Following up on my spoof post about the complexity of making bets when it comes to new pharma products, worries about Tamiflu continue. The original issue was one of unexplained deaths in Japan, which were, more or less, subsequently explained. Now there are questions about its efficacy in the case of the H5N1 flu strain, and this is important since Tamiflu is known to have side effects. All in all, something to treat with caution.
The study raises new questions about the drug, which more than 50 governments have ordered in significant quantities in recent months to stockpile as a potential prophylactic and treatment in the case of a flu pandemic.
An accompanying article in the Journal reinforced calls for alternative approaches to treatment for a pandemic, including the stockpiling of the rival drug zanamivir, or Relenza.
Dr Anne Moscona wrote that individualsâ€™ stockpiling of Tamiflu was â€œpotentially dangerousâ€ because it could lead to insufficient doses and inadequate courses of therapy, in turn accelerating the development of resistance.
So maybe Brad Delong’s hypothetical government would have been better off nationalising Zanamivir (or maybe not). The thing is it is really hard to know in advance, and that is an in-principle problem. At the end of the day this is why the private sector solution may be better, because at least you have different horses in the race.
The title is really an ironic (if somewhat affectionate) reference to this post from Brad DeLong. Reading the news this morning, it seems that Tamiflu may not be such an unambiguously good thing as it was being made out to be:
Roche, the Swiss pharmaceutical group, on Thursday moved to reassure investors after US regulators said they would on Friday examine reports of up to 12 deaths and 75 cases of children who suffered health problems after using Tamiflu, the company’s anti-flu drug.
The US Food and Drug Administration said it was in â€œactive communicationâ€ with regulators in Japan, the country with the widest use of Tamiflu for regular seasonal flu treatment, and where all the deaths and most of the other incidents of side-effects occurred.
I think a number of points could be made here. Firstly in this game there will be no free lunches. There are risks one way and there will be risks the other. Individuals may have to take decisions based on the best available information. Secondly, at the end of the day Tamiflu is not going to be virus-specific for any possible variant of avian flu simply because we don’t yet know the variant, so forward planning and risk assessment is inherently a complicated business here.
Lastly, when Brad said this: ” Low-probability but high-payoff projects are likely to be underfunded by the government–but properly funded by private companies willing to roll the dice. However, these ex ante considerations vanish ex post when an epidemic threatens…”, ( maybe I would say better rather than properly, but that’s a detail) – he was both right and wrong, IMHO, since the real issue which lies behind the argument is the moral hazard one. If you let the market regulate drug development, but then when you have a drug which is a big winner you immediately take it over, it isn’t clear that the market will work as well as you want it to next time round. On the other hand, governments can’t just stand back in the face of a real and present danger to their citizens. So I guess the only answer is negotiation and consensus, and maybe this consensus would include compensating those companies who are given the green light to go ‘full speed ahead’ if it turns out that – post ante – that decision was a bad one.
All in all a complex situation where prudence is indicated.