Hey Thomas, thanks for sharing this very interesting article. It seems like a well thought out proposal, and I really appreciated some of the finer details, like the endogenous nature of the reward. Overall I liked it a lot; the nature of comments of course is they tend towards the negative, so please don’t take the below harshly!
My main concern with this proposal, which seems overcomeable, is the potential for fraudulent inflation of impact numbers by local officials in third world countries.
In an ideal system where patients pay for their own treatment, this does not occur, because people only buy the drugs if they intend to use them. In a third party decider/payer system the issue is at least reduced, because even though you could bribe an official to order too many, at least his employer, who would be on the hook to pay, is incentivised to prevent fraud.
The situation here seems worse. As far as I can see, a drug developer could bribe some officials in an extremely poor and remote region (who need to be included in order to achieve global eradication) to say that a certain number of locals have been cured (or a certain amount of emissions avoided). This lie will benefit the drug developer, benefit the local officials, and not harm anyone they know or any organisation they are accountable to. The only loser is a faceless international pot of money that they are collectively conspiring to loot; hardly a novel situation in international aid.
Separately, and I don’t think it affects your conclusion, I am not sure this analysis of the current system is correct:
Under the current monopoly regime, new pharmaceuticals that are just slightly better than existing alternatives can earn as much as first-in-class innovations; and duplicative products that do not improve the state-of-the-art at all can still capture a large market share, thereby garnering huge profits and reducing the rewards of a preceding break-through innovation.
It is true that, once they have launched, new drugs that offer only a slight improvement can earn as much as the origional drug (in cases where switching costs are low). However, once there are two drugs on the market, competition will generally significantly reduce the revenue either can earn. The innovator drug will have earned its higher profits during the period prior to the launch of the second drug. This time period might be short, but if so the innovator drug didn’t have that large a counterfactual impact anyway.
Finally, and least importantly, I think your description of the breakthrough HCV cure sofosbuvir (Sovalid/Harvoni) is incorrect.
An example is sofosbuvir, an effective hepatitis C drug introduced in the U.S. in 2013 at $84,000 per course of treatment, 3000 times the estimated cost-based generic price of $28.
Something of a nit-pick, but while it is true that $84k was the list price, because of the convoluted nature of US drug regulations this was not the actual price Gilead sold it at. The average US price even for the first year was $60k, but it was always lower internationally and fell from there as competition entered the market and they quickly burned through the highest willingness-to-pay subpopulations.
More importantly, Sovaldi/Harvoni is not just an example, it is a uniquely extreme example; no other drug has had quite the impact it had, in terms of representing a massive improvement over the standard of care (a cure! rather than mere treatment) for a serious condition that affected a lot of people, where the prior standard of care treatment was extremely expensive, allowing for large total revenues. I haven’t been a pharma analyst for several years know but to my knowledge essentially every other drug is significantly smaller.
In lower-income countries, patented medicines are often much cheaper but still widely unaffordable on the also much lower incomes there. In its first five years, sofosbuvir had reached only 5 million patients worldwide; the other 66 million remain infected and continue to spread the disease.[13] Such disease proliferation benefits the patentee who, were its new drug deployed in a global population-level strategy of disease eradication, would find its future sales decimated. Each year, hundreds of millions suffer, and millions die, from lack of access to pharmaceuticals that competing generic firms could and would supply quite cheaply if patent enforcement did not prevent them from doing so.
I think this is misleading. If we quote the full sentence from the report you cited, it is clear that the majority of those 66 million people did not even know they had the disease!
Despite its high prevalence, morbidity and mortality, only 19 percent (~13.1 million) of people living with HCV knew their status in 2017 4 , and only 7 percent (~5 million) received treatment worldwide as of 2017. (link)
Nor do I think this failure-to-erradicate helped Gilead. In 2014 they made over $12bn in revenue from Harvoni/Sovaldi; by 2019 their total HCV revenue, which also includes some other things, had fallen to under $3bn. A rapid global eradication strategy might have been much better for Gilead, because it would have lead to high demand quickly, instead of a slower, lower level number of patients while competitor products entered the market and competed the price down.
In this case, the patent system did provide a product which could cure and, if used widely enough, erradicate the disease, but (largely government) payers were (irrationally) willing to pay more for treatments than they were for a strictly better outcome of a cure. The problem was myopia on behalf of payers, who cared about their one-year costs, neglecting the longer-term benefits of curing people, and egalitarian prejudices against paying high prices for high value.
Hi Larks, a very big thank-you for your close reading and excellent comments on the Health Impact Fund (HIF) proposal. Let me go through them in order.
1. You write that the HIF is vulnerable to “fraudulent inflation of impact numbers by local officials in third world countries.” Note that the HIF does create parties incentivized to prevent fraud, namely all the other innovators with HIF-registered products. If one registrant claims greater health gains than it actually achieved, then the payouts to other registrants are thereby diminished (the HIF’s disbursements are unaffected). Some of these competing registrants operate in the same countries and may well be able to spot fishy claims. The HIF’s audit staff could then check up on reported suspicious claims. Should fraud be detected, the HIF would trigger statutory penalties – and there would also be the sanction of public opinion: it would look truly awful if a pharmaceutical company’s claims to have achieved health gains among the world’s most impoverished populations turned out to be based on bogus evidence obtained through bribery. So I am confident that this sort of misconduct could be effectively deterred down to a trickle.
2. You write that, under the current system, “once there are two [patented] drugs on the market, competition will generally significantly reduce the revenue either can earn.” I agree. My points were (i) that, with appearance of a patented competitor, the break-through drug will, under the current system, lose its advantage, and (ii) that this can greatly weaken incentives to swing for the fences i.e. to do R&D on possible break-through treatments. By contrast, the HIF preserves the advantage of the first-in-class product by treating it more generously than later copy cats. This is how: the health gains achieved with the break-through drug are always assessed against the standard of care prevailing when it entered the market – whereas the health gains achieved with later (also HIF-registered) copy cats are always assessed against a (higher) standard of care including the earlier breakthrough product.
3a. I accept on your word that the actual price charged for sofosbuvir in the US was often well below list price ($60k per course of treatment rather than $84k). That’s “only” 2143 times the estimated cost-based generic price of $28, still a markup of 214,186%.
3b. I also accept that the majority of the 66 million who did not get sofosbuvir did not even know they had hepatitis C. But this is not a context-independent fact! Had there been an affordable cure, many more potentially hepatitis-C-infected people would have sought diagnosis; and had it been paid for health gains, Gilead would have made, and pushed for, concerted efforts to identify patients with hepatitis C (in order to improve the health of these patients and of those they might infect).
3c. Holding fixed how it was getting paid, Gilead would not have fared better with a strategy aimed at suppressing and eradicating hepatitis C. Pursuing such a strategy, it would indeed have sold much more product – but at very much lower prices and profit margins.
3d. I completely agree that the root problem lies with myopic and irrational payers. This is the central insight behind the HIF: the pharma industry would achieve vastly more for human health if some of its monopoly rents were replaced by impact rewards as defined.
Hey Thomas, thanks for sharing this very interesting article. It seems like a well thought out proposal, and I really appreciated some of the finer details, like the endogenous nature of the reward. Overall I liked it a lot; the nature of comments of course is they tend towards the negative, so please don’t take the below harshly!
My main concern with this proposal, which seems overcomeable, is the potential for fraudulent inflation of impact numbers by local officials in third world countries.
In an ideal system where patients pay for their own treatment, this does not occur, because people only buy the drugs if they intend to use them. In a third party decider/payer system the issue is at least reduced, because even though you could bribe an official to order too many, at least his employer, who would be on the hook to pay, is incentivised to prevent fraud.
The situation here seems worse. As far as I can see, a drug developer could bribe some officials in an extremely poor and remote region (who need to be included in order to achieve global eradication) to say that a certain number of locals have been cured (or a certain amount of emissions avoided). This lie will benefit the drug developer, benefit the local officials, and not harm anyone they know or any organisation they are accountable to. The only loser is a faceless international pot of money that they are collectively conspiring to loot; hardly a novel situation in international aid.
Separately, and I don’t think it affects your conclusion, I am not sure this analysis of the current system is correct:
It is true that, once they have launched, new drugs that offer only a slight improvement can earn as much as the origional drug (in cases where switching costs are low). However, once there are two drugs on the market, competition will generally significantly reduce the revenue either can earn. The innovator drug will have earned its higher profits during the period prior to the launch of the second drug. This time period might be short, but if so the innovator drug didn’t have that large a counterfactual impact anyway.
Finally, and least importantly, I think your description of the breakthrough HCV cure sofosbuvir (Sovalid/Harvoni) is incorrect.
Something of a nit-pick, but while it is true that $84k was the list price, because of the convoluted nature of US drug regulations this was not the actual price Gilead sold it at. The average US price even for the first year was $60k, but it was always lower internationally and fell from there as competition entered the market and they quickly burned through the highest willingness-to-pay subpopulations.
More importantly, Sovaldi/Harvoni is not just an example, it is a uniquely extreme example; no other drug has had quite the impact it had, in terms of representing a massive improvement over the standard of care (a cure! rather than mere treatment) for a serious condition that affected a lot of people, where the prior standard of care treatment was extremely expensive, allowing for large total revenues. I haven’t been a pharma analyst for several years know but to my knowledge essentially every other drug is significantly smaller.
I think this is misleading. If we quote the full sentence from the report you cited, it is clear that the majority of those 66 million people did not even know they had the disease!
Nor do I think this failure-to-erradicate helped Gilead. In 2014 they made over $12bn in revenue from Harvoni/Sovaldi; by 2019 their total HCV revenue, which also includes some other things, had fallen to under $3bn. A rapid global eradication strategy might have been much better for Gilead, because it would have lead to high demand quickly, instead of a slower, lower level number of patients while competitor products entered the market and competed the price down.
In this case, the patent system did provide a product which could cure and, if used widely enough, erradicate the disease, but (largely government) payers were (irrationally) willing to pay more for treatments than they were for a strictly better outcome of a cure. The problem was myopia on behalf of payers, who cared about their one-year costs, neglecting the longer-term benefits of curing people, and egalitarian prejudices against paying high prices for high value.
Hi Larks, a very big thank-you for your close reading and excellent comments on the Health Impact Fund (HIF) proposal. Let me go through them in order.
1. You write that the HIF is vulnerable to “fraudulent inflation of impact numbers by local officials in third world countries.” Note that the HIF does create parties incentivized to prevent fraud, namely all the other innovators with HIF-registered products. If one registrant claims greater health gains than it actually achieved, then the payouts to other registrants are thereby diminished (the HIF’s disbursements are unaffected). Some of these competing registrants operate in the same countries and may well be able to spot fishy claims. The HIF’s audit staff could then check up on reported suspicious claims. Should fraud be detected, the HIF would trigger statutory penalties – and there would also be the sanction of public opinion: it would look truly awful if a pharmaceutical company’s claims to have achieved health gains among the world’s most impoverished populations turned out to be based on bogus evidence obtained through bribery. So I am confident that this sort of misconduct could be effectively deterred down to a trickle.
2. You write that, under the current system, “once there are two [patented] drugs on the market, competition will generally significantly reduce the revenue either can earn.” I agree. My points were (i) that, with appearance of a patented competitor, the break-through drug will, under the current system, lose its advantage, and (ii) that this can greatly weaken incentives to swing for the fences i.e. to do R&D on possible break-through treatments. By contrast, the HIF preserves the advantage of the first-in-class product by treating it more generously than later copy cats. This is how: the health gains achieved with the break-through drug are always assessed against the standard of care prevailing when it entered the market – whereas the health gains achieved with later (also HIF-registered) copy cats are always assessed against a (higher) standard of care including the earlier breakthrough product.
3a. I accept on your word that the actual price charged for sofosbuvir in the US was often well below list price ($60k per course of treatment rather than $84k). That’s “only” 2143 times the estimated cost-based generic price of $28, still a markup of 214,186%.
3b. I also accept that the majority of the 66 million who did not get sofosbuvir did not even know they had hepatitis C. But this is not a context-independent fact! Had there been an affordable cure, many more potentially hepatitis-C-infected people would have sought diagnosis; and had it been paid for health gains, Gilead would have made, and pushed for, concerted efforts to identify patients with hepatitis C (in order to improve the health of these patients and of those they might infect).
3c. Holding fixed how it was getting paid, Gilead would not have fared better with a strategy aimed at suppressing and eradicating hepatitis C. Pursuing such a strategy, it would indeed have sold much more product – but at very much lower prices and profit margins.
3d. I completely agree that the root problem lies with myopic and irrational payers. This is the central insight behind the HIF: the pharma industry would achieve vastly more for human health if some of its monopoly rents were replaced by impact rewards as defined.