Monday, October 04, 2010

The best evidence is no evidence

Science is based, ultimately, on evidence. Unfortunately, 'evidence' is not a monolithic construct and there are varying levels of evidence from the weak to the stronger. Based on your understanding of science, in what order would you put these various forms of evidence?

  • Meta-analysis

  • Clinical trials

  • Case studies

  • Anecdotes

If you guessed that they're ordered from weakest to most powerful evidence, then you're in good company.

EoR has already noted that glucosamine and chondroitin provide nothng more than a placebo effect. Some vets, however, argue that this doesn't apply to horses (where the use of such products is extremely popular).

Vets and supplement manufacturers have said there is no evidence that two dietary supplements declared ineffective for humans last week are also ineffective for horses.

That's an interesting stance to take: absence of evidence is proof of effect. Of course, drugs ('neutraceutical' and 'dietary supplements' are weasel words) for humans don't necessarily function the same way in equines. The fact that they don't work in humans, however, doesn't automatically mean they do work in equines but this is the argument being put forward.

"Humans and horses have different digestive systems and absorb substances differently," she said. "I like to have scientific evidence of efficacy. But people like to do something, and, as far as we know, glucosamine and chondroitin do no harm."

Other than wasting money, of course. As far as EoR knows, psychic healing also does no harm. Presumably, "leading vet" Sue Dyson would also recommend that. The continued support for a drug with no evidence for its efficacy is based on the claims of manufacturers:

Eliane Bennington of Healthspan, which owns VetVits, said: "We are not going to stop selling it — people who use it know it helps joints."

And personal anecdote:

And H&H forum users last week spoke out in support of the substances — the majority saying they found joint supplements effective. Vet Andrew McGonnell added that anecdotal evidence that the products work should not be "written off". "Many of my clients have seen benefits, and, crucially, have seen deterioration when treatment stops," he said.

Of course, if a Big Pharma company was selling a massively popular drug that had no evidence trials (indeed, trials indicating it was ineffective), EoR wonders whether the reaction would be similarly supporting.

No comments:

Post a Comment

Note: only a member of this blog may post a comment.