Adapting Decision Analytic Models to Provide New Perspective

San Diego—Existing healthcare decision models may not always address real-world needs of decision makers, according to information presented during a Contemporary Issues session at the AMCP meeting titled Adapting Decision Analytic Models to Meet the Needs of the Health System. The presenters were Steve Duff, MS, president, Veritas Health Economics Consulting; Joe Gricar, MS, president, Apo-MED; and Prakash Navaratnam, RPh, MPH, PhD, senior director, DataMed Solutions LLC. The use of adaptive models, they told audience members, can provide new perspective and may help to inform health plan decisions.

Common decision analysis models (cost minimization, cost effectiveness) are used to help estimate clinical and economic outcomes of interest and act as a conceptual framework to aggregate different data elements. There are several key issues that affect decision models, which can indicate the need for a new modeling paradigm: (1) complexity (healthcare problems are multifactorial); (2) transparency (model credibility depends on it); (3) bias (sponsorship by industry); (4) uncertainty; (5) perspective (numerous stakeholders); (6) interpretability (deciphering results of economic analyses can be difficult); and (7) relevance (do the models/economic analyses enable a more informed decision and address the key issues of concern for the decision maker?).

Tier Placement Models

Tier placement models focus on finding the optimal product placement within a clinical pathway, providing an alternative to restricting access to new, high-cost products to avoid high upfront acquisition costs. They also seek to evaluate the impact of product placement on the overall pharmacy and medical costs as well as patient outcomes. Appropriate tier placement can maximize the cost effectiveness of a product’s use and may result in patients being treated more aggressively initially to avoid creating medical issues later on, thereby improving the patient’s experience.

Tier placement models can allow decision makers to make choices about expanding or contracting access using information that aligns with clinical pathways, which is their main advantage. However, they require more flexibility, and tend to be more complicated. Assumptions may need to be made regarding the impact of treating naïve patients versus failed patients.

Facility Models

Facility models can consist of simple accounting of facility revenues and expenses before and after introduction of a new technology, or become more complex by integrating other perspectives and interactions. As organizations evolve to take on greater financial risk, the presenters said, a clearer understanding of the impact of facility economics and provider behavior on health plans will be crucial for decision-making.

Key questions that can be addressed through the use of the facility model are (1) what should the health plan pay for the new device/procedure? and (2) are there reimbursement levels that can balance the needs of all stakeholders (plan, facility, clinician, and patient)?

Advantages of facility models include a more granular understanding of facility resource use and economics, evaluation of how financial drivers of clinical decision-making may impact health plans, and exploration of how new technologies can impact facility efficiency. Conversely, facility models largely exclude nonfinancial domains (quality of life, patient satisfaction), and may require extensive data collection/analysis.

Portfolio Models

Portfolio models seek to optimize the mix of products and services offered to meet desired endpoints over a distinct time window, and are conceptually similar to portfolio management models derived from finance. The return on investment for a health plan may be increased efficiency by minimizing costs or improving outcomes, or both.

There is increasing pressure on health plans to ensure products and services offered to their patients yield optimal returns for the investments made by the health plan and/or realized value savings for health plan clients (government).

Portfolio models consist of distinct products and services that can be priced in a discrete manner and can be tracked over time, and require a detailed understanding of the patterns of care and the relative impact of competing interventions on each other. These models also can become very complex and carry a perception of being overtly bottom-line driven. However, the presenters concluded, they do allow decision makers to weed out potentially unnecessary procedures or medications and provide the ability to simulate a new technology or service to determine its impact on the overall portfolio.