The interpretation of evidence

doi:10.1136/bmj.39465.493090.3A

BMJ 2008;336:174 (26 January), doi:10.1136/bmj.39465.493090.3A

Letters

The treatment paradox

The interpretation of evidence

The first 150 words of the full text of this article appear below.

Jenkinson and I have an article on a similar theme to Spence’spiece in this month’s Student BMJ.¹ ² Any trial generatesfour summary numbers: relative risk reduction, absolute riskreduction, number needed to treat, and personal probabilityof benefit.

Each number is useful and gives some information, but no onenumber gives us the whole truth about the information. Usingone figure on its own, particularly the relative risk reductionabove all others, is very risky.

Each figure takes a different viewpoint on the evidence. Therelative risk reduction is a public health (area wide) prediction.

The absolute risk reduction puts the starting risk back intothe frame.

The number needed to treat measures the workload needed to achievethe relative risk reduction. It’s the beginning of healtheconomics.

The personal probability of benefit answers the patient’squestion, "What’s in this for me?"

All the figures are contained in . . . [Full text of this article]

Peter G Davies, GP principal

¹ Keighley Road Surgery, Halifax HX2 9LL