Genomic medicine has increased heterogeneity in patients’ responses, which may soon be reflected into pricing schemes. Indication-based prices may increase drug expenditure; marginal value-based prices may reduce the incentive for the industry to assess the differential in effectiveness. This calls for more research into the effects of drug price regulation on innovation.
Personalised medicine is going to change the shape of the pharmaceutical market in the years to come. The advances in genomic medicine make new active principles be very effective only for an increasingly smaller number of patients, with effectiveness declining rapidly if the patient does not match the characteristics of the targeted one.
This revolution is going to reshape drug pricing. The market for drugs is highly regulated, and prices are not defined according to the usual rules of perfect competition. Most economists would agree that value-based schemes are probably the best alternative to set the price of new drugs since costs are far more difficult to be verified.
The idea behind value-based prices is to reimburse the drug according to its societal value, which in turn depends on its effectiveness (usually measured through the expected life years that the drug allows to gain) and society’s willingness to pay for such an increase. However, which value should be considered (the marginal value, the average value, or some other definition) is still debated, and the choice is going to be quite important in the light of the changes produced by personalised medicine.
Marginal value-based schemes have received quite a lot of attention in recent literature and work in this way. Let us assume that N patients, normalised to one for simplicity, could be treated with a drug that is about to be listed. For a group q of patients the drug allows to gain yH of years of good health while for the rest of the group (1-q), the number of years gained is lower (yL). Let us also assume that society is willing to pay ?$ for each unit of effectiveness. Under marginal value-based prices, the price of the drug will be ?yL. If, for simplicity, we assume that the marginal cost to produce the drug is zero, the profit of the industry (revenue minus cost) is equal to the cost (the price) borne by the provider. The interesting feature of this scheme is that, by paying for a price that is equal to the marginal effectiveness, society is able to get “good value for money”. For the first q patients the price paid is lower than the value society attaches to the years gained, while the industry still has a profit. However, this model may have important consequences as concerns the incentives the industry has in investing in research & development (R&D) in order to determine the effectiveness differential. To show this, let us first assume that the industry knows the effectiveness differential. In this case, the industry has three different options for listing:
Ask for listing for both indication. In this case the price will be equal to ?yL
Ask for listing only for the indication for which the drug is more effective. In this case the price will be equal to q?yH
Ask for listing only for the indication for which the drug is less effective. In this case the price will be equal to (1-q)?yL
The third option will never be taken into account: it entails a profit lower than alternative 1. The choice between 1 and 2 depends on the effectiveness differential between the two groups and its relative size. The profit for the first alternative is ?yL while for the second is q?yH, which is higher if the effectiveness differential is sufficiently high .
From the point-of-view of the healthcare provider, alternative 1 is preferable to alternative 2 since the price is lower and the number of people treated is higher. However, the choice does not depend on the regulator, but with the industry, and in this case, the price for listing on both markets does not represent the best alternative for the industry.
An interesting example of this conflict is the case of bevacizumab, an active principle that has proved to be effective both for colon cancer and age-related macular degeneration (AMD). The drug has, however, been approved only for colon cancer and some physicians have started using it for AMD off label(not without quite a lot of discussions and High Court rulings). For AMD, given the quantity of active principle needed to treat the patients, the drug becomes rather cheaper compared to other alternatives, and its price no longer reflects its value.
It is also interesting to note that if the industry decides to list only for the more effective group of patients, the full value of the drug is not exploited. There are (1-q) patients that could benefit from the drug, but they will never be granted access because the drug may not be listed for their indication.
Another interesting perspective on which to evaluate this payment scheme is related to the incentive that the industry may have in investing to assess the effectiveness differential; several additional costs should in fact be incurred: randomised clinical trials must be larger, and more investment is needed in research. Marginal value-based schemes may not offer enough incentive since they tend to reduce the payment of the drug (the price will be set according to the effectiveness of the marginal patient).
If the industry does not invest in finding the effectiveness differential, the price of the drug will be set according to the average expected effectiveness across patients. From the regulator point-of-view, this means an increase in cost, but also patients may be worse off. With respect to the previous system, it may in fact seem that patients are better off if the industry does not invest in finding the differential in effectiveness. Both groups are treated and some will receive a benefit higher than expected ex post (the q group), while other will be disappointed.
However, let us broaden the range of drugs and also consider that a second drug may be available, but the latter is more effective for the second group. Knowing the effectiveness differential, the first group should be the first active principle while the second group would benefit from using the other active principle. However, if the return from the extra investment is not sufficiently remunerated, it may be possible that both drugs are marketed for both groups of patients with a loss of health for both groups.
In the recent past, some authors have proposed new schemes that are commonly known as indication-based prices. These value-based schemes foresee selling the drug at a price that depends on the actual effectiveness of the drug. In the example before, this would mean that the drug is sold for the price ?yH to the first q patients and to ?yL to the other (1-q).
This scheme is known in industrial economics as price discrimination; in competitive markets it is expressly forbidden by antitrust agencies (it will be seen as abuse of dominant position).
The most important drawback of this scheme is that it allows the industry to reap most of the societal value of the drug. It may also seem unfair that if the industry could have a profit selling at a price equal to, it should choose to market the drug also to a higher price simply because the willingness to pay is higher for a specific group of patients.
This principle is somehow accepted across countries. Drugs in low-income countries are sold at prices lower than in high-income countries, but less across patients within the same country. On the other hand, it should be considered that discovering the effectiveness differential has a cost, which may be reimbursed through this price difference. This shows that a trade-off exists between value for money and information: if the goal is to reduce expenditure, it may well be that less information on differences in effectiveness across patients will be known. In this environment, regulators will play an even more strategic role than in the past.