Disproportionality in Pharmacovigelence

#1
I am working on a project involving pharmacovigilance that was passed on to me by another individual. I am a little skeptical about the values used to calculate the proportional reporting ratio, so I would like your opinion on the method. First, the data is a matrix of m rows of adverse events (AE) AE1.......AEn by n columns of drugs drug1....drugn.
PRR = (Drug&AE/ (Drug&AE)+(Drug&noAE)) / (noDrug&AE/(noDrug&AE)+(noDrug&noAE)). The values, however, for the noDrug and no AE are being assumed from either the other columns or the other rows. meaning if we are calculating a PRR for drug1 the values for noDrug&AE are the row sums of drug2 through drugn in the matrix for the AE row being evaluated. Similarly, the values for Drug&noAE1 are the column sums of AE2 through AEn. I guess my biggest issue would be that the values for noDrug&AE are the row sums of the AE being evaluated and all the other drugs in the matrix except for the drug being evaluated.

Any insight would be greatly appreciated.
Thank you.