Indication-specific pricing of pharmaceuticals in the US healthcare system

Some manufacturers have addressed the challenge of marketing a single drug with widely differing clinical uses by gaining regulatory approval for different brands for each indication. For example, sildenafil was initially approved for male erectile dysfunction under the brand name Viagra^® (Pfizer, NY, USA) in 1998, but in 2005, it gained separate approval for the treatment of pulmonary arterial hypertension, and marketed for this use under a new brand – Revatio^® (Pfizer, NY, USA) [11,12]. Viagra has an average wholesale price of $57.93 per 25 mg tablet and Revatio has an average wholesale price of $43.66 per 20 mg tablet [13]. This multiple-brand approach facilitates separate value assessments and provides a mechanism for different pricing for each indication of the same compound. It has been used selectively in the US either when the indications are distinctly different and separate brands are viewed as commercially attractive to help define the market for each or when two manufacturers license the same compound for different uses. For similar indications, however, such as treatment of different forms of cancer, multiple brands may be too burdensome or contribute unnecessary confusion.

This option establishes different prices for each indication and administers these through a direct link to drug usage. We are unaware of any examples of this approach in the US, however, work by the Office for Health Economics in England suggests that Italy has had the most experience with this approach [14]. For example, in Italy, separate risk-sharing agreements apply on an indication-by-indication basis for the cancer drug bevacizumab, with a specific additional 7% discount applied when used in advanced colorectal cancer. This approach, though conceptually simple, can be difficult to implement, as it requires robust data systems for successful execution.

Instead of trying to capture indications to assign differential pricing, an alternative is to use ex ante population estimates to establish a single weighted-average price, and then through retrospective review of actual utilization across all indications adjust the rebate accordingly. This approach may address payer concerns about ‘overuse’ for lower value indications while still providing access for higher value indications.

The weighted-average method appears to be administratively simpler than other approaches and may be easier to communicate to other stakeholders such as clinicians, patients and the public. However, since this approach requires retrospective review of claims to determine reconciliation based upon actual usage by indication, it may be best suited for organizations with robust data capabilities.

Capturing indication information in a precise and reliable manner can be problematic. Clinicians are not always required to provide the indication while prescribing a drug and pharmacies may not know or record the indication for which the drug is prescribed, even if the prior authorization process required by the insurer needs the information. Even when the indication is known at the point of dispensing, it may need to be verified for accuracy with the physician if it is the subject of a manufacturer rebate agreement. Medical claims may have this information but lag time and dissociation from the pharmacy benefit may limit usability. Thus, few payers currently have the data capabilities to implement ISP models that require patient-level indication information.

Drug formularies are typically organized by drug category and may not have the capability to place the same drug in different tiers per indication. Administrative challenges of implementing more flexible formulary designs may make it difficult to align patient cost-sharing with value-based pricing in a transparent fashion. In principle, patients prescribed a drug for a ‘higher’ value indication should not face the same cost-sharing as patients prescribed the same drug for a ‘lower’ value indication.

Prescribing physicians and hospitals purchase certain drugs from the manufacturer (or an intermediary) at a set price per milligram in a process uncoupled from indication. Reimbursement is based upon a CMS-issued volume-weighted average sales price (ASP) derived from data submitted by manufacturers 6 months earlier [16]. Because of this delay and dissociation from actual use, circumstances could arise where the medicare reimbursement rate could be lower than the acquisition cost for some indications.

In exchange for coverage in state Medicaid programs, manufacturers enter an agreement with the Department of Health and Human Services to provide a quarterly rebate to state Medicaid programs based upon a statutory formula: for ‘innovator’ products, the base rebate amount per each unit is 23.1% of the average manufacturer’s price, the price that manufacturers charge retail pharmacies before any negotiated discounts. However, if the manufacturer offers a rebate to any qualified purchaser in excess of 23.1%, Medicaid must also receive that ‘best price’ rebate [17,18]. ISP agreements with commercial payers could interfere with the Medicaid Drug Rebate Program if one of the indications is linked to a discount that exceeds the basic Medicaid rebate; this would trigger a new ‘best price’ for all uses of the drug in all state Medicaid programs and would impact the mandated price to 340B eligible entities [17,18]. Absence of an explicit exemption to the best provision from CMS, some manufacturers may be unwilling to enter into a commercial ISP contract.

Manufacturers can only negotiate reimbursement contracts for FDA-approved indications. Drugs that have significant off-label uses, including ones that may be supported by research, guidelines and compendia, are unlikely to be suitable candidates for ISP since a decision must be made regarding which price will be used for off-label uses, and manufacturers cannot enter contract negotiations that give the perception of promoting off-label use.

A federal Anti-kickback Statute prohibits offering or receiving remuneration to induce or reward referrals for items or services paid for by federal healthcare programs [19]. It is possible that an ISP arrangement under which the manufacturer accepted risk for ‘overuse’ of a drug for a lower value indication could be viewed as remuneration offered to encourage the health plan to favorably cover the manufacturer’s product. Statutory and regulatory safe harbors protect certain arrangements from Anti-kickback Statute liability, but it is unclear how enforcement agencies would apply these safe harbors to ISP contracts.

Whether using a single weighted-average price or separate, differential discounts by indication, ISP requires robust data capabilities that are not available system-wide today. Nonetheless, some payers have the data capture capabilities and cross-platform integration to support innovative reimbursement models that depend upon claims information. Other payers can seek to improve these capabilities.

Drug markers or ICD-10 codes from claims data could be used as surrogates to help identify the indication, but this would be limited to indications that can be clearly identified and are easy to distinguish from each other, such as rheumatoid arthritis and cancer. Prescribing physicians, either through prior authorization, as part of a new prescription format or both, could be required to report the indication for each patient receiving a drug. Specialty pharmacies that serve as intermediaries between manufacturers and patients have extensive data tracking capabilities that could be adapted to capture indication.

For infused products, the issuance of different billing codes for physician-administered drugs (usually referred to as J codes and issued by CMS) for different indications could be a simple way to differentiate uses. For oral products, National Drug Codes are the universal product identifier assigned post-FDA approval and used to bill payers [20]. Creation of additional National Drug Codes for billing purposes would be difficult and is less realistic than the assignment of different J codes.

The inflexibility of drug formulary tiers can be addressed in the short run by selecting drugs for which pricing can vary by indication but formulary tier remains consistent.

Although oral drugs are not entirely unaffected by ISP, oral drugs are more likely to avoid entanglement with infusion-based ASP-plus pricing policies that might create financial losses for practitioners using the drug for its lower value indication. To address concerns about ASP, infused drugs could be distributed and coded based on their separate indications to accommodate the process of physicians and hospitals buying and then billing after the drug is administered. But this approach would require extensive administrative work.

It is unclear whether CMS will proceed with its proposed National Part B demonstration program in which regions of the country would be randomized to either retain the current ASP + 4.3% payment structure or shift to an ASP + 2.5% + a fixed-fee model. With either model clinicians risk losing money if they predominantly prescribe a drug for its ‘lower value’ indication in an ISP approach.

As the largest payer of provider-administered drugs, any changes to Medicare reimbursement in the buy and bill system is likely to have a big impact on the healthcare system overall. Thus, commercial payers may be reluctant to make significant changes until more is known about the Medicare Part B Demo. If CMS opts to use ISP itself for some Part B drugs, private insurers and manufacturers to follow suit for those drugs. If CMS does not move forward with ISP, payers and manufacturers interested in trying to implement ISP contracts for provider-administered drugs should work together to present to CMS formal requests to create exemptions to the ASP-based reimbursement mechanisms that currently obstruct ISP pilots.

Similarly, the Medicaid Best Price policy could be modified to accommodate ISP models by specifying an allowance for different best prices for different indications, and that any new, lower best price created by an ISP agreement will affect only the best price for that indication. The consequence of this modification would allow ISP contracts without affecting Medicaid best price for drug prices that are undifferentiated by indication.

Communication between payers and manufacturers regarding off-label use of pharmaceuticals is an ongoing area of debate: policies are unclear and challenging to interpret [21]. However, off-label use should not be an insurmountable barrier to ISP. Recognizing that limitations exist on promotion of off-label indications, and that some utilization will fall outside the FDA-approved drug label, payers and manufacturers can address these concerns by selecting drugs for ISP that have minimal off-label use, by applying indication-specific price adjustments only to labeled indications, and by using a weighted-average approach to ISP.

Table 2 presents a cumulative list of the statutes and regulations that may impinge on ISP agreements.