This is the second post in a 6-part BarMittzva Romba series.
Every product is built from a raw material. The raw material puts constraints on a product developer. There may be difficulties fashioning the product from the material, or the material may be costly or scarce. There is the delicate matter of how the mark-up from raw ingredient to product is perceived – the market will only bear so much.
In the later decades of the twentieth century, bottling water produced a close to perfect product. A few years earlier no-one had imagined that something as ubiquitous and inexpensive as water could be bottled and sold at such a mark-up in places where tap-water quality was good. In the case of bottled water, except in cases of water shortage, almost all the value of the product comes from its marketing, and lies in the eye of the beholder.
In the nineteenth and early twentieth centuries patent medicines, which often contained little more than water were equally perfect products. The mark-up on these proprietary products was even greater than for bottled water now. The creation of a market in patent medicines was perhaps the single greatest contributor to the development of modern marketing, and the greatest influence on the marketing of medicines to this day.
Compared with patent medicines and bottled water, prescription-only drugs are complex products that require two sets of ingredients. First, they require a chemical that forms the basis for a medication. Second, they require information to transform the chemical into a medical product. This information specifies the conditions under which the chemical might best be used, and its likely effects at particular doses and in particular circumstances.
The costs of chemicals today are little more than the costs of bottling water. The information required since 1962 to make a chemical into a drug at least superficially seems not so readily manufactured. But in fact, the current constraints on the informational component needed to produce a medicine are so lax as to make this information close to the perfect raw material. It took companies a while to realize that they are all but able to invent this information.
The invention of this information adds apparent value that supports mark-ups of an order not seen in any other market. Were the value real rather than apparent, these mark-ups might be justifiable but far from being real what appears to be added value in many cases increases rather than reduces the risk from the chemical.
Problems with the informational component of medicines lie at the heart of the therapeutic paradox and are the reason why the more we access medicines, the more problems we have.
At the center of this information is the requirement introduced in 1962 that drugs would only be licensed if they could be shown to be efficacious. The development of randomized controlled trials (RCTs) made it tantalizingly possible to insert this requirement into the system – a requirement that seemed to force the financial camels of the pharmaceutical industry to get through the eye of a scientific needle if they were going to make money out of sick people.
Controlled trials were introduced in the 1950s after they had demonstrated that they could weed out unwarranted claims for treatment efficacy. This appeared to raise the bar to entry into the market in a way that would keep the purveyors of patent nostrums out of healthcare. Trials have since been portrayed as a method that was going to give medicine the objectivity and certainty linked to sciences like physics and chemistry. Medicine could become evidence-based, and in the process would harness the pharmaceutical industry to its purposes.
If the use of trials had been restricted to keeping drugs that don’t work off the market, this use would have been a major contribution to drug safety. If trials had been rigorously used to weed out ineffective treatments they would have helped produce a dramatically different market to the one that has developed. For better or worse a number of drugs we currently have would not be available.
More to the point RCTs would not have become the vehicles used to sell drugs, hide side effects and drive clinical practice they have become. Instead as evidence based medicine has developed we have entered a stranger and stranger world.
First, the primary use of RCTs – to disprove claims of efficacy – has been completely subverted. Even if a preponderance of the RCTs undertaken fail to show a drug is of benefit, if any trials show hints of a positive outcome the treatment is likely to be permitted on the market. The way the system in fact works, the makers of snake-oil would have little problem getting on the market. In the case of some of the best-selling drugs in modern medicine the preponderance of evidence suggests the treatment may be no more helpful then and distinctly more harmful than snake oil.
Bizarrely, the treatments supported by independent reviewers taking an evidence based medicine approach are in many instances treatments that have the greatest amount of negative trials. This happens as guidelines have to be based on published trials and can neither take into account the negative trials that remain unpublished nor access the data that shows many negative trials are published as positive studies.
This has been clear for decades. Finally last year Peter Doshi and colleagues who recently reviewed the evidence on Tamiflu came out straight and said it is not possible to assess the efficacy of a drug without full access to all the studies that have been done. Yet on the basis of the original published data, governments throughout the world spent several billion dollars stocking Tamiflu in 2007-2008.
Second, running studies in which there are huge numbers of participants makes it likely that some irrelevant benefit will be shown for a drug. Snake oil could almost certainly be demonstrated to be of benefit in large studies. Far from seeing through what is happening doctors can be relied on to be more impressed by a study that contains thousands of patients than by one that needs a handful of patients to demonstrate a benefit. The most prestigious journals are also more likely to take a multicenter study with thousands of patients showing a trivial clinical benefit, than they are to take a small study showing a clear cut clinical effect.
Third, the only data from a trial that can ordinarily be generalized are the estimates of the reliability of a trial’s primary outcome measure. But in practice almost anything that turns up in the course of a trial – that is of use to a company – is taken as having been established simply by virtue of the fact that it happens in the course of an RCT.
Few trials include the right instruments to measure outcomes other than the primary outcome. This is particularly true for adverse events where in addition the data are commonly creatively coded or relocated or otherwise massaged to make problems vanish.
Fourth, trials are primarily run by pharmaceutical companies. From the start these trials are aimed at producing knowledge to serve a purpose rather than aimed at producing knowledge. Any inconvenient knowledge that turns up in the course of the studies is likely to be discounted.
RCTs are indeed a gold-standard method as is so often claimed by the adherents of evidence based medicine. However, for the reasons just outlined, they are in practice a gold-standard method for hiding adverse events rather than for demonstrating efficacy. They might have made a significant contribution to safety if their use had been restricted to weeding out ineffective agents, but in practice their use has been to conceal adverse events and as such they have been detrimental to the development of a comparative safety market.
There is however little or no recognition that the mantras of gold standard and statistical significance and evidence based medicine are essentially rhetorical tropes that need dismantling.
A new default has been created. Where before drugs were viewed as poisons with the art of medicine lying in an ability to find the right dose in order to balance the risks and benefits of treatment, now supposed proof that they “work” has transformed these chemicals into fertilizers to be administered as widely as possible. Where once the greatest art in medicine had lain in knowing when not to treat, now it seemingly lies in knowing how to get people on as many drugs as possible for as long as possible.
This stems from the use of controlled trials. Extraordinarily a technique that was introduced to contain company claims has become the means by which companies create knowledge within health care and drive the sales of drugs. We have almost arrived at the point within healthcare where if parachutes hadn’t been through an RCT no one would be let use them.
But it is the actual conduct of trials that makes them into the alembus in which pharmaceutical company alchemists can transform base metal into gold.
First, an initial set of trials are conducted in healthy volunteers. These trials uniquely reveal the hazards of drugs in a way that clinical trials in patients do not. Perhaps because this is the case, the data from these trials are impossible to access. While the data from clinical trials are also concealed, there are now registers of trials undertaken in clinical conditions. There is no register of healthy volunteer trials and publications from these studies are commonly deeply misleading.
The trials that doctors and others hear about are undertaken on patients. In these, patients volunteer to participate. If the treatment is new, they unwittingly take the risk of ingesting chemicals that will likely prove too toxic to market. They do so without being informed of these risks. As the exercise is billed as scientific, most participants likely believe that the data that results from their participation will be made available to experts and will contribute to a knowledge base that is incrementally driving medicine forward.
In fact while clinical trial registers now offer some evidence that the trial took place, the actual data is sequestered by companies. When it comes to transforming the data into information that will shape clinical practice, companies can select which trials they wish to publish. They can also select the data from these trials that suit their purpose. In some areas of medicine, a third of trials may remain unpublished and of those published up to one third are portrayed as positive when regulators or others who have seen the data deem them to be negative studies of the drug.
The data are written up to produce a “publication” which is the primary marketing tool of companies. Owing to a supreme sleight of hand, these marketing aids are designated as scientific articles, although they fall at the first scientific hurdle by not making the data on which claims are based publicly available.
Once the publication is complete, in order to add value to the marketing copy, ghost writers add the names of distinguished medical academics. Academics provide this service in return for very modest amounts of money. The publication is then sent to a journal. Unlike quality newspapers which check the integrity of the primary sources on which a story is based, journals never do so. Even after the claims in an article have been shown to be fraudulent, journals refuse to retract these publications so that they continue to influence medical practitioners.
These publications are then embodied in guidelines that de facto require doctors to use the latest on-patent and more expensive drugs rather than older, less expensive and more effective drugs. The guidelines process is one that enables companies to co-opt even the most independent and company hostile academics into endorsing their products.
Another confusion may demonstrate how badly efforts to constrain commerce within the clinical trial process have infected therapeutics. In 1962 the idea was to demonstrate a drug worked before it was let on the market. As a result studies are conducted in disease states. If the trial is not negative, companies are then licensed to advertise the fact that drug Y works for condition X. But the drug in fact may work for many other conditions or purposes, and indeed work better for these other conditions. Imipramine for instance is a far more effective treatment for panic disorder that any of the drugs that have been licensed for this purpose. SSRIs are more effective treatments for premature ejaculation than they are for depression.
If a company promotes its drug for some purpose without undertaking a study in that condition this is called off-label promotion. There is a great deal of concern about off-label promotion. As a result of this concern, a majority of doctors think that they cannot prescribe off-license. And guidelines only endorse the licensed indications of a drug. This effectively hands medicine over entirely to pharmaceutical companies.
Having a resource like this that can be molded into virtually anything that the company wants might be expected to cost money. But in fact it costs very little. The patients who generate the data are paid nothing. They take risks with compounds that have unknown unknown risks, known (hidden) unknown risks, as well as known risks, for free. They have been sold the idea that it is their civic duty to participate. In the 1960s this civic effort did liberate us from scourges that had plagued humanity for millennia. The same effort today is inflicting harms on people that will take decades to eliminate.
Doctors also come cheap. In many jurisdictions, doctors have been told that it is government policy that they “partner” industry and one way in which they can do this is to participate in clinical trial networks to make the testing of new drugs quicker and easier. Policies like this stem from government efforts to keep pharmaceutical business “in country” even though it is difficult to see the economic return on this. The doctors running the trials for companies are paid by the state. Any notional fees they get are a fraction of the true costs of the exercise.
Meanwhile, for economic reasons pharmaceutical companies are relocating most clinical trials to countries such as India, where the oversight and costs are less. The publications that come out of these Indian trials are still likely to have Western academics listed as their authors.
As a symbol of how the field has developed, it is hard to beat the fact that as of 1962 the only drug that had been demonstrated to be effective and safe in a placebo-controlled trial before being brought to the market was thalidomide. Fifty years later antidepressants have become the most commonly used drugs in pregnancy, despite increasing evidence they double the rate of birth defects and miscarriages and likely cause significant cognitive delay in the children born to mothers who have taken them through pregnancy. The increased prescriptions for antidepressants is happening on the basis that these drugs have supposedly been shown to be efficacious and withholding something that is efficacious is increasingly portrayed by public relations and media companies as unsafe and unconscionable.
In terms of policy, the one indisputable fact for both conservatives and liberals is that this market is not free. Companies are able to sequester data so that it is impossible for consumers (either doctors or patients) to know what the risks and benefits of a treatment are. The consequent labeling of most drugs is deeply misleading, has in some cases been charged to be fraudulent, and is in all cases in breach of the norms of science.
These norms include access to the data and a commitment to empiricism. They make science democratic and the ultimate free market. When this market was free, it worked to enrich and liberate us all. Within healthcare however companies have been able to create the appearances of science and deploy these to their benefit but our detriment. As healthcare costs escalate dramatically based on a subversion of the scientific market, while actual health deteriorates, this deployment of the appearances of science poses an increasing threat to the economies of developed countries.
This deployment of the appearances of science enables companies to generate mark-ups not seen since the days of patent medicines. In Anthony Powell’s Dance to the Music of Time, when young the author faces what he thinks are unique and individual situations only to find later in life that even the most distinctive of individuals and unlikeliest of events replay themselves.