P value

p-values used

• a statistical test cannot absolutely prove anything - all a statistical test can do is quantify the likelihood that an observed result in a study is a real effect rather than due to chance
  • tests of significance (hypothesis tests) in clinical studies are undertaken to assess the probability that an observed difference between interventions could have occurred by chance - the tests actually check the hypothesis that no difference exists between interventions (referred to as a 'null hypothesis')
  • the p-value is the probability that no difference exists between interventions for a given endpoint ('null hypothesis')
    • probability can take any value between zero (no chance at all) and 1.0 (certainty), and this is also true the p-value
    • there is an arbitrary convention of using a p-value of 0.05
      • this means that if the p-value is < 0.05 (which means that the probability of the effects of two interventions being the same is 1 in 20 or less) the effects of two interventions are said to be statistically significantly different and the 'null hypothesis' is refuted (i.e. there is evidence that a difference exists between the interventions)
      • conversely, if the p-value is >0.05, this, by convention, would indicate there is no statistically significant difference in effect between the interventions
      • note that significance tests alone do not indicate the magnitude of the observed difference between treatments that is needed to determine the clinical significance of study results

Reference:

1. MeReC Briefing (2005);30:1-7.