Krystallized

BBC Radio Four’s Today program ran a piece on August 2 in response to an NHS report showing a startling 500% rise in prescriptions for antidepressants since the advent of SSRIs and a 9% rise last year. Close to 47m prescriptions were dispensed in the NHS in 2011 for anti-depressants and sleeping pills. There has been a rise year on year for the last two decades.

In response to questioning on Today, Professor Clare Gerada, of the Royal College of General Practitioners, trotted out the standard lines “antidepressants work” and “antidepressants save lives”. Would we be worried if there was a rise in the use of statins or anti-cancer drugs?

The UK twittersphere exploded with tweeters taking sides. When challenged on the issue of antidepressants working Dr. Gerada insisted they do. There were predictable tweets from individuals who stated antidepressants saved my life – although as is often the case with such comments the person will often add that it was only on the fourth or fifth antidepressant that they were helped.

Claims that antidepressants work and antidepressants save lives are not evidence based. There is no question that antidepressants do things and in this sense “work” (DBM position paper on antidepressants). If the trials had undertaken to demonstrate SSRIs have an acute onset anxiolytic effect this could have been conclusively demonstrated. But the clinical trials that were done were undertaken to see if antidepressants “work” for depression and this has not been shown. They haven’t even been shown to “work” marginally better than placebo, as is often claimed.

What has been shown is that these drugs have “effects” on rating scales that regulators such as FDA can interpret as evidence they “might” work. This is quite different to demonstrating that the drugs get people back to work or save lives. It is indeed a moot point whether the law has been broken in the licensing of antidepressants in that the 1962 FDA statutes state that the agency will approve drugs that have been shown to be effective.

The rating scales moreover have been rating scales completed by doctors where some of the benefits may stem from sedative, anxiolytic or appetite enhancing side effects of antidepressants. On rating scales completed by patients, or scales less sensitive to the side effects of the medication, there is no evidence even for a rating scale benefit.

Pharmaceutical companies will quickly sue anyone who says things that the company believes it can prove in court are unsupported or injurious to the reputation of its product, as happened to Prescrire. (See also Welcome to Data Based Medicine). It is against this background that I am saying that antidepressants such as Cymbalta or Pristiq or Zoloft have not been shown in clinical trials to work.

Having made this point, I also have to say that I use antidepressants. They can have clear therapeutic effects. These effects are typically concealed by trials. If used judiciously I hope I can sometimes put such effects to good use. Whether on balance I do more good than harm is something neither I nor any other doctor can know, in the way we can know when we give an antibiotic for a life-threatening infection or a benzodiazepine for catatonia. Appealing to the literature won’t do it, as this is largely marketing copy and the better the marketing copy the more misleading it is likely to be.

The second statement that antidepressants save lives has simply got no evidence base. In 2006, the FDA in the USA got all (or most) placebo controlled antidepressant trials from companies (FDA 2006). In 2004, the MHRA in the UK got all (or most) placebo controlled trials for some antidepressants from companies (MHRA 2004). There were large discrepancies between the two datasets. But a common finding was that there were more deaths in the antidepressant arm of these trials compared to the placebo arm.

So what is happening in the case of those who tweet that an “antidepressant saved me”? There are two options. One is that they are wrong. The other is that they are right but there are a slightly greater number of voices who might have tweeted an “antidepressant killed me”. All anyone hears though are the voices saying the pills saved me.

“Antidepressants saved my life” are like the wonderful stories of people whose lives were saved when their religious medal or medal for valor deflected the bullet away from their heart. We never get to hear the stories of those who die when the medal deflects the bullet the wrong way.

It’s the same with those who swear a statin or other medical drug saved them when trials show an excess of deaths on these also. The evidence base comes back to bite us in the case of  treatment induced death – Pharmacosis – which has become at least the fourth leading cause of death. Some drug must be responsible for this, but everyone – psychiatrists, cardiologists, primary care doctors and others – look the other way and say “not me”.

If antidepressants don’t work very well, why are prescriptions rising so relentlessly? A great part of the rise stems from the fact that an ever greater number of people are stuck on these drugs permanently. There is likely no more people being put on an antidepressant for the first time each year now than there was 10 or 15 years ago, but the total on treatment continues to build because of the people who cannot stop.

A year ago in the New York Review of Books Marcia Angell wrote a two-part review of recent books including Bob Whitaker’s Mad in America, Irving Kirsch’s The Emperor’s New Drugs, and Daniel Carlat’s Unhinged that, based largely on the issue of do antidepressants work and save lives, asked the question are Americans in the middle of a raging epidemic of mental illness.

A year later organized psychiatry is still fuming. The latest response is from John Krystal, the President of the American College of Neuropsychopharmacology. After the usual arguments which amount to the claim that the evidence is not what it appears to be, that only we guardians of the flame can interpret it, and that the media should stop alarming people, he comes to an extraordinary conclusion:

 “By stigmatizing a field progressing toward a scientific foundation and by disparaging treatments that show signs of efficacy, Dr. Angell’s facile criticism of psychiatry could do harm.”

This is astonishing on a few levels. Herbalism and a range of other fields could be characterized as progressing toward a scientific foundation with treatments that show signs of efficacy. Put a sufficiently large number of people in a trial and choose your measures carefully and even snake oil will show efficacy. Dr. Krystal seems to have consigned psychiatry to a lower level in the pecking order than physiotherapy and other paramedical disciplines. Many in other branches of medicine will likely smile wryly.

If psychiatry is only progressing toward a scientific foundation and its treatments only show signs of efficacy, the harm lies in the fact it has sold itself as a fully scientific discipline with treatments that work and save lives. If this isn’t the case, people need to be told.