|
|
|
|
front |1 |2 |3 |4 |5 |6 |7 |8 |9 |10 |11 |12 |13 |14 |15 |16 |17 |18 |review |
 |
The Odds Ratio and the Relative Risk are normally tested versus unity. It is not sufficient for the confidence interval not to cover One but we must test with a given power that it is greater than one. In the formulae appear p1 and p2. P2 is the prevalence of exposure in the population, which can be estimated easily. P1 the exposure in the cases, can be deducted from the OR ( or RR), so for a given OR (RR) we have a sample size for a given error alpha and beta. |