This article isn't very good IMHO because it focuses on emotional considerations rather than the underlying math. The core problems with the PSA test is that it has relatively low statistical specificity and it hasn't been shown to have any statistically significant effect on mortality in aggregate, and the resulting treatment options do have statistically significant morbidity. Hence all the debate about when, how often, and for whom. All the arguments about increased anxiety are pretty unhelpful and obscure the underlying statistical challenges.