Institutional corruption and off-label drug use

Today, I attended a presentation by an invited professor on institutional corruption and off label drug use. He introduced the problem of misalignment of interest. Public wants prescriptions that will solve their health issues. The pharmaceutical industry wants to maximize profits.

In particular he discussed the problem of off-label drug use. The idea is that the drug that was approved for indication A (indication is kind of a fancy word for illness) is used for indication B. This is obviously good for the pharmaceutical industry, because they can sell more. More importantly, they do not need to conduct very expensive clinical trials. However, this may not always be in the interest of the patient. It may be in the interest of the patient. Just because there is no direct scientific evidence for it, it does not mean that it is not effective.

Kind of interesting in this problem is the role and interactions of say four independent bodies. First, there is the patient of course that supposedly knows nothing. Second, there is the doctor prescribing the medicine. Third, there is the manufacturer/distributor of the medicine (pharmaceutical companies). Fourth, there is the medicine regulator – the government agency/law (say FDA).

Suppose the patient was being prescribed something off-label. This is the doctor’s judgement (and taking these meds are patient’s judgement, but I assume here that they do as doctors say). The doctors are the final decision makers in this model. Pharmaceutical companies can influence their decision by advertisement and I guess some sort of more direct financial way. By law, they are not able to directly promote their medication for off-label use. However, they can indirectly promote their medication by providing evidence from other bodies suggesting efficacy for off-label use. They may for example fund independent parties to conduct such research.

The FDA has practically no role other than regulating the initial approval and subsequent off-label advertisements. However, even after paying these fines, it seems off-label advertisements can pay off. Once off-label use becomes the norm, marketing is not needed.

Dr Rodwin proposed three things to address this problem:

First, that off-label use should be tracked. Currently, it is difficult to manage or even understand the true implications of off-label use (although he did cite some statistics). This can be done by restricting reimbursements unless the purpose of the drug is provided.

Second, using this data, that if off-label use becomes a problem, the pharmaceutical industry (by law) conducts clinical trials.

Third, to remove the incentive for promoting off-label use. Only the on-label use should be earning them that patent premium. Using the off-label use data, only the manufacturing cost is paid to the pharmaceutical company and the remaining is taken back.

Overall, the proposal is very interesting. Obviously the caveat is that it may find a lot of resistance from the pharmaceutical companies. However, he believes that it is the same as with the implementation of previous regulations such as full disclosure of ingredients. Apparently in the 1900s, some new laws that potentially increased cost for the pharmaceutical companies have been passed. He believes that this seem to be at first difficult, but eventually become the norm.

Even aside from this caveat, some other profs in the room seemed to disagree slightly. They argued that the phara industry faces larger problems. For example, on-label use drugs may not be as safe as expected and the fact that it is on-label provides false sense of security. Making small adjustments makes new loop holes and fixes nothing. This was a very interesting idea, but for now I will hold back on this and talk about my opinion of the original claim.

I find that the proposal focuses too much on what the pharma industry should do. He believes that they will do things as long as they can make money. I like his general approach. He objectifies every entity and simplifies the system into a game almost. Each has their own roles. In this game, his role is to propose rules and regulations that will force the other entities (pharmaceutical industries and others) to act in a way that would benefit the patients. Although he has not formalized the system into a game theory setup, I can easily see this happening.

Following this scheme, he seems to think that the pharma companies have an upper hand over the doctors. His resolution involves only the pharma companies and not much of the doctors (the first is about the doctors reporting but he aims to incentivize this behaviour by blocking reimbursements unless the information is present). I particularly disagree with the second proposal. I do not understand why the pharmaceutical companies should pay to conduct such research just because they are making money from it. What if they believe that the research would result in a negative outcome? The clinical trial would seem like a complete waste of money (and of course they would also be blocking subsequent sales from this off-label use but let’s suppose they would rather NOT do research but stop the doctors from prescribing –which they don’t even have the ability to do!)

Ethically speaking, I think the doctors should be taking more precaution. We just need better doctors… not ones that induce babies just because they want to go home at 5pm -_- (not sure if this is true since I just read it on reddit). But I think as a policy maker, it is more reasonable to track the larger small numbered entities (ie. The pharma companies).

Back to the proposal, I think the off-label use incentive idea is quite good. Companies should be paid premium (patent associated price) for their on-label drug use only. That will motivate companies to acquire on-label drug use status for more stuff. It’s not the drug that needs approval. It’s the drug use. It’s like getting a patent for IDEA. It’s not the substance but the substance that materializes the idea that matters. The patent should not be for the drug but for the IDEA that the drug solves some problem.

There are just so many things I would like to change and I find this interesting. What is theoretically reasonable and good? What can actually be done about it?

At the end, who holds that power?

Thermal oscillation cooking

So today I'm having guests over. I'll be serving pork rib rack. I love to read, so while I was reading about the best way to cook this, I found many differing opinions. To make this discussion more "scientific", I will qualify the best meat and limiting the goodness of meat to the tenderness and the moistness of meat as well as the flavor absorption.
So let's think about the meat cooking process more rigorously. First and most obvious I think is the the act of "cooking" and the associated denaturation of proteins. In more general terms, the proteins that make up the meat is deformed and often this leads to a shrinking of meat. Also obvious is the rendering of fat. The fat dissolves and somehow is expelled. There are many other levels to cooking, including the breakdown of collagen.
In our case, we are using heat to cook, so heat is the central driving force. Suppose we place the meat in the oven. What happens? Well, the meat heats up and the meat is cooked. Easy? Not so fast. The heat doesn't heat up the meat instantaneously. The heat must propagate to the center to cook everything. That's why the size of the meat is important. larger meat -> longer cooking time.
Similarly, the flavor absorption doesn't happen all at once. The flavor must diffuse into the meat. Thicker meat -> more time needed for similar flavor absorption. The absorption is faster for more porous material right? Poke holes. That works, but one possible problem is that the juices can also escape through these holes faster.
So let's talk about moistness. Moist meat is so because it has high water content. To make this possible, two things must happen. First, there should be sufficient moisture around. Second, the meat stuff must be able to hold on to moisture. The second part is often emphasized in cooking forums. Over-cooked meat cannot hold as must moisture (apparently). It makes some sense when we think about how moisture might be retained -between the flesh of the meat. If the meat contracts, this space is reduced.
Just a fun thought:
What if I oscillate the temperature so that I'm cooking at high temp and then low temp. If I choose some perfect frequency (or frequencies as a function of time), I think the cooking should be more even throughout since the diffusion would match the injection rate.
Or another possibility is to increase flavor absorption by expanding and then contracting. as the meat expands back, the flavors surrounding it will be sucked in.
Just a thought. I'm not really sure how reversible meat contraction is. I would guess that it is somewhat reversible since the purpose of meat resting is so that the juice will be retained when cutting. I venture to guess that the relaxation of the meat (reversible contraction) promotes moisture retention by increasing space between the flesh.

Mathematical models

I think I have emphasized the importance of models. In fact, I am currently trying to create a mathematical description of an observed physical phenomenon.

What is a model? In short, models are abstractions of the real world. Models provide a framework in which we can think about the problem. Using the framework, we can make predictions and explain more complex phenomena by combining different models. Every time a prediction is found correct, our models are confirmed.

Sometimes, models may lead to wrong predictions. But first, we should repeat the experiment to confirm that there were no errors. It happens... sometimes. I happen to be more of a theoretician, so I trust the model. However, as all scientists understand, the model is based on hypotheses and observations. Strictly speaking there is no reason why it should apply in an entirely new situation.

Anyhow, the incompatibility between the model and "reality" can be addressed in one of two ways. First, we can change the bounds where the model applies. For example, F=ma is a well known equation proposed by Newton. We now know however, that the Newton's second law (F=ma) does NOT apply when the speed of the moving object approaches the speed of light. To account for this, we can say that the F=ma provides sufficient match in "normal" regime. This means that it's NOT possible to describe movement of electrons using Newton's laws. As previously mentioned, the model is an abstraction. Consider some equation whose "true form" is G =  sin(x). You however have only encountered situations where x << 1. Through various measurements you conclude that G=x... this approximates sin(x) well. When you somehow become capable of measuring x > 1, you realize that the function G = x does not describe G well. However, it still holds that G = x describes the situation well x << 1.

The other way to fix the problem is somewhat obvious. We need a more "true" description that holds. This can be a modification of the previous model to cover new situations. In some cases, it might be a proposition of an entirely new model that can explain all the stories.

As a theoretician, I've always suffered from the possibility that models are wrong. At the end of the day, infinite number of models are capable of explaining a particular set of data. How do I know which one is correct?

My personal conclusion is that correctness is not important. It's the utility of these models. If a model is useful, it's correct at least partially. I can make predictions using this model with some specified confidence. This confidence comes from the experiments that confirm our model.

conservation of mass

When you lose weight, where does it go?
1. disappears
2. converted to energy
3. I poop and pee it out :)
4. I don't know
5. other: ________________
... the correct answer is 3 and 5. Surprisingly, you remove a LOT of waste through the mouth and skin. Even now, you are probably removing a lot of waste.
This waste is carbon dioxide. During respiration, carbon dioxide and water is produced as a byproduct. It's also obvious where the byproduct ends up. These are well established and known... I thought at least at first. If you didn't know this, you are not alone. Derek has asked many "normal people" and has found that some people think it's converted to energy (https://www.youtube.com/watch?v=lL2e0rWvjKI). But as a chemist, I know that conservation of mass must hold and thus that mass cannot be converted to energy (unless nuclear events take place in my body... which probably doesn't). More surprisingly, even doctors couldn't say where the lost weight ends up (http://www.bmj.com/content/349/bmj.g7257). The paper talks more indepth about the calculations. Honestly, though, I'm not sure if the "novel calculation" is new since assuming CO2 and H2O are the only byproducts, it's textbook material... Nonetheless, it is an interesting read. My only (tiny) objection would be that the fat may be transformed into other substances also. The fat burning that results in weightloss is definitely excreted as CO2 and H2O, but my guess is that some of the fat can end up as other storage form... maybe glycogen?