Clinical Trials – Tricks of the Trade

Comments (1,915)

Drug Approval – Clinical Trials – Lies, Damn Lies and Statistics

Bringing a new drug to market is an expensive business. Although these days an entirely new type of drug is comparatively rare and the majority of new patents are tweaked versions of existing medicines, they still need to get approval from the relevant licensing authority (e.g. the FDA in the US and the MHRA in the UK).
Before the drugs can be approved they have to be tested and evidence produced to show that they are both effective and safe. The way that this is done is by undertaking clinical trials.

Drug Approval – Clinical Trials

The first stage in the clinical trials process is normally to give the new drug to some healthy paid volunteers (often young men) and to watch what happens to them.
If all goes well and these volunteers do not experience what are deemed to be unacceptable side effects from their exposure to the drug, the clinical trials will proceed to the next stage and a larger group of patients (maybe around 200 of them) with the relevant condition will be treated with the drug to determine whether or not it seems to be effective and at what dosages.
Assuming that Phase II of testing has been completed satisfactorily, the drug company will move on to Phase III of the clinical trials, where the drug is tested on hundreds or even thousands of patients.
It is at this stage, where things have the potential to get really interesting. So let’s start off by considering how good clinical trials should really be run.

Drug Approval – What Makes for Good Clinical Trials?

Good clinical trials – ones that are fair and transparent – start at the design stage. The trial design should clearly identify things like: what factors make a patient suitable for inclusion in the trial and who should be excluded; how patients are assigned to treatment and control groups; what outcomes will be measured and so on (this is far from an exhaustive list).
Trial protocol document should also be produced to ensure that the clinical trials are performed in the same way, on patients with the same characteristics, regardless of which researcher is actually involved in the administration of treatment and monitoring of outcomes. This document is especially important for clinical trials where more than one group is participating in the study.
Clinical trials should be randomised – that is to say that any participant will be randomly assigned to either the treatment group or control group. It should be “blinded” – whereby the people administering the treatment do not know whether they are giving the patient the new drug or the control. The control treatment should ideally be the current most effective treatment available, but is more commonly a placebo (often referred to as a sugar pill).
All of these steps are required to try and prevent bias (whether unintentional or not) from affecting the clinical trial’s outcome.

Drug Approval – Why Clinical Trials Are Not Always What They Seem

The trouble with the clinical trials process is that most of them (around 90%) are run by the pharmaceuticals manufacturers themselves (or commissioned by them) and they, of course, have a vested interest in getting their new drugs licensed.
So what can go wrong with Clinical Trials? The simple answer: lots of things!
At the most fundamental level, the design of clinical trials can be flawed. There are lots of methodological errors that can be made.
However, in terms of deliberate manipulations by the pharmaceuticals manufacturers, it is far more likely that the analyses of theoutcomes of clinical trials will be distorted and there are plenty of ways in which this can be done.

Control Data

One of the core precepts of a scientific approach to studying the effectiveness of a medicine is that it is compared against something. In an ideal world, where the aim was purely to come up with better treatments, the new drug would be compared with the most effective treatments currently available.
Since we live in a real world that is far from ideal, and profit is the primary objective for new medicine development, many clinical trials compare the new drug against a placebo.
A placebo is something that appears to be a medication, but which has no active ingredients. The reason that some form of treatment has to be seen to be being given is that it is well documented that patients’ health can improve simply because they believe it will.
If you give a patient a treatment that he believes will make him better, then the chances are that his health will improve, even if the treatment is simply a sugar pill or a saline injection. This is known as the placebo effect.
If every patient involved in clinical trials believes that he is being treated with a drug, then trial and control groups are both as likely to have patients benefiting from the placebo effect, so that any improvement in the drug trial group (over that of the control group) can be assumed to be as a result of the treatment.
While the results may show that the new drug is more effective than no treatment however, it is much harder to establish whether it is any better than existing treatments.
This is important because new treatments tend to cost a lot more than well established ones. Also with existing drugs, the likely side effects will have been established over the drugs lifetime and this includes problems which may only appear after long-term use. With new drugs, the only evidence is of problems that may have arisen during the clinical trial and frequently a much wider range of side effects emerges when a drug appears on the open market.
If the new drug is compared with an existing treatment as part of its clinical trials, this may seem to give the results much more credibility. However there are still a few tricks to watch out for. The first thing we need to consider is whether like is being compared with like. Is the new drug being given at a comparable dose to the control drug?
Some clinical trials produce what seemed to be very favourable results for the new drug being trialled, but further investigation revealed that they were being compared to a drug that was being prescribed at half the dosage. Hardly a fair or accurate comparison.
There are other manipulations that can make the new drug seem better than the control. One reason a new drug may be preferred to an existing treatment is because it is believed to cause less side effects. Drug companies can stack the odds in their favour during the drugs clinical trials e.g. by giving very high doses of the control drug, or by increasing its dosage faster than recommended, both of which are likely to result in more side effects.

Selective Reporting

This is a common approach to presenting the results of clinical trials in the most positive way possible.
If a drug company has results that are not as good as hoped across the whole range of people in the clinical trials but has good results for one particular sub-group (say young, otherwise healthy, men) then this is the evidence they will present to demonstrate effectiveness.
Now, when it is prescribed in the real world, the people receiving the drug may not be much like this small sub-group, in fact they may be far more like the overall group, in whom the results were unimpressive, but the evidence that shows this will be carefully buried.
Another approach to selective presentation of data is known as “cherry picking”. If a drug has been tested in multiple clinical trials and some results have been more positive than others, then only the positive one’s will be submitted in evidence. The drug companies will try and hide the results of the not so positive clinical trials (although now that they are legally required to publish all data, at least in the US, this is harder to do).

Fudging The Issue

I’m afraid that this is where I get onto a bit of a hobby horse of mine: statins and cholesterol. They are excellent examples of fudging the issue.
The theory is that high cholesterol levels are a risk factor for cardiovascular disease and that lowering cholesterol should therefore help to reduce heart problems (and I won’t get into all the potential issues there are with that theory, at least not here).
So if you have a drug that you want to use to treat/prevent cardiac problems, what you should be measuring during clinical trials is the number of people in the treatment group suffered from a serious cardiac event or died from heart attack, compared with the control group.
So, what if your drug turns out to be not much good at preventing and reducing heart problems, even in people who have high cholesterol? Well first off you start by focusing on those people who are most likely to benefit from treatment – that is to say the most high risk patients. And when those results are, frankly, not that impressive, then it’s time to fudge the issue.
So instead of focusing on how good your drug is at dealing with heart problems (not very), you look at how good the clinical trials show that it is at lowering cholesterol and it turns out that it is very good. Well, the argument then goes, high cholesterol is bad since people who have serious heart problems have been shown to have higher cholesterol levels, so lowering cholesterol levels must be beneficial – so the drug is effective. (For more information about statins see the article: Statins for Lowering High Cholesterol.)
While I’m on the subject of statins, cholesterol and heart disease lets consider another issue: statistics.

Statistical Manipulation

Correlation is a commonly used statistics term which means that two data sets seem to be linked in some way. For example evidence shows that high levels of cholesterol are common in patients suffering from serious heart conditions.
The trouble is that correlation is not the same as causation. Just because two things have an association doesn’t mean that the one caused the other. It is possible that high cholesterol may cause heart problems, but it is also possible that higher cholesterol levels appear because there is a problem – cholesterol is part of the bodies healing mechanism (to understand more about how cholesterol works, see the article: Cholesterol Explained).
So, back to clinical trials. Just because one thing happens at the same time as another doesn’t mean that the one caused the other, you have to consider all the other factors that are likely to be involved.
If a study reported that children who were overweight tended to have more fillings, you could argue that being overweight increased the likelihood of tooth decay (or vice versa). In which case you might conclude dieting could improve dental health or improved dental health would result in weight loss. But then you might be ignoring the fact that the children with the most dental decay were the ones that ate most sweets. Or perhaps it was the level of parental involvement that made the difference… parents who supervised their children’s brushing were more likely to ensure their children had a healthy lifestyle. Or some other reason altogether. (By the way, this example is spurious – something I just made up.)
That’s the trouble with statistics. They are only as good and objective as the person presenting them and yet they are presented as science and tend to give credibility to a report that it may not deserve. There is a lot more that can be done with statistics to help massage a clinical trial’s results.
Statistics can help you lose inconvenient clinical trials results in a variety of ways. Ignoring people who have dropped out from the treatment group is one method. The reasons for this are pretty straightforward – people are more likely to drop out of a study if they have been having problems – perhaps their condition has worsened or they have experienced unpleasant side effects, they may even have died (it’s happened). Clearly, excluding drop outs will help to make your clinical trials data look better.
Another way of tinkering with the data is to see if there are any results that are markedly different from the majority of the results. If these results help to make the study look better then they can be kept in. On the other hand if they suggest something negative, then they can be ignored as anomalies and not presented as part of the clinical trials results.
Even though the control and trial groups are randomly assigned, it can happen that when measures are taken at the beginning of the trial, one group appears to be doing better than the other.
If the evidence is that the control group is for some reason doing better at the commencement of clinical trials, then this can be adjusted for, to level the playing field so that subsequent measures reflect the change from the starting point.
Now, in theory, if it turns out the treatment group is the one that is doing better, then an adjustment should also be made for that… but since it makes the data look better, well it just might not happen.
The last manipulation of clinical trials that I’m going to look at (although not the last of the tricks pharmaceuticals companies get up to in order to get new drug approval) involves mucking around with the timescale.
The original trial design probably specified the intended duration of the clinical trials, but if altering this may offer better results then that’s probably what will happen. Now there are good reasons why clinical trials may be stopped earlier than originally intended and one of these is to do with the negative events associated with a treatment.
A trial of a cholesterol lowering drug showed that there were significantly more cancers and cancer deaths in the treatment group than in the control group and the trial was therefore stopped on ethical grounds.
There are however less ethical reasons why clinical trials may be stopped early. For example, if the early trial results are looking positive, then the drug company may decide to quit while they’re ahead. After all longer term results may not be so great and the longer the trial goes on the more side effects may emerge.
On the other hand, if the trial is not a total disaster, but not quite good enough to be deemed as having significant results (meaning that any positives are likely to be more than chance) then the drug company may choose to extend the trial, to see if that gives them the result they are looking for.

The Results of Manipulating Clinical Trials

If you were thinking that only a few drugs have been given approval as a result of drug companies using these tricks, then think again.
According to an independent review, only about 15 percent of new drugs work as stated, and most (whether they offer a benefit or not) come with serious side effects.
An analysis of 111 final drug applications showed that 42 percent were missing adequate randomized trials, 40 percent had inaccurate dosage testing, 39 percent failed to show drug efficacy, and roughly half revealed the drugs to have serious adverse side effects.
In another study, which looked at 192 trials of statin drugs, it was found that almost half of the trials lacked adequate blinding, an extremely important aspect of clinical trials design. Not surprisingly, the researchers found that those studies with adequate blinding were less likely to report results favoring the test drug.
This study also found that where 2 drugs were directly compared, the drug which emerged as the best depended on who was funding the trial. If the reported results favored the test drug, the trial was about 20 times more likely to be funded by the maker of the statin rather than the comparison drug company. And if the interpretation of the results favoured the test drug, the trial was about 35 times more likely to be funded by the maker of that drug rather than the comparison drug.
In an effort to try to improve the accuracy and completeness of drug trial reporting, the FDA Amendments Act 2007 was introduced. Under this Act, all trials supporting FDA-approved drugs must be registered when they start and the results of all the outcomes declared, as well as specific details about the trial protocol. All this information must be publicly posted within a year of drug approval on the US National Institutes of Health clinical trials.gov site.
In addition, in 2005, the International Committee of Medical Journal Editors began requiring that a trial be registered before enrolling patients in order to be considered for publication, thus creating a record of the planned study outcomes before the study’s initiation.
The theory is that if trial protocols are made public in advance, a trial sponsor is less able to manipulate or selectively publish the findings.
Although these requirements may eventually help to improve the standards of the performance and reporting of clinical trials, it may take some time in coming.
A study iIn 2008 showed that over half of all supporting trials for FDA-approved drugs remained unpublished 5 years after approval. And in 2010 an investigation into unpublished stroke research data revealed that 19.6% of completed clinical trials were not published in full.
Even where the study details have been registered at (or prior to) commencement of the clinical trials, there is no guarantee that the details that are published in medical journals will accurately reflect this.
More than one study has found a surprising number of clinical trials whose results were registered with the FDA, did not fully match what was later published in a journal - the statistics looked better, or certain failures were omitted, in the published version.
Since drug companies have so many stratagems to improve their chances of new drug approval, change’s to the system are essential to help redress the balance and ensure that the new drugs that are licensed are both safe and effective.
In an ideal world, all clinical trials would be independently funded and conducted; but at the very least there should be far more independent scrutiny of the way that clinical studies are designed and conducted, with critical analysis of any evidence the drug companies present as part of the licensing process.

Pills for Profit Part 01

Comments (248)

No one wants to be ill, but when you think you’ve got more than a slight ailment your first port of call is normally your doctor. I guess most of us believe that they offer us the best treatments to deal with our illnesses as effectively as possible.

I know that this is what I used to believe, but nowadays I’m a lot more cynical. However I don’t blame the doctors, they can only work with the treatments that are available to them. No, I put the responsibility for the lack of truly effective treatments squarely on the shoulders of the pharmaceutical industry.

Once upon a time, new drugs were developed that really did have a radical impact on our overall health, because they worked and nothing like them had existed before. I’m talking about drugs like antibiotics, which have helped to prevent huge numbers of deaths over the decades, by making it easy to treat diseases that regularly used to kill thousands or millions of people.

From such astounding steps forward in the treatment of diseases came the widespread belief that new drugs must offer better and more effective treatments – after all, newer is better in so many other aspects of science and technology, why should it be any different for drugs development?

Well let’s start by reminding ourselves that drug companies are, first and foremost, money making machines. They have shareholders who invest in them for the returns that they expect to see and the company directors’ primary responsibility is to run a profitable business. Pharmaceutical companies are not philanthropic organisations with the goal of improving human health and wellbeing.

You may argue that by developing better medicines, drug companies will become more competitive, win more of the market share, and make higher sales, increased turnover and profits. And it is true that genuine innovation is one of the ways that drug companies can make profit. However true innovation is a high risk strategy, as the amount of R&D required to develop a completely new approach is always going to be high, while the chances of a dramatic success may be low and a long time in coming.

Instead drug companies spend a lot of time tweaking existing medicines, looking for marginal improvements to maintain their competitive edge, or products similar to the successful drugs developed by other companies, so they can grab a chunk of the market share. Real innovation does exist, but it is the exception not the rule. Most of the major breakthroughs in treatments come from dedicated research establishment often, funded by charities, to search for a cure/treatment for a particular treatment.

In any case, are drug companies primarily interested in looking for cures?

With a cure the cause of the underlying problem is removed and the symptoms disappear and don’t come back (or not unless it’s something infectious that you’re unfortunate enough to catch again).

A treatment however, is something that tackles the annoying/distressing symptoms of an illness without actually removing the underlying cause. Treatments offered by modern medicine typically have to be used for prolonged periods and symptoms may reappear if the treatment is stopped.

From the drug companies’ perspective, producing treatments is far more lucrative than developing cures. In terms of human health and wellbeing this is very bad news indeed.

So, if the pharmaceutical companies are focusing on treatments, at least these treatments should be getting better and more effective, shouldn’t they? After all why else would a doctor prescribe them?

Well, there is quite a bit of evidence to suggest that costly newly licensed medicines offer minimal or no new benefits when compared to existing treatments (which may be available in generic form and therefore much cheaper). The trouble is, when medicines are tested for effectiveness, they are generally tested against a placebo (sugar pill) and not the most effective (or indeed any) currently available treatment. This makes it hard to judge whether a new drug is in fact better.

When a new drug is launched, it is sometimes claimed that it is better than other treatments because it causes less side effects. This may not, however, be strictly true. Just because fewer side effects have been identified during the clinical trial stage, doesn’t mean that they won’t occur when the drug is used by a wider population. Trials are often carried out with a specific type of individual and the trial group may not reflect the general population to whom the drug will be prescribed.

It is very common for a host of previously unidentified side effects to emerge once a drug has been in use for a while. And that is only the side effects that develop in the short term. Taking a new drug that is intended for long term use when it has only been tested during comparatively short term trials, must be considered a risk. But all too often patients are unwitting guinea pigs.

So, given that many new drugs may little or no better than existing treatments, why do they get prescribed? There is a very simple answer: marketing. This is an area I shall be looking at in detail another time.

About PharmaceuticalsIndustry.org

Comments Off

The top companies in the pharmaceutical industry are amongst the most profitable businesses in the world. The largest pharmaceutical companies hold the rights to drugs worth many billions of dollars a year.

So how do these major pharmaceutical companies earn their money? It certainly isn’t by finding miracle cures! The ‘holy grail’ of the pharmaceuticals industry are drugs that treat (but not cure) the symptoms of some of the most common chronic conditions so that continued drug use is required throughout a patient’s life.

And if you think the pharmaceuticals industry is run on an ethical basis, then think again!

Marketing and PR are by far the biggest costs for these multinational pharmaceutical companies. They spend only about half as much on ”research”, including clinical trials, to justify the use of these drugs. And the studies may be structured in such a way as to exclude data (such as patient death) which won’t look good in the results. You just have to learn a little about the suppressed data and biased analysis that allows drugs to pass the regulatory tests, to know that it’s the bucks that count, not the patients health and well-being!

The release of the drug is not the end of the drug trial period. The first years of a drug’s release are also used to assess the drug, during which time many more side effects may come to light. People don’t realize that they are acting as guinea pigs and that the medicine they’re taking hasn’t been studied for effects of long term usage. Drug companies allocate a huge fund for out-of-court compensation payments for the problems that emerge during this testing in the market place.

Once the drug company has amassed the positive “evidence” for a drug, experts are offered substantial incentives to validate the research and, similarly, regulators may be incentivized to expedite the release of the drugs. Incidentally, the regulators view the drug companies as their customers, not the public, and it is the customer they wish to keep happy!  Once approved, the drug companies then focus huge resources to get health care professionals to prescribe their drug in preference to another.

The same experts that are funded by the drug companies also sit on the committees that set the levels at which it is recommended that symptoms/conditions should be treated. The lower the thresholds are set, the greater the number of patients treated and the more revenue the drug companies earn – some of which then ends up in the experts’ pockets.

Learn more about how pharmaceutical companies control almost every aspect of the pharmaceutical industry, including influencing the supposed watchdogs! Get to grips with just how ineffective some of the top-selling, highest-earning prescription medications really are.

Now that we are in the know, we will never take a prescription medicine without first of all researching it. Once you’ve read the articles about how a pharmaceutical company operates, you may feel the same!

DISCLAIMERS:

Please seek professional medical advice when considering any new course of treatment. Be aware that different treatments may affect individuals in a variety of ways and that what works for one person may not work for another. Likewise a medicine that causes problems for one person may produce good results for another and remember that some medications may interact with other drugs and even food stuffs.

The comments posted on this site represent the views of individuals and do not constitute our endorsement of any product or service. Whilst we have made every effort to ensure that the articles published on this site are accurate and supported by research, they are for information purposes only and do not constitute a recommendation and should not be relied upon when making desicions about your health care.

We do not take responibility for the material posted on our websites by our contributors, nor for any comments, remarks and opinions that may be posted anywhere on this website. Any such contribututions, comments, remarks and opinions should not be considered to be representative of the views of the websites’, hosts, owners and operators.