Scenario #3

Standard Discounts

Under this scenario, the federal government requires companies that want their branded products covered under Part D to provide a specified discount—in effect, creating single-payer price controls. This model has long-term risks to pharma because the only option that would be available to CMS to address cost increases is to demand greater discounts. Moreover, it would permanently place pharmacy into a separate budget because it isolates pharmacy as a single cost category, without any linkage to hospital or physician services. Pharmacy would compete with physicians and hospitals for funding.

To sell this benefit to Medicare enrollees, Congress could sweeten the deal by decreasing co-pays and filling the doughnut hole. It also could streamline the program by eliminating PDPs and choosing one to two administrators to manage claims and reimbursements (similar to how Part A and B are administered).

This scenario is more likely than the first two, only because it is an easy political solution to cost increases in Part D. To understand this, think about the year 2009: In this scenario, the political pressure to address the cost of Medicare and other social programs, combined with high federal deficits and dissatisfaction from consumers about the problems with the donut hole and confusing benefits, prompts Congress to act quickly. The political answer to all of these pressures is to attack private insurers and pharmaceutical companies, and propose a new benefit that is funded with deeper discounts from pharmaceutical companies.

Scenarios #4 and #5

Reference Pricing and/or Therapeutic Maximum Pricing

Because their overall threats to the pharmaceutical industry are the same, it's useful to think about Reference Pricing and Therapeutic Maximum Pricing together. Here, the primary risk to the pharmaceutical industry is that the government quickly reduces reimbursements to the generic level, and only the most cutting-edge, innovative treatments have pricing leverage.

Reference pricing, which is already used by several countries—Germany and New Zealand, for example—would operate largely like a Maximum Allowable Cost list. CMS would fix the maximum level of reimbursement for classes of drugs. Retail and mail-service pharmacies would receive only the specified per-unit reimbursement—no matter what they dispensed—and pharmacy margins would be driven by the ability to purchase drug products below the class price.

Therapeutic pricing is the equivalent of DRG reimbursement for hospitals. The goal under this model is to identify the optimal treatment regimen and price for a specific disease. For example, treatment for a "simple patient" with high cholesterol may be $1.00 per day, but more complex cases would have higher reimbursements based upon the drug and other therapies required.

These models attempt to address the issues of quality of care, value of outcomes, and variances in patient need. They are, in theory, better than straight price controls for the industry because they offer a place at the table for addressing issues of optimizing outcomes and total cost of care—price controls do not.

For these scenarios to occur, the following needs to happen:

• Cost increases for Medicare, and Part D in particular, must remain between five and eight percent per year for the next three to five years, taking some pressure off the need to act quickly
• The MA-PD model demonstrates success at measuring population-based health outcomes, but fails to attract more than 10 million lives
• Strong alliances of private and public groups must form to demonstrate how to implement and measure these designs before the cost debate begins.

The probability of these scenarios occurring is very low unless action is taken now to lay a foundation for these complex programs. The greatest barrier to these two scenarios is the complexity of developing and maintaining the evaluation and pricing mechanisms. Either of these models would require a significant amount of time to develop and test. What's more, reference-based pricing and therapeutic maximum models would be far more difficult to describe and explain to consumers than the current choice of MA-PDs and PDPs.