I am also going to suggest that open science faces a pair of problems called Scylla and Charybdis. I am going to introduce the health impact fund proposal, and suggest how it could complement biomedical research and I am going to suggest possible directions for research.

First of all, what are the problems in a really long list of problems in pharmaceutical markets? The first one is that our paradigm for research is deeply inefficient. We're missing a lot of great opportunities because of the system that we have. So there's secrecy in research and this secrecy in research is created ass a sort of by product of the patent system. This kind of secrecy leads to a lot of duplication, excessive costs, and fewer promising new clinical products. At the same time, because information isn't shared, you end up with fewer collaborative, less collaborative work and fewer good outcomes for people. The research paradigm also directs research into the product areas that are going to be profitable, not those that will have the highest health impact. Companies will spend a lot of money working on drugs for rich people and ignore the needs of the poor.

It's also true that the paradigm for pricing is deeply inefficient. Since companies are awarded for developing therapies by the grant of the patent- the idea of the patent is that you get to exclude others from the use of innovation, then you keep prices high, and then this results in massive under-use of new drugs, to the detriment of poor people. Those are what I think are the key problems.

An important open science proposal is Aled Edwards and others in the Structural Genomics Consortium. So uh, the new model that Aled Edwards proposed, in combination with government- develop new drug targets in a consortium, um, with all research output to be open access. I think I don't need to say why that has benefits in terms of the science. It's pretty clear that the Structural Genomics Consortium using this model has had considerable success. Aled is proposing a push further forward than the SGC has gone so far. He's really thinking of pushing the Consortium out into first and second stage clinical trials so that you have products being developed to a proof of concept stage. He's proposing a pre-competitive scientific experiment whose results would be available to all, and enable development at the next stage into a commercialized product.

So now we return to Scylla and Charybdis. For those of you who have forgotten what was in the Odyssey; according to Homer, Odysseus had to steer his boat between two horrible monsters named Scylla and Charybdis, and Scylla was a horrible monster with six long necks with grisly heads. Charybdis had a face that was all mouth, she would suck in water and vomit it out every few hours. So, I wanted to get a picture, but I couldn't find one. I think open science has to deal with its own Scylla and Charybdis in the pharmaceutical markets.

Here's the Scylla. You go through with this consortium, you develop a product to the proof of concept phase. But the product becomes unpatentable. And so companies looking at this are going to say "how am I going to make money on this when I don't have patent protection?" And they decide to leave the product, not spend enormous money on clinical trials, required to bring it into the market with FDA approval, and this is not a good outcome in the sense that you've developed a product up to the proof of concept phase, but you have got nothing out of it. This is the worst possible outcome. This would not be an attractive result.

Suppose on the other hand that you are successful, in the sense that you get the proof of concept phase, and then some company is able to get exclusivity. This doesn't look good either (this is Charybdis). The final product is protected by patents, but then sold at high prices. This doesn't really go at the open science concept, which you started development with. You end up excluding many poor patients from the use of the product.

A brief word on clinical trails- clinical trials are really essential to product approval. Stage three trials are expensive. Some people claim that the typical cost for a phase three trial which brings a product to market approval, the average cost is $80M. That's the figure I hear. They are risky. There's a reasonable probability that even if they are successful, you might not get an approved product at the end. And it's not an organized open science framework. Governments could fund it, but historically that hasn't happened.

Here's where I think the health impact fund fits with open science. I am going to switch tracks for just a minute and introduce the idea. The idea of the Health Impact Fund is that each year there is a fixed sum for registered pharmaceutical innovations. Innovators could choose whether to register their product with the Health Impact Fund. If they decide not to, they have the usual options. If they do, they are obliged to sell the product at a low price, or else to license it openly, and then they would be compensated for this through rewards paid by the health impact fund. The key idea is these rewards in the fund.

How would the rewards be set? There would be a fixed reward pool each year. Each participating firm would obtain a share of the reward pool in each 10 years, following the introduction of the product, and each product's share would be equal of the reward funds, its share of the total qualities of all participating products. Companies generating a lot of health impact would have a big share of the reward pool. Measuring the health impact would be quality adjusted life years. You would be looking back each year and look at how the product has been used, to figure out its health impact, and therefore its share of the reward pool.

Innovators might like this idea because you choose products to opt into this thing that high expected health benefits, but low expected profitability under the usual approach. Especially products that are mainly used by relatively poor patients who can't normally afford to pay very much money. Another class of firms that might want to register their products are in cases where you had a product developed under a public-private partnership that would require a low pricing, or if the firm otherwise was reluctant to charge high price.

If the product lacks effective patent protection; if you don't have good patent protection, how are you going to make money? Well, Health Impact Fund has a way to make money. They have a way to reward the product moving through market approval. Let me tell you a little bit about health impact assessment. If you are going to assess health impact, you want to use all available information, you have to engage in lots of inference, as to the benefits created by a particular drug, the kind of information that would be available- like data from clinical trials, and then you have to add to this, like data from pragmatic trials, like how is it used in practice? Sampling of use in the product in different environments, survey data on demographics and clinical characteristics. You put all of this stuff together to assess the health impact of the product.

We think that the Health Impact Fund can be financed by governments, because they have the financial ability to make the commitments. If this is going to work, it has to be in the billions of dollars. Countries can contribute proportionality to their income. It's important to note that this isn't just a huge new resource cost, since the cost of rewards would bee at least partially compensated by savings in the low prices on the drugs that were registered. On the one hand you would be paying out for rewards, but they would be leading to low priced drugs that you wouldn't have to spend lots of money on.

An important feature of this is that the rate of rewards is self-adjusting. Firms know that the rewards that they are going to get are going to be reasonable, since the reward rate going to be too low means that other firms aren't going to enter. At the same time, tax payers know that the rewards aren't going to be too high, because then lots of other firms will enter.

Let me zip through to the open science bit. In a way then, the health impact fund offers firms a mechanism to be rewarded for developing products, through to approval, and then ensuring that the products are effectively developed and distributed. The products don't have to depend on patent status. The firms that finance this through the approval process, they get rewards; based on sales on all other firms that are selling that product, not just their own sales.

To suggest where we are going next- we are working on a pilot, where we have a particular drug sold under these terms in a particular country. Let me just mention a couple of other things. First, the product, this proposal doesn't just stem from me. Jamie especially was instrumental in thinking about this general idea. And there have been many other people who have worked on this idea over time. We've been working on developing this proposal, by the way I should say that Jamie Lovett is going to speak next, he doesn't agree with everything in this proposal, but a lot of the underlying ideas are his, so there. We've been in talks with governments and industry and many others.

Okay, I guess I'll close there, and hopefully we'll have some time for discussion. Thank you.