2016 Winter Symposium

The Pharmaceutical industry is moving fast towards more lean and efficient sales and marketing business model. Gaining further insight into ever changing market dynamics including disease specific utilization on treatment pathways in Oncology has become more of imperative in planning and implementing sales and marketing strategies.

This study will build upon access to broad and in depth data with IMS health, as well as communications with industry business leaders, to demonstrate examples of market dynamics and insights that provide performance based analytics and insights for nontraditional oncology markets.

This discussion will be applicable to many aspects of commercial decision areas such as Business / Product Development, Launch Readiness, Detailed physician Targeting, and Market Performance Assessment with a focus on:

• Commercial viability in terms of market sizing
• Launch forecast validation from treatment pathways in targeted tumor types
• Market performance metrics for targeted tumor pathways
• Patient Progression Prediction and Targeting

Anonymised patient-level data is rapidly growing in both volume and complexity. This presents exciting opportunities to help drive clinical insight and commercial value. At the same time, the very volume and complexity of the data poses analytical challenges. This session will explore how novel analytical methods from the field of predictive analytics / machine learning can enhance the value of real-world data. Examples will be used to motivate the session, including illustrations of how predictive analytics can be used to I) detect undiagnosed patients with rare diseases and II) improve salesforce physician targeting by identifying physicians with patients immediately prior to key treatment switches and / or disease progression.

Advances in biotechnology continue to enhance the power and speed of genomic capabilities through the rapid generation of large, high quality, genomic datasets. Genomic data, combined with our knowledge of biological pathways and our expanding Real World Data capabilities will play an increasingly critical role in all aspects of the drug development life cycle. This discussion will focus on current and future applications of this data in the generation of Real World Evidence.

Each cancer has unique properties which are considered when choosing a treatment for a patient. Just as each cancer is unique so are the methods and data sources required to answer the multitude of questions posed to the brand, sales and analytics teams when marketing and tracking an oncology product. Understanding these unique differences and how best to leverage the various sources and methods can strengthen a brands strategy.

The aim of this workshop is to illustrate how disruptive innovations in healthcare data, particularly lab data integration, can be used to inform commercialization strategies and improve brand success. This workshop will address how the healthcare space is evolving with the growing mountain of data that is available today. We’ll discuss what challenges will result when we are presented with many sources of data and what disruptive innovations in healthcare data will help your brand to tackle those challenges. When properly integrated, the various types of healthcare data can provide crucial insights all along the patient journey. In this session, Medivo will go over the basics of data integration, including benefits, challenges, and application.

The aim of this class is to bring the participant up to speed regarding one of the most fundamental techniques of our industry, Promotion Response modeling, and to explain how this technique has evolved over the years to accommodate the new reality we live in. Topics will include:

• The availability of new data assets such as patient claims data, EMR, and registries
• SP and hub data, in relation to specialty drugs
• The coming to the fore of managed care as reimbursement surges to prominence
• The consolidation of providers into IDN's and ACO's whose level of control ranges from absolute to laissez-faire

This class consists of 3 parts. The first part presents the nuts and bolts of promotion response, and ensures that the participant has the requisite vocabulary and concepts in place. The second part is a blinded case study where the participant is asked a series of thoughtful questions. By contemplating the problem, the participant will get a better appreciation of the finer points of Promotion Response modeling. The last part concludes with lessons learned in a playful Q&A format. This will reinforce key messages conveyed throughout the class.

This discussion will cover a variety of advanced topics in setting up a promotion response model, interpreting the coefficients, and deriving optimal strategies from the results.

Examples of ideas to be discussed:

• How results are driven by a mixture of cross-sectional (between physicians) variations and by individual physician-level variation over time (in response to variations in promotion). Which is “best” for robustness of results? What strategies are available for shifting this in one direction or another?
• Choice of segment membership
• “Standardization” or “normalization” of results by physician prior to modeling
• What are the implications in your choice of dependent variable to measure: Share, sales, or something else? NRx, TRx, or ...?
• Role of autoregressive terms (if you use them). How do they change the interpretation of other coefficients?
• Mixing of promotion variables on the “independent variables” side of the model
• How does your choice of functional form imply what you can, and cannot, see in the results?
• How does your model “see” possible interaction effects? Or does it implicitly deny their existence?
• How does your model account for possible substitution effects among different promotional tactics? Are there other ways to do this? Would they be better?
• “S-shaped” or “Threshold” models: They’re probably more realistic (maybe), but are they worth the analytical headaches to use them?
• Non-Parametric models: do we really have to believe the model follows a specific functional form at all? What if we “set it free”?

Active audience participation will be welcomed.

Integrating New Data Assets to Go Beyond Account Level, Cross-line/Patient type PRM
Few things are more exciting to an industry than the prospect that new data assets can be harnessed to provide insights that can drive competitive advantage. There are well-known gaps in data assets for injectable products for physician level valuation and response modeling for specific lines of therapy and subsets of relevant patients. The default in most cases is an approach that lacks precision - insights at higher levels of aggregation (i.e., account level) across all lines and patients that lack precision.

The presenter will cover a framework for addressing these gaps, and a brief case study on applying this integrated data ‘triangulation’ approach.

Integrating Marketing Insights for Enhanced PRM Precision for Personal Promotion
For most brand’s detailing costs are still the most important factor underlying its P&L. The most promising paths for enhancing the effectiveness of that investment is to systematically and continuously evaluate targeting for those encounters, messaging conveyed in those encounters, and in some markets the timing of those encounters, and to calibrate quickly.

The presenter will include an integrated framework for evaluating a brand’s targeting efficiency, taking into account structural factors, behavioral propensities, and willingness to engage in promotional efforts for the brand, and its communication effectiveness using IVA tablet data. A brief case study on the application of this framework will also be presented, along with ‘backtest’ personal promotional response modeling results substantiating the value of acting on those insights.

This session will focus on modeling local influence in local healthcare markets, promotional response modeling in networks, and empirical findings in the U.S. market.