I have 115 cases in breast cancer group of which 8 are HPV positive. In control group consisting of 31 patients there was no case of HPV postivity. What is the OR risk value?
Can't be done. I suggest you select some more referents by the same sampling frame until you get at least one "exposed (ie HPV +) referent.
If this is impossible ,a bit of a cheat would be to change the criteria of HPV positivity to admit a borderline referent as positive . This could have implications on the case HPV classifications.
Yes, the number of positive cases in control group is zero, but is neccesary always numbers in this control group.
One method accepted generally by investigators and computer programs is to add 0.5 to to each cell and calculate.
(https://books.google.co.in/books?id=ZLDWAgAAQBAJ&printsec=frontcover#v=onepage&q&f=false) From the book Intuitive Biostatistics, page 249. Got it by searching for "How to count the OR and RR if I there is zero value in one cell"
This is NOT cheating. You can do this if you can't increase the numbers available.
Your finding could be important. Were they age matched? There may be other factors too, remember
Narayanan