prioritized prescriber detailing Optimizing sales force structure and organization Understanding medication effects Segmenting patients to better target medicines Assessing the impact of postlaunch activities (e.g. marketing, phase IV trials) Source: The Boston Consulting Group 21 3 rd International Systems Biomedicine Symposium M. Said 28 October 2015 Luxembourg
...as well as on patients and healthcare costs... Examples with outcomes focus and increased HC value Organization Patients Impact Cataract PE-rates 1-80% Costs 2-1% Outcome Coronary artery bypass grafting Complic. -21% Costs -28-36% HC Value = Cost Chronic diseases Mortality -53% Costs - 151/pat. All patients Patient. sat. +20% Costs -25% Lung transplant Survival +3% Costs -6% 1. Postoperative endophthalmitis 2. Swedish saving estimated based on US cost data Source: NCR annual reports 2000 2009: Schmier, J. K. et al (2007); Friling et al (2013); ProvenCare SM, Ann Surg (2007); Geisinger Health System, ACO (Oct 2010); Integrated care experiences and outcomes in Germany, The Netherlands and England, Health Affairs, 2014: Best Care at Lower Cost: The Path to Continuously Learning Health Care in America, IOM, 2012; Cleveland Clinic transplantation 2011 report; McClellan, and Kent et al, WISH accountable care report(2013); NHS Confederation, The Search for Low-cost Integrated Healthcare, The Alzira model from the region of Valencia, 2011
...offering a potential solution to a rising and unsustainable problem 23 3 rd International Systems Biomedicine Symposium M. Said 28 October 2015 Luxembourg
Thank you for your attention