Choose here
Information in this site comes from
a review of the evidence 2006. A new review is expected in 2011- we will update it when this is out.
What the ratings mean
  • Very likely to help

    Scientists are very sure about this way of helping.

    Why this rating?

    Several different scientists have done the most careful kind of study where they split people into different groups, each of which got a different type of help. The groups people were put in were decided 'randomly', e.g. someone might toss a coin to see who goes in which group - so that each group has roughly the same mix of people. This is thought to be the fairest test of whether something helps because people in the groups are similair in every way, except for the type of help they get.

  • Likely to help

    Scientists are pretty sure about this way of helping.

    Why this rating?

    Scientists may have done several good studies where they have split people up into different groups and given them different sorts of help. People got to choose – or were chosen – to be in each group. This is not as fair a test, because people in the different groups may be different in some way that affects whether they get better or not.

  • Might help

    Scientists are not sure about this way of helping.Why this rating?
    Scientists have not yet done many good studies, or they have only looked at what people are already doing and compared one group with another. This is thought to be the least fair test because it is quite likely that the people in the different groups are different in some way that affects whether they get better or not.



    For more information about how scientists decide what helps: Testing Treatments: Better Research for Better Healthcare Imogen Evans, Hazel Thornton and Iain Chalmers (2006, British Library)

  • This site rates what scientists have found helps, based on how sure we can be about what they have found out.
  • All the tests by scientists reported here have been checked to see how much we can trust the findings.
  • Ways of checking scientific findings include looking at: how scientists have tested something; who paid for the research; whether other scientists have found that the same thing help different people in different places.
  • How well you get on with the person trying to help you is likely to be important – let them know what you find helpful and unhelpful.
  • There are things for and against all types of help and they all mean you doing something new – ask about what is involved.
  • Knowledge is power