Partnering to Share the Risk

Statements of fact and opinions expressed herein are those of the individual author(s) and are not necessarily those of the Society of Actuaries or the respective authors’ employers.

Health care actuaries pay attention to large costs and areas of rapidly changing costs. In the past several years, growing drug budgets have been associated with pharmaceutical innovations that have transformed the treatment of many challenging diseases. While these advances have greatly benefited patients, health plans often struggle to predict and manage the impacts of these therapies on their budgets. As a result, actuaries are paying much closer attention to the financial risk associated with pharmaceuticals.

With new pharmaceuticals coming onto the market and the overall health system’s move away from a fee-for-service environment, there is an increasing focus on value-based payments in which reimbursement for pharmaceuticals is linked to treatment outcomes via benefit design.¹ These so-called risk-sharing agreements (RSAs) may also be referred to as performance-based risk-sharing arrangements (PBRSAs), managed entry agreements, patient access schemes, coverage with evidence development (CED) and outcomes-based risk-sharing agreements (OBRSAs).

The Emergence of OBRSAs

OBRSAs are one approach that health care payers and pharmaceutical manufacturers can use to manage risks. This article explores the emerging use of OBRSAs for pharmaceutical contracting; the role of actuaries; and a multiyear learning initiative designed to develop, test and refine solutions and new types of models for RSA development. Through our collaboration, we discovered similarities and differences related to culture, language, method and value drivers between health actuaries and health economic outcomes researchers, who are often the ones responsible for quantifying the economic value of pharmaceuticals.

Traditional financial-based schemes focus on the financial arrangements between the pharmaceutical manufacturer and the purchaser/payer (e.g., health plan). These agreements are not typically associated with specific performance metrics, but rather rely on traditional terms such as rebates, discounts, price-volume agreements and/or quantity limits. In contrast, outcomes-based schemes are tied to specific performance metrics such as biomarkers, clinical outcomes and/or medical resource utilization (e.g., hospitalizations), and often include coverage with evidence development and guarantees.

Currently, treatment costs and clinical outcomes are not necessarily aligned, leaving uncertainty in the correlation between the two. Through OBRSA agreements, patients, payers and pharmaceutical manufacturers have an opportunity to clarify the relationships among treatment costs, clinical outcomes and possibly related financial outcomes. As such, OBRSAs for pharmaceuticals are becoming an increasingly popular topic of discussion in the United States, especially as the broader health system moves from a fee-for-service model toward a pay-for-performance model. Despite the growing interest, a recent study found only 12 percent of global risk-sharing agreements have been executed in the United States.²

Historically, significant barriers have hindered the development of OBRSAs, such as defining relevant performance metrics, adequate data infrastructure, lack of operational and adjudication capacity, policy and regulatory barriers and their implications, an often fragmented multi-payer health care system, uncertainty about the real-world performance of the pharmaceuticals, and ambiguous contractual terms and conditions. These barriers require creative and novel approaches to OBRSA development and execution.³ Despite these challenges, OBRSAs present opportunities for both payers and pharmaceutical manufacturers. See the sidebar for details.

A Unique Learning Opportunity

Over the past two years, two leading organizations in the U.S. health system—one a pharmaceutical manufacturer and the other an information and technology-enabled health services business—collaboratively conducted an iterative, exploratory and data-driven initiative meant to inform the methodology behind, and design of, OBRSAs. The overarching purpose was to inform the development and execution of OBRSAs over the next three to five years. Structured as a “learning laboratory,” the initiative focused on immediate learning rather than immediate success or failure in designing and testing innovative OBRSA models. The learning laboratory is the first phase of a potential multiyear plan to develop and execute OBRSAs. The next phases could focus on prospective OBRSA piloting and then broader pilot implementation based on key learnings.

In the learning laboratory, real-world data⁴ was used to develop and test predictive outcomes-based and financial models in order to better understand the types of variables, populations and clinical characteristics that are most predictive of clinical and financial outcomes. The data was also used to explore new models and modeling methods, define stakeholders for whom such agreements may be most promising and develop OBRSA contracting archetypes. We also identified prevailing and evolving regulatory challenges that should be considered in the development and prospective testing of OBRSAs. Contributors to the learning laboratory included not only actuaries, but also experts in health economics, health policy and health care data, along with the health plan, pharmacy benefits management (PBM), and medical and pharmacy leaders.

There is a lot to be gained from OBRSAs for payers, pharmaceutical companies, patients and ultimately the health system at large. Exploring these innovative models requires a commitment of time and resources, but we are all encouraged by the progress we’ve made so far.

Areas of Scope

The learning laboratory focused on two therapeutic areas, each of which included several work streams (e.g., actuarial, health economics and outcomes, contracting strategy and policy). The therapeutic areas included a chronic condition predominately managed in an outpatient setting and a more acute condition that can be treated in either inpatient or outpatient settings. A governance committee included senior leaders from both organizations to align on key strategic and technical topics and oversee the project teams. A series of modeling and retrospective analyses were then conducted to inform an OBRSA simulation model, which was developed to estimate potential payer and pharmaceutical manufacturer financial value under specific guarantee, formulary and contracting conditions.

Concurrently, an extensive forward-facing analysis (consisting of literature review, conference presentations⁵ and interviews) was conducted to ascertain how health care culture, policies and regulations shape the OBRSA environment. Example topics included federal and state pricing reform initiatives, government best price regulations, potential changes in the federal 340B drug pricing program, the impacts of alternative payment models (APMs) and value-based insurance design (V-BID), HIPAA and related privacy and security rules pertaining to personal health information, applicable statutory authority for risk-sharing, and biosimilar and orphan drug regulation.

Return-on-investment modeling is one way that actuaries, and the organizations they represent, prioritize risks—to determine which risks might be worth engaging with nontraditional approaches.

The View of “the Payer”

While a variety of OBRSA opportunities and challenges have been documented in the public domain, the learning lab distinctively focused on the perspective of “the payer,” including risk-bearing entities and other influencing stakeholders. Health care actuaries describe and measure financial risk. The aggregate size and probability of the risks are important in determining if and how specific risks are managed. The actuarial analytics quantifying the size and probability of the risk are critical in establishing the level of uncertainty, and thus whether and how risks should be mitigated or leveraged. Actuarial analytics were specifically included in the learning laboratory to complement the health economics approaches that are more familiar to pharmaceutical manufacturers. Ultimately, the role of the actuarial analytics was to estimate payer financial risks and opportunities with methods and language that are familiar to payer-community stakeholders.

Return-on-investment (ROI) modeling is one way that actuaries, and the organizations they represent, prioritize risks—to determine which risks might be worth engaging with nontraditional approaches. ROI models developed for health care payers estimate the potential incremental financial “investment” (costs) relative to the “return” (cost-offsets/savings) for
specific cohorts of health plan members. ROI models incorporate net-cost, utilization management and medical-cost offset scenarios, leveraging “proxy client data” from claims databases, with sensitivity testing of various population and assumption scenarios. This initiative used two types of analyses to inform the ROI estimation: payer addressable burden analyses (PAB) for the “return,” and formulary design modeling (FDM) for the “investment.” The data source for these analyses was a large proprietary claims database, which included access to commercial and Medicare Advantage Part D claims for approximately 27 million members.

The FDMs provide insights into key levers that payers may use to leverage or mitigate financial risks. These actuarial models were designed to replicate the analytic modeling payers use when determining pharmaceutical formulary placement, medical drug benefits, and clinical program and utilization management policies. The models estimate health plan per-member-per-month (PMPM) costs potentially associated with the OBRSA, incorporating key utilization and net-cost levers that health plans may use to exploit or mitigate financial risks. Specific payer perspectives include:

• Understanding the current pharmacy PMPM cost for the therapeutic class
• Estimating the net financial impact of changing coverage and clinical program decisions for the therapeutic class
• Estimating the risk that could be attracted based on the OBRSA design, terms and conditions

We’ve made major headway in establishing mutual understanding between the actuarial and health economics disciplines. This opened up the necessary dialogue to help us figure out a productive approach to these very complex financial models.

PAB analyses describe the total cost of care curve, its trends over time and the primary drivers of those trends. PAB analyses are ultimately used to identify opportunities to address those costs and trends, should the expected clinical benefit of a pharmaceutical and the hypothesized benefits of an OBRSA materialize. For this initiative, disease episode grouper technology [Symmetry Episode Treatment Groups (ETGs)] was used to bundle claims into episodes of care for the diseases being evaluated plus comorbidities, for three sequential yearly cuts of the medical and pharmacy claims data. ETGs provide a condition classification methodology that combines related services into medically relevant and distinct units describing complete episodes of care and costs. Each episode can be composed of multiple claims clusters. Each cluster has only one anchor record and may have multiple claims. Each claim line can be assigned to one, and only one, episode of care. The ETG software aggregates costs into five cost categories for care management (services, surgery, facility, ancillary and pharmacy), which are then added together to produce a total cost for a specific episode of care.

Key Learnings

Throughout the course of the OBRSA learning laboratory, team members paid close attention to similarities and differences in culture, language, methods and value drivers traditionally used within the health care payer and pharmaceutical industries. The goal was to build a mutual understanding beyond the terms and conditions of OBRSAs. Through our work together, we identified several common foundations shared by the actuarial and health economic outcomes research disciplines, including calculus-based statistical theory, skills in measuring results, focus on health economic impact and ability to deal with uncertainty. However, there were also divergent approaches and applications. Whereas actuaries leverage the law of large numbers to minimize statistical variation, the pharmaceutical industry approach is to use characteristic-matched or controlled samples. Additionally, actuaries tend to focus on lines of business (e.g., Medicare, Medicaid, commercial large group, small group and individual), whereas the pharmaceutical industry has more of a disease-based focus. We also found that actuaries tend to use models with shorter-term time horizons, as they are more relevant to the business decisions they are addressing, whereas the pharmaceutical industry typically uses models with longer-term time horizons such as spending for a specific disease area or capturing the full value of a medicine over a lifetime.

We’ve made major headway in establishing mutual understanding between the actuarial and health economics disciplines. This opened up the necessary dialogue to help us figure out a productive approach to these very complex financial models.

A Novel Opportunity

Health insurance is a promise to pay for certain health care expenses incurred over a specified period of time in the future. Both the number and frequency of claims are moving targets, with regular changes including but not limited to charges for services and products, the frequency with which products and services are utilized, and the emergence of new products and services. Accurately predicting the uptake and cost of new products is among the most important issues facing health care payers, or “risk-bearing” entities, which work to meet the needs of their members in a constantly evolving environment.

The pharmaceutical industry continues to develop new medicines, often delivering on the great clinical promise associated with substantial pharmaceutical development efforts. The associated costs and frequency with which these innovations are occurring has brought increased scrutiny and financial risk to the provision of pharmacy benefits.

OBRSAs offer a novel opportunity to manage and mitigate risks, or leverage them as appropriate. In their role measuring clinical outcomes and managing financial risks, health care actuaries may wish to further explore this emerging area of risk modeling as the U.S. health care system increasingly moves toward value-based payment models and new therapies continue to come on the market.

Copyright © 2018 by the Society of Actuaries, Chicago, Illinois.