Pricing Drugs Based on Their Outcomes

The American Consumer Institute Center for Citizen Research (ACI) has been highlighting the outrageous levels of drug pricing over the past few years.  Later, Congress voiced its alarm and recently presidential candidates have jumped on the criticism bandwagon.  Some are late converts to the movement, but unfortunate consumers have long been aware of the stratospheric pricing problem as they struggled to pay for prescribed medications.

Anticipating some scuffles with feisty elected representatives in the post-election period, some pharmaceuticals firms are proposing a new approach to pricing.  The new approach would be to price based on the treatment’s performance rather than the amount of the drug consumed.  At first blush the approach seems rational, but establishing the value of “performance” is more difficult than it may appear.  Furthermore, the eagerness to discuss performance pricing seems like an effort to change the topic away from today’s high prices and on to how best to arrive at those high prices in the future.

In some respects, the value of performance is already baked into the prices of newer medications, especially the biologics.  Biologics include vaccines, blood components, gene therapies and other agents that are not the straightforward “molecules” that comprise most prescriptions.

One example is Opdivo, an immunotherapy (biologic) that reduces the size of lung tumors and can elongate life for cancer patients.  Opdivo seems to offer an additional 8.6 months of life.   Opdivo was in a clinical trial to demonstrate its prowess as a first treatment for lung cancer.  Unfortunately, the trial failed, possibly due to technical issues that Bristol-Meyers Squibb could have avoided.  Opdivo was already approved for use on non-small cell lung cancers that have previously been treated by other drugs.  The trial failure cut the expected revenue for Opdivo by $4 to $5 billion, a heavy penalty for failing to prove its performance when it’s already approved to treat lung cancers as a second line drug.  Opdivo treatments were priced at $143,000 per year when it was first introduced.

Another example is Solvadi, an antiviral drug that can actually cure Hepatitis C.  Untreated for 20 years, Hepatitis C can lead to cirrhosis and eventually to liver cancer, then death.  Solvadi was originally introduced at more than $84,000 for a course of treatment, but the price is now a little lower because there is competition (e.g. Harvoni).

Some medications for rare diseases and conditions can be ultra-costly because there are so few potential buyers.  “ADA-SCID is a severe immune disorder that is usually fatal in the first few years of life… Strimvelis has cured this in each of the 18 children it has been tested on over 15 years.” Fortunately, there are just 15 children diagnosed with ADA-SCID each year, but the procedure for making Strimbelis (a gene therapy) is so complex and labor intensive to serve this small market that it is priced at $665,000 for each cure.  Other rare diseases and conditions can be priced even higher and at least one drug had price of $1 million but was purchased just once.

These three real world treatments are biologics, highly priced, and deliver both a reduction in suffering and/ or varying amounts of additional life.  Opdivo delivers an additional 8.6 months of life (over 11.5 months of which 2.9 months is attributable to the drug YERVOY) for $137,000 (the price for the average 11.5 months of treatment).  Strimvelis delivers almost a full lifetime for $665,000, and Solvadi delivers your life beyond the next 20 years for $84,000.

Could those treatments pass a future price for performance standard?  The federal government agencies identify a statistical value of life so they can perform a cost-benefit analysis for regulations that promote health and safety.  In 2010, the EPA placed a value of $9.1 million on each human life.  The EPA feels tempted to increase that value if the death were by cancer or from terrorism.  Clearly, there is more than a little sentimentality in the EPA’s thinking.  The Federal government would prefer to adjust its estimate of a life’s value before applying it to drug pricing – but today we have a firm number to work with and the $9.1 million figure has been used in regulating other industries.  We can use EPA’s life value to gauge whether these treatments are economical performers.

Strimvelis seems to meet our standard for value.  It delivers almost a full life (perhaps 95%) for just $665,000 for just 9.3% of an EPA life’s value.  Opdivo delivers about 1.2% of a lifetime for 1.5% of the 9.1 million.  In other words, Opdivo seems to fall short of the value standard.   Opdivo could be repriced at $109,000 per year or less to meet the standard.  Assuming the average Hepatitis C patient is aged 25 when treated, Solvadi probably delivers about 35 extra years of life (44%) for just $84,000 or 0.9% of a life’s value.  Solvadi meets our value standard with ease.

This simple illustration ignores the many other medical inputs that play a role in reducing suffering and saving lives.  A less capricious estimate of the value of a life would be a good start.  As well a monetary value of suffering reduction would be useful.  The Centers for Medicare and Medicaid Services are working on mechanisms that will promote higher quality care from physicians. In a more complete assessment, physician quality along with hospital care and the impact of other drugs used (e.g. YERVOY in the Opdivo example) would each be ascribed a portion of the value of the outcome. That is a more complex problem to untangle, but we believe it can and should be done.