Pharmaceutical Management Science Association

PMSA Board Members Jason Carlin and Karthik Chidambaram 
present the Lifetime Achievement Award to JP Tsang, Bayser

General Session

PMSA Monday night dinner cruise

Incentive Compensation today continues to be a critical function in sales operations, and is one of the key levers for driving field force effectiveness. Expectations from incentive compensation team leaders have now become more demanding. Priorities are changing at a rapid pace, with new customer types, new selling models, advent of new technologies, new data sources they need to contemplate, data restrictions, cost pressures, and emerging global captive CoEs (Centers of Excellence). IC team leaders today need to be aware of best practices that drive high impact.

In this tutorial, we will focus on three main themes:

• IC Plan Design Analytics in Specialty Brands: We will highlight how IC plan designs and best practices differ in specialty markets, small patient population situations, and how data challenges are more acute, and are there new data sources being used. We will also discuss common plan design templates that can be effectively rolled-out, new metrics that are being contemplated, and will discuss pro-con analysis including situations where specific plan design templates score better. There will be a focused workshop on quota/goal setting including how seasonality and data patterns need to be addressed, how and should you have floors/caps etc. We will end this segment with a discussion on payout curve designs and best practice on financial risk analytics as you look to deploy IC plans for future POAs.
• Impact on IC from Reduced Access and Multi-Channel Marketing: Digital strategy is making advances from a more holistic MCM approach to reach HCPs. The industry is also encountering drastically reduced access, something that is more pronounced in specialty markets. We will discuss how IC plan design needs to react to these challenges.
• Perspectives from Other Industries: We will discuss how IC plan design and operations differ in other industries (financial services and banking, retail etc.), and some projections that may come true in our industry

There is a convergence of new analytic methods and data availability that can advance the understanding of how drug choice and treatment decisions are made and influenced. This workshop shows how newly available data sources in combination with similarity clustering techniques can enable a researcher to determine which institutions have the highest influence over prescribing within a community of physicians with similar behaviors.

The objective of this tutorial session will be:

• To demonstrate Excel - implementable methods for measuring prescribing similarity
• To demonstrate Excel – implementable methods for clustering physicians into communities with similar behaviors
• Demonstrate Physician “roll up” and aggregation methodologies to assign proportionate influence metrics to institutions and payers.
• Determine Impact of ACO risk sharing agreements on quantifiable models of physician activity.
• Discuss strategic implications of Health Care consolidation and strategies for maximizing MSL and sales force effectiveness.

John J. Castellani is President and Chief Executive Officer of the Pharmaceutical Research and Manufacturers of America (PhRMA), which represents America’s leading biopharmaceutical research companies. The biopharmaceutical sector directly employs over 810,000 Americans, and invested more than $50 billion in 2013 alone to develop new medicines that help patients fight disease and live longer, healthier lives.

At the nexus of public policy, health and business, Mr. Castellani leads PhRMA’s work to preserve and strengthen a healthcare and economic environment that fosters medical innovation, new drug discovery and access to life-saving medicines. He is a passionate advocate for a strong American biopharmaceutical research industry that plays a critical role in helping improve patients’ health in the U.S. and around the world.

Recognized each year by The Hill newspaper as one of America’s top health care policy experts, Mr. Castellani is a member of the Board of Trustees at Sibley Memorial Hospital and a member of the Board of Directors of the National Health Council. He was also honored with the Bryce Harlow Foundation’s prestigious Business-Government Relations Award, recognizing his leadership and exemplary, life-long contribution to the public affairs and advocacy profession.

Mr. Castellani is a former President and CEO of Business Roundtable, an association of chief executive officers of leading U.S. corporations with a combined workforce of nearly 12 million employees and $6 trillion in annual revenues.

Prior to Business Roundtable, Mr. Castellani was Executive Vice President of Tenneco, Inc.

Mr. Castellani, who began his career as an environmental scientist at General Electric, earned his bachelor’s degree at Union College in Schenectady, New York.

This session will present the second part of the outputs from the joint PMSA-ZS research initiative on the emerging EMR data vendor landscape. Since the presentation in 2013, the EMR data market has evolved rapidly with several recent M&A activities and the announcement of joint partnerships. In addition, data reliability and maturity are improving. The first part of the session will cover key trends on the EMR landscape, including a framework for evaluating EMR data suppliers. The second part will present several cases to provide a current overview of the robustness of the data to support key applications, including recommendations to manufacturers and data suppliers. We will also provide access to deliverables from this initiative such as the joint ZS PMSA white paper, list of EMR data vendors and their specialties.

The pharmaceutical marketing landscape has been changed significantly by the emergence of various types of “health systems”, variously known as ACOs, IDNs, IHNs, etc. These organizations create varying degrees of centralized control and decision-making for pharmaceutical or device and diagnostics products. Pharmaceutical companies have generally responded to this emergence by creating (or increasing) the role of Account Managers calling on key decision makers at these health systems.

The activity of the Account Manager will vary widely based on the products being sold and the type of health system they are calling on, but in almost all cases, part of their goal is to influence policies and formulary decisions at the target account in a way that improves the selling environment for the “regular” sales reps calling on the various target accounts or physicians that are influenced by that health system.

Until recently, the ability of the Account Manager (AM) to produce that favorable impact on the selling environment was pretty much taken on faith . . . it certainly seemed like it ought to make a difference, and there was occasional evidence that it mattered . . . perhaps a big contract closed after a prolonged sales effort. But there was no direct evidence of a generalized impact, and no evidence of what kinds of health systems were most likely to be influenced by the activity. Nor was there any robust targeting strategy (which health systems should the AMs call on, and how hard?).

Recently, we have completed work that demonstrates a robust way to quantify these impacts, and demonstrate the actual value created by AM activity. The model offers all of the following components:

• Quantification of both indirect and direct impacts from AM activity
• Relative prioritization of target health systems based on their size, degree of centralization, and the mix of account types that they influence
• Identification of the role of “regular” sales rep pull-through, offering the opportunity for stronger coordination of activity across the sales organization.

This presentation will discuss the modeling approach and offer examples of impact effects.

This presentation is based upon a specific analysis of a recent launch in preparation for a next generation launch for the same market. The goal of the analysis was to improve messaging and targeting through a better understanding of new and existing prescribers. The products were high priced acute treatments for a chronic disease that has a high number of undiagnosed patients and a long and asymptomatic disease progression, treated by a variety of prescriber specialties. This presentation will provide an overview of the specific case, the insights gained from the analysis, and a summary of how those insights were made actionable.

We will share an approach that demonstrates how to use patient-level data in a robust way to estimate market opportunity and product penetration in the field of Oncology. We will be sharing examples of results from a specific application of this approach for an Oncology drug that is indicated for 3 different tumor types and specific treatment regimen within each tumor type. Our presentation will show how patient-level data from Rx and medical claims, can be used to develop an approach for estimating market opportunity and product penetration at an account level for Oncology products. The presentation will also highlight some key aspects of the approach/analytics including:

• Accurate identification of a defined market, by using a deep clinical knowledge of the disease area to determine indicators in available patient level data.
• Identify specific treatment regimens based on longitudinal history of patient treatment, and also understand product usage by lines of therapy.
• Align patient treatment activity to accounts, by linking patients to physicians and physicians to accounts
• Generate estimates of penetration by product at an account level, by triangulating across different data sources to avoid potential for bias due to limited or unrepresentative samples
• Size account opportunity, through a scaling process that considers coverage at an account level

No single source of data can answer all business questions, and our industry is constantly faced with trade-offs when evaluating data solutions: Closed-system or open system? Projected pharmacy data or adjudicated life-cycle pharmacy data? Electronic Health Record data with provider notes or chart pull? Each data source offers its own unique strengths and limitations.

The evolution of Management Science will require linking disparate data sources that traditionally sit in silos to answer more questions more efficiently. One such example – Integrated Claims and Electronic Health Record data – offers a more complete and useful understanding of the clinical narrative, addressing questions such as primary adherence. Opportunities for data integration extend to other data sources including health risk assessment data, socio-economic data, and self-monitoring health data like continuous glucose monitoring. Integrating such data assets will create opportunities for new analyses, and along with these opportunities, challenges in defining elegant methodologies.

We start from a simple and fundamental premise. There is no value added from what you do unless the great analytic work leads to action that delivers business impact. No value – period!

Today the job of a marketing scientist or marketing researcher is to drive business impact, and doing this requires more than technical skills. We call these skills “Power Skills” – being a consultant; having the ability to influence others; developing insights with business impact potential, increasingly integrating multiple information sources; leveraging graphical and visual excellence; and communicating for impact.

In the 2014 Cambiar Future of Research Study, both clients and providers rated Storytelling the number 1 skillset need in the next five years. Ian Lewis will focus on How to Communicate for Business Impact. He will state mandates, and introduce principles, practices and tools that you can start using right away, emphasizing a structured storytelling approach. His talk deals with the content of a presentation, NOT on delivery skills.

The Open Payments database, implemented under the 2010 Affordable Care Act, is the only national resource of its kind to report the financial interactions between physicians and health care manufacturing companies. The existence of this new data source raises questions as to how pharmaceutical, biotech, and medical device companies can leverage this new data and incorporate it into their selling strategies. One strategy in which the Open Payments database will have a significant impact is physician selection processes—the Open Payments database provides a view of physicians previously unavailable and therefore can result in a more multifaceted view of a prescriber in order to evaluate whether or not he/she would be a good candidate for selling activity.

Patient assistance programs (Co-Pay buy downs) have quickly become an industry standard as the cost of prescription drugs continue to rise and patient economic issues have applied pressure on patient adherence to chronic medications. It is hard to find a product and/or market that is not deploying or researching the potential of a buy down program. For those that have deployed, are their programs having the desired affect? Are they effective? As these programs can be a costly expense to the pharmaceutical companies marketing budgets, it is equally important to determine and measure program effectiveness? But how should one measure effectiveness? Is it merely by balancing the increase in patients versus the cost of the program?

This presentation will focus on a real-life case study that measured voucher program effectiveness beyond the traditional methods, and will provide the attendee with a new way of looking at program effectiveness that can be used with their own brands.

As quality-based reimbursement metrics roll out and new provisions of the Affordable Care Act take effect, new tools to increase Patient engagement is one of the byproducts. The tools are designed to help providers demonstrate better outcomes. Technology has created new options for healthcare providers that increase efficiency and promote engagement that may lead to better outcomes. One of those tools are patient portals. They are now available for many patients to review test results and other information from their physician online. As of December 2013, about 57% of providers offered a patient portal. They are also a new and potentially important part of your research plan.

For patients with sufficient motivation, technology can lower the barriers to taking action. Health professionals can provide important support to their patients by educating and encouraging them in the use of appropriate technology to improve their health. Patients, their caregivers, and the entire health team have an opportunity and a responsibility to share new technologies, gain experience, and learn from one another.

We have entered an era that allows patients easy access to their medical records. They can now simply view their x-rays and computed tomography scans, look at their electrocardiography results, and read consult reports from other specialists on a secure server. If laboratory findings are normal or abnormal, then the portal can trigger the communication. It could be a simple, timely note relaying that a patient’s laboratory results were normal or one indicating that the physician is requesting an appointment for further evaluation.

And it’s that consistent and timely dialogue that is going to help improve health outcomes, from managing chronic conditions to catching a problem during a routine physical examination.

In a recent study by Delbanco et al. published in the Annals of Internal Medicine, researchers invited patients to read their doctors’ notes. Of 5,391 patients at three sites who opened at least one note and completed a post-intervention survey, 77% to 87% reported that access to the notes helped them feel more in control of their healthcare. Sixty percent to 70% of those taking medications reported increased adherence.

This technology also allows the provision of the electronic patient record, their portal information and the ability to survey they immediately following a moment of truth – such as a diagnosis or a first script.

As pharmaceutical marketing continues to evolve with the more widespread adoption of digital, mobile, social and other emerging media, one phenomenon remains the same: TV is still the dominant media force in DTC.

And while TV continues to command the lion’s share of DTC media budgets, pharma marketers are encountering increased scrutiny and accountability in optimizing their TV buys and proving return on their investment. Fortunately, pharma marketers can now take advantage of data-driven, privacy-safe approaches that empower their TV campaigns to reach more relevant audiences, drive incremental sales impact, and maximize ROI.

In this session, Crossix will explore how the combination of predictive data, innovative analytics methodologies, and new audience targeting technologies are revolutionizing TV for DTC. The presentation will include a recent brand case study that demonstrates how predictive audience models are developed and applied to optimize TV campaigns. Join us to learn much, much more about this exciting, transformative opportunity for DTC marketing!

As the US healthcare landscape continues to evolve, Managed Care Organizations (MCOs) have been facing growing pressures to manage healthcare costs. Consequently, MCOs are more actively managing access to healthcare providers/services and prescription drugs. The utilization management of drugs has greatly affected the availability of products to patients and the ability of pharmaceutical companies to drive growth. Increasingly, pharmaceutical companies are actively employing analytics to measure the impact of managed care on their portfolios and incorporate this thinking into their managed care negotiations. However, as the frequency and magnitude of impact of managed care events increases, pharmaceutical companies have found it increasingly challenging to align commercial planning with ongoing changes in the managed care landscape. While incorporating managed care impact is commonly quoted as a major commercial planning challenge by pharmaceutical companies, only a few companies have been able to actively integrate this aspect in both upstream and downstream activities such as forecasting, customer segmentation, targeting, tactic planning, and incentive compensation.

In this presentation, we will share case studies to highlight insights and learning from adapting our managed care analytics methodologies to support various aspects of the commercial planning process including:

• Forecasting: Factoring for the impact of managed care events (access wins/losses)as well as the managed care ripple effects of other market events (competitive launches, LOEs, new indications, studies, etc.) into demand and gross-to-net projections
• Product Launches: Integrated approach to factoring managed care access and patient assistance programs into launch sales & marketing planning, to optimize launch trajectories via limiting perception of poor managed care access
• Event Based Pull-through Initiatives: Optimizing targeting, access messaging, and estimating pull-through potential to guide pull-through execution (call planning, etc.)
• Incentive Compensation: Prospective and retrospective approaches to account for payer formularies and access changes when evaluating sales force performance, taking into consideration various factors driving differences in formulary adoption across sales territories, ensuring a fair and robust approach to incentive administration

Through these real-world cases, we will outline a collaborative approach to translate insights from managed care analytics into actionable inputs for commercial planning processes. We will conclude with an overview of the potential benefits as well as reflections on additional opportunities to integrate managed care into the planning cycle.

Overview: More frequent updates of medical claims data facilitates decision making and promotes swift action for Pharmaceutical Sales & Marketing teams.

Why Important: Historically, the age of medical claims data has carried a stigma for being outdated, particularly when compared to prescription claims data. This has made it difficult to use medical claims to support timely business decision making. However, as healthcare technology evolves, medical claims data is becoming more readily available. This shift has opened a door to new and innovative ways of using daily, weekly, and monthly claims feeds to derive key insights and enable more agile sales and marketing teams.

Supporting Case Studies: This talk will walk through a series of case studies that highlight how pharma companies are gaining competitive advantage with “near real time” medical claims data:

• Measure promotion response and campaign effectiveness for infused therapies through weekly or monthly trending reports based on medical claims activity (diagnosis and procedure) at the physician/facility/payer level.
• Give sales reps the opportunity to influence a physician’s therapy choice before treatment by pinpointing newly diagnosed patients at the physician level.
• Identify patient candidates for clinical trials to supplement recruitment efforts and meet critical timelines.

Given many tumor sub-type segments and intense competition in some areas, effective and credible identification of “cause and effect” for an Oncology product’s performance can be especially challenging. With any product’s perceived or actual non-performance typically being multi-factorial, internal organizational dynamics also quickly become challenging to navigate. This interactive session will focus on a real-world case study from the presenters’ recent experience, where an effective integration of data mining, primary research, team sessions, and strategy building resulted in measurable turn-around of product performance and improved satisfaction of cross-functional team stakeholders. Session participants will see specific examples of hybrid methodology analyses and data integration, team workshops design and output, and resulting business strategy. In addition to the review of analytical methods, we will illustrate bridging from research findings to business implementation, including re-framing key metrics, developing a new strategy, and sales force training.

Electronic medical records (EMRs) contain a wealth of real-world clinical data and can be a source of deep, actionable insights for life sciences companies. Significant challenges currently exist, however, in accessing EMR data - including integrating across fragmented providers, combining data across different EMRs, and accessing valuable data in both structured fields and unstructured sources. This problem is particularly important to solve in oncology, where the market is segregated into cancer subpopulations only identifiable through the use of datapoints like biomarkers and histology, which are captured in hard-to-reach unstructured documents such as clinical notes and pathology reports.

This session will focus on a practical approach to extracting data efficiently and accurately from unstructured EMR sources in the oncology setting, creating a rapidly growing dataset that is representative of contemporary cancer populations and treatments received. Real-time analytics on top of these rich, complete data can transform commercial decision-making. The data can then be combined with other data sources such as claims, tumor response variables, and genomics to provide a 360-degree view of the cancer market.

This presentation will explore a new approach to identifying key influencers. First, it goes well beyond the workhorse referral network inferred from patient-level data, and encompasses several networks such as professional associations, local community gatherings, country clubs, colleagues at work, high school and college reunions, etc. Communication and, as a result, the ability to influence each other increases with the number of networks through which two physicians can interact. A physician starts imitating other physicians when the physician realizes that a good number of physicians of significant stature exhibit a certain behavior, namely, the adoption of a product. There is an urge to adopt which builds up as more and more connected physicians adopt the product. The physician jumps on the bandwagon when the urge exceeds a certain threshold (specific to that physician).

Unlike a neuron that fires an electric signal down the axon when the combined upstream potential exceeds a certain threshold, the physician stays on and keeps on prescribing the drug. Three things are very clear at this point. First, investing time and money to build physician networks is a worthwhile and rewarding enterprise. Indeed, the more networks we have at our disposal, the greater the accuracy with which we can gauge the information conductance that connects two physicians, and, as a result, our ability to establish if a physician is indeed influential and if so to what extent. Second, identifying key influencers in the context of a product launch is extremely relevant as it is primarily in this context that we will be deploying these insights (segmentation and targeting). Third, the algorithm currently assumes that a physician adopts a drug when the weight of the social proof (number of high caliber physicians adopting the drug) exceeds a personal threshold. Some physicians are mavericks and do not need any social proof to adopt a product. In other words, their threshold is very low. Our inability to recognize this is what's going on when the maverick adopts late in the game may lead us to overestimate the stature of connected physicians. A more sophisticated algorithm is in order.

Multi-source claim level data linked through diagnosis, treatment, patient, physician, or facility ID will provide much more powerful information than just using them independently. For example, medical claim data capture all activities within a doctor visit, but with limited coverage on Rx data; pharmacy claim data provides a close to complete picture of Rx drugs, but with no disease information. When combining both, we can trace back Rx with the related disease type and provide a more complete treatment cycle. Many drugs now are treating more than one indication. Targeting with only information from the Rx drug will not serve the exact need of the complicated therapeutic area. Disease population distribution requires more critical information to understand market penetration and market potential.

The VNS Therapy System is FDA-approved for the treatment of refractory epilepsy. In addition to Rx data, it is better to utilize procedures and/or diagnosis to further define potential patients and size the market. The ultimate goal is to optimize the VNS prescriber target list.

We will use VNS as one example through out to illustrate how to apply multi-level regression with multi-source claim data. Traditionally, only prescriber level Rx data is used to provide the doctor targeting list. Because Rx data has no visibility of procedures in the doctor's office, the recommended approach is to leverage Rx, patient RX claim and medical claim data to build a predictive modeling. The modeling will provide the probability for each patient under every doctor if they need the procedure or not. By linking patients from both data sources, a sample of patients with both Rx drug treatment history and medical office visit history were created. The dependent variable is the event that the patient had the procedure of interest for, in the past or not under the doctor. The fixed effects are information related with the patient, a co-morbidity condition, drug treatment, treating physician, and physician office type.

Because in many cases, doctor and patient correlations vary by nature, office IDN influence on doctor could also be significant; a multi-level regression was highly recommended. The doctor and/or affiliated facility type could be treated as random effects. This provides more advantage than a simple patient level modeling. Doctors with affiliation to key locations of Epilepsy services would have higher probability treating patients with both drugs and procedures. Epilepsy is a chronic disease, therefore doctors with more patients with procedures in the past would have more chance treating patients with procedures in the future. Other information, like various patient co-morbidity conditions were also significant predictors for the need of procedure. Model results were validated through hold off sample, which was randomly selected and repeated many times. Selected significant predictors proved to make business sense. An optimized targeting list was tested to be effective by the client sales force.

In the current pharma commercial paradigm, marketers must do more with less and there is greater focus on accountability and the return on marketing investments. The commercial organization has responded by attempting to further understand the impact of promotional activities but most companies have not been able to “crack the code” and fully understand how effective these tactics are, especially with many of the non-personal promotion / digital tactics. Our solution design includes a novel approach, methodology, tools and insight “inventory” to identify and assess marketing ROI performance gaps and solutions and can address a significant un-met market need related to Non Personal Promotion Effectiveness.

With the increase in the number of orphan and specialty drugs being launched, biopharmaceutical companies are struggling to apply traditional analytical methods to identify treating physicians for personal promotion. As therapies become more targeted towards patient sub populations (specific lines of therapies, narrower indications, specific biomarker stratification, patients with incomplete response to standard of care etc…) where there are no existing product markers, it becomes more difficult to identify physicians who treat targeted patients. Furthermore, patients in these specialized markets are often managed by multiple physicians. How can management science help identify the controlling physician?

Current techniques of identifying target physicians typically reply on valuing physicians based on competitor drug usage, using weights to value scripts and estimating the ‘potential’ for each physician. In orphan and specialty disease areas, secondary data is sparse the treatments are often non-specific (i.e. single drug used for multiple indications) and this creates a lot of noise when estimating potential. We have seen several companies assess their targeting post launch, and seeing uptake across all ‘deciles’ being very similar.

In this session we will describe (via cases) a patient focused micro targeting approach which yields a more reliable targeting strategy in these disease areas.