False Friends

‘Evidence’ is what the French call a false friend. You think you understand the word but you don’t. In French or Dutch, the Evidence in Evidence Based Medicine means that something is self-evident – as in using a parachute when jumping from a plane or penicillin for septicemia or an antipsychotic to tranquilize. You don’t need a clinical trial to work out what the right thing to do is or to prove what has happened.

What you see is what you get.

In English, the Evidence in Evidence Based Medicine (EBM) means something that needs proof. If it has to be proven, then it’s not evident. When your doctor practises according to EBM, you are usually being asked to take something that is not obviously good for you. Even if the evidence is provided by completely independent sources there is still considerable scope for experts to get things wrong.

But in the case of EBM most of the evidence is produced by pharmaceutical companies who have as many tricks to maximize the positives as they have tricks to minimize the negatives (as outlined in recent blog posts). Where once we might have put our Trust in Evidence, now we are being Trussed in Evidence (see Pharmageddon).

Under the influence of EBM, the English use of the word is spreading, which plays right into drug company hands. They increasingly manage to get doctors to get patients to jump out of planes with little more than a few ghostwritten articles to bear them up.

There is a lot at stake.

It is this gap that charlatans and quacks exploit

There is no doubt that doctors and patients can get the evidence of their own eyes badly wrong. Both are biased to seeing cures and hope where either none exists or the improvements have come about naturally. It is this gap that charlatans and quacks exploit.

Randomized controlled trials (RCTs) were designed to save us from the exploitation of charlatans or the false hope of the naïve. They did this originally by showing that patent medicines or other panaceas didn’t in fact work. They were designed to buttress skepticism.

Using trials to keep drugs that don’t work off the market would be a way to promote safety but this is not what trials are used for.

When it comes to the things that count, RCTs are not the gold standard

Clinical trials have been turned inside out and have now become the primary mechanism by which treatments with minimal or no effects are trumpeted as working. They have become a means to prevent us from believing the evidence of our own eyes. They are inducing a psychosis.

And far from making us safer, RCTs compromise our safety because trials as we use them now have never been a good way to establish safety, as this series of Spin & Data blog posts will bring out. When it comes to the things that count, such as discovering obvious benefits or hazards to a treatment, RCTs are not the gold standard.

Consider this: When we get access to the data from trials we find that SSRIs or related drugs have about the same effects for primary care nerves as clinical trials for opiates, benzodiazepines, stimulants, or nicotine have shown and could doubtless be shown for alcohol.

Imagine giving nicotine or alcohol as a matter of public policy

Imagine on the basis of 6-to-8-week-long trials, we were to let companies heavily promote nicotine or alcohol as antidepressants – almost to the point of making it public policy that they be used widely. And certainly having their use written into treatment guidelines that make life a lot more difficult for doctors who might have doubts about the wisdom of treatments like this being given for months or years.

What would we find? That many people would apparently recover on nicotine and alcohol and be able to stop after a few months of treatment. Some patients and doctors would swear by the benefits. Any critics of treatment who said the effects were all in the mind (just placebo effects), would be dismissed as plain wrong. Anyone who has a glass of wine knows that it self-evidently has very real effects. And in the French sense of Evidence, the doctors and patients would be right and the critics wrong. But in this case both doctors and patients would know the pitfalls of both French and English Evidence.

Many others wouldn’t do well. Some others would do reasonably well but would not be able to stop. They would feel more anxious when they tried, and their problems would be relieved by going back on treatment, attributed by doctors to their illness: “Think of it this way, you have the equivalent of diabetes and you have to continue taking the pills.”

We would find an increased rate of birth defects and other complications when these treatments were used by women of child-bearing years. Academics would say that leaving these nervous problems untreated causes birth defects.

We would find an increased rate of suicides – academics would blame these on the illness.

The long-term RCTs of alcohol and nicotine have never been done

When given for years or even decades, who knows what we would find. The long-term studies of alcohol and nicotine have never been done. The English kind of Evidence just isn’t there.

Is this a fanciful comparison? We have no way of knowing. If the risks are as bad on antidepressants as they are for alcohol or nicotine, it might still be worth taking these risks for severe mood or anxiety disorders. The benefits of the original antidepressants for melancholia, or antipsychotics for acute psychoses or manic states, for instance, were Evident; they were not discovered in or proven by controlled trials. The trial evidence we have amassed since has in fact obscured rather than revealed the Evident, just as trials for alcohol or nicotine or stimulants would obscure rather than reveal the obvious effects of these drugs for primary care nerves.

Either way people should be aware of what we do and what we don’t know. They should know that in most cases these drugs are not self-evidently a good thing. They should be told they are likely to be at least as risky as alcohol, nicotine, stimulants, benzodiazepines, and opiates. Used wisely, an evident benefit can be helpful, but a blind adherence to evidence is not wise.

We throw caution to the winds, almost to the point of jumping out of planes without a parachute – how else to explain giving these drugs to children?

Faced with a doctor persuading them to take alcohol or nicotine for their depression on the basis of Evidence in the English sense of the word, most people would be appropriately cautious because of their familiarity with the Evidence in the French sense of the word for these drugs. But when it comes to SSRIs, we throw caution to the winds, almost to the point of jumping out of planes without a parachute – how else to explain giving these drugs to children?

This was the basis for constructing the Data Based Medicine (DBM) papers on antidepressants – see Call for Papers. These are a mixture of Evidence Based Medicine (French) and Evidence Based Medicine (English).

As for RCTs, these are wonderful when used to eliminate treatments that don’t work or have minimal effects. Used for any other purpose they are the ultimate way for drug companies to hide bodies, as the next blog posts will bring out.

Just as problematic, RCTs are a way to snare anyone concerned about overmedication into disputes about just how many angels there are on the head of that pin. A battle that the critics typically lose in the public domain at least because people who have had a glass of wine have Evidence they feel comfortable with.