I have done a post of meta-analysis using the Proc Freq to produce cross tabulation of 2 by 2 table, the odds ratio, and the relative risk (column 1, column 2). In this post, I will show how the stats for odds ratio and relative risks are calculated using Excel.
A. SAS output
- Frequency table
- Odds table
B. Excel
- Frequency and Odds Ratio
- Relative Risk Column 1
- Relative Risk Column 2
C. Interpretation
Drug: peole who feel better/people who did not improve = 58/30 = 1.93333
Placebo: people wo feel better/people who did not improve = 24/3 = 8
Odds ratio: of 0.241667 means patients who take the drug are 0.24 times likely to improve.
The probability of the improvement with the drug is 0.74 time the probability of the improvement with the placebo.
The probability of the no improvement in the symptoms with the drug is 3.7 time the probability of no improvement with the placebo.