Hi, is anyone able to help me?
I have a maths problem Population Y and with disease X with 2 tests (1 and 2) where the results can either be Positive, Negative or Inconclusive. This depends on the sensitivity, specificity and prevalence of the disease which I have calculated from the probability of testing positive (sensitivity) and negative (specificity) but I am struggling to get the maths to work for inconclusive results. I think I need two probabilities: one for the probability of testing inconclusive if they're actually positive, and one for the probability of testing inconclusive if they're actually negative. I've put below the probabilities for test positive and negative but can anyone help me with how to create a calculation for test inconclusive for both categories? I thought it could be 1-(prev2*Se1*Se2+(1-prev2)*(1-Sp1)*(1-Sp2)) and 1-(prev2*(1-Se1)*(1-Se2)+(1-prev2)*Sp1*Sp2) respectively but now I don't think that can be right.
prob positive in both tests = prev1*Se1*Se2+(1-prev1)*(1-Sp1)*(1-Sp2)
prob positive in one negative in the other = prev1*Se1*(1-Se2)+(1-prev1)*(1-Sp1)*Sp2
prob negative in one positive in the other = prev1*(1-Se1)*Se2+(1-prev1)*Sp1*(1-Sp2)
prob negative in both = prev1*(1-Se1)*(1-Se2)+(1-prev1)*Sp1*Sp2
I have a maths problem Population Y and with disease X with 2 tests (1 and 2) where the results can either be Positive, Negative or Inconclusive. This depends on the sensitivity, specificity and prevalence of the disease which I have calculated from the probability of testing positive (sensitivity) and negative (specificity) but I am struggling to get the maths to work for inconclusive results. I think I need two probabilities: one for the probability of testing inconclusive if they're actually positive, and one for the probability of testing inconclusive if they're actually negative. I've put below the probabilities for test positive and negative but can anyone help me with how to create a calculation for test inconclusive for both categories? I thought it could be 1-(prev2*Se1*Se2+(1-prev2)*(1-Sp1)*(1-Sp2)) and 1-(prev2*(1-Se1)*(1-Se2)+(1-prev2)*Sp1*Sp2) respectively but now I don't think that can be right.
prob positive in both tests = prev1*Se1*Se2+(1-prev1)*(1-Sp1)*(1-Sp2)
prob positive in one negative in the other = prev1*Se1*(1-Se2)+(1-prev1)*(1-Sp1)*Sp2
prob negative in one positive in the other = prev1*(1-Se1)*Se2+(1-prev1)*Sp1*(1-Sp2)
prob negative in both = prev1*(1-Se1)*(1-Se2)+(1-prev1)*Sp1*Sp2