2
$\begingroup$

I am trying to replicate some odds ratio calculations which are reported in this paper. For example in the paper on page 3 in the Strome et al. data for tonsil: it states there are 52 cases and 48 controls. HPV % cases = 40% and controls = 6%.

Converting the percentages to raw figures, I get the following 2×2 table:

               Case  Control
HPV positive     21        3   
HPV negative     31       45

Therefore the odds ratio should $=\frac{21\cdot45}{31\cdot3} = 10.16$.

However the paper reports this particular odds ratio as 18.2 on the following page. Does anyone know why this is? Are they using a different formula?

|cite|improve this question
$\endgroup$
  • 3
    $\begingroup$ Actually if you look at figure 1, the Strome 2002 data shows odds ratio of: 10.2 (2.8, 37.0), so that 18.2 looks like a typo to me. Also in figure 2 they report the OR as 10.2 from Strome data. $\endgroup$ – bdeonovic Mar 9 '14 at 0:09
0
$\begingroup$

Answered in comments: Actually if you look at figure 1, the Strome 2002 data shows odds ratio of: 10.2 (2.8, 37.0), so that 18.2 looks like a typo to me. Also in figure 2 they report the OR as 10.2 from Strome data. – bdeonovic

|cite|improve this answer
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.