In my study each person either tests positive or negative - in state you can look at the proportion testing positive by the usually tab function. And we have looked at this by various sociodemographic factors. However, the diagnostic test used in the study had imperfect sensitivity and specificity. The formula for the adjustment is: p = (q + specificity – 1) / (sensitivity + specificity – 1)
where p = adjusted proportion positive and q = observed proportion positive
Is there a way of carrying out this adjustment automatically / efficiently in STATA which doesn't involving manually calculating the adjusted proportion after running the tab function. Perhaps by generating another variable for each individual which in some way uses this function?
Help much appreciated - thanks
Is there a way of carrying out this adjustment automatically / efficiently in STATA which doesn't involving manually calculating the adjusted proportion after running the tab function. Perhaps by generating another variable for each individual which in some way uses this function?
Help much appreciated - thanks
0 Response to How to adjust proportions for imperfect test performance
Post a Comment