Antidepressants work. Why can’t we ask something more useful?

Earlier this week, The Guardian published the results of their survey, “Do antidepressants work?” Overall, responses from all five countries were remarkably similar – about 75% of people thought they did work, based on their own experiences.

This may sound exciting but actually, it’s largely a waste of time. Not only are the responses biased beyond the point of usefulness – but we already know the answer to the question.

Get some perspective

When you run a survey like this, only people who’ve had memorable experiences of antidepressants will respond – the ones who’ve had their lives transformed for better or worse. People who’ve had forgettable experiences will simply decline to fill in the form, making the results an unrealistic sample of two polarised camps.

Also, there is no guarantee that the improvements and deteriorations that people reported were due to their antidepressant. No matter how miraculous the recovery or horrific the side effect, it could’ve happened on a placebo or without a tablet at all, as part of life itself. There’s no way of telling by simply asking someone.

As a sensible scientist once said, “the world looks flat from where I’m standing”. Our own points of view are hopelessly biased, so it’s only when we look at the bigger picture that we can accurately see what’s really going on. This is why we have clinical trials – and when you look at those, the answer is clear – antidepressants work.

Even the staunchest of detractors agree. Joanna Moncrieff, who usually states that antidepressants only work via a generalised “numbing” effect, published a meta-analysis with Simon Wessely reporting that they do actually seem to work better than a “numbing” placebo. And the much debated meta-analysis of Irving Kirsch, who has written a whole book vehemently attacking the whole concept of antidepressants, suggests that they do still work in severe depression.

Of course they don’t work for everyone, no drug does. But in comparison to many other drugs in medicine, they work damn well. You’d need to give 50 people a statin to stop one of them having a non-fatal heart attack or stroke. The same number applies to aspirin. But the relapse rate in depression drops from 41% to 18% if you take tablets – the “number needed to treat” is less than 5.

In fact, an elegant paper by Leucht (2012) notes that overall, most psychiatric drugs are at least as effective as other medical drugs – but perhaps that’s a discussion for another day.

Always read the label

Sure, antidepressants have side effects, but again, so do all drugs. Here’s the side effect list for Ibuprofen, which I assume no one is debating the effectiveness of:

  • increased risk of heart attack
  • increased risk of stroke
  • nausea
  • vomiting
  • diarrhoea
  • indigestion
  • abdominal pain
  • headache
  • dizziness
  • fluid retention
  • raised blood pressure
  • gastritis
  • duodenal or gastric ulcers
  • allergic reactions
  • bronchospasm

Just because a drug has side effects doesn’t make it ineffective and certainly doesn’t justify sensationalism – just a level of caution when we prescribe it and high quality research to learn more.

Ask something more helpful

Persisting in asking if antidepressants work, when we know that they do, is an insult to those who clearly benefit from them and need them to live their lives. To tell people that they are being duped into sedation for “problems of living” by drug companies minimises their distress and devalues their right to an effective treatment. This “pillshaming” is a form of stigma which only mental health patients suffer.

We are wasting valuable time and column inches which could be spent discussing far more fruitful questions, albeit ones which wouldn’t sell papers:

  • Are we prescribing antidepressants too readily for problems that could be fixed without them – often in primary care?
  • Why is accessing talking therapy as an alternative option so difficult? Is this affecting antidepressant prescribing rates?
  • Do we discuss the indications, effects and side effects of tablets well enough with patients?
  • How close are we to tests and scans that can tell us if someone will benefit from a certain antidepressant?

We need to move on from asking if antidepressants work – and starting asking how they work best.

%d bloggers like this: