Screening Settings VS. Diagnostic Settings

Click here for Audio/Video Version  *requires RealPlayer - free download

Dr. Kees Jansen: “At present we are at the European Society meeting in Lausanne in 2001. We have just heard a very interesting lecture by Dr. Hajo Wildschut who is sitting next to me, and my name is Kees Jansen. This lecture was such that I would like to interview him on some of the important messages he had to convey. Now, Dr. Wildschut, could you tell me something about, for instance, the difference between prenatal diagnosis and screening?”

Dr. Hajo Wildschut: “Yes, thank you very much. Indeed, I think with the influx of all new medical technologies there is a tendency to use diagnostic tests in screening settings, and I try to make a point that you should carefully interpret your test results in a screening setting because they’re not necessarily the same as in a diagnostic setting. This sounds fairly complicated but to cut it short, when you have a positive test result in a screening setting that doesn’t mean that you have the disease since the likelihood that in a screening setting the woman has the disease is very low, and I’m talking about, for example, infectious diseases like chlamydia. You should really do a diagnostic evaluation of positive screening tests - an office positive test results - in a screening setting to rule out or to confirm disease. That’s very different from when a patient comes in to your office with symptoms or signs of, for instance, infectious disease because when the test is positive for chlamydia then the likelihood that you will indeed have chlamydia is very high. So one of my thesis during this lecture was that you really have to know the pre-test probability of the disease prior to testing - so are you looking at someone who’s coming from a screening setting or someone from a diagnostic setting like your own clinic.”

Dr. Kees Jansen: “I think that’s a very important message because the outcome is determined by the patient population that you’re looking at. Now one of the other things that struck me in your lecture is that, let’s say, the sensitivity and specificity of the test, the value of it, and the number of false-positive and false-negatives is very much dependent on the incidence of the disease itself in the population you have. Would you comment on that for me?”

Dr. Hajo Wildschut: “That’s very right, you have a fairly good test and using that in a diagnostic setting then, indeed, the likelihood that you have the disease is high. However, using the same test with the same sensitivity and the same specificity in a screening setting then the likelihood that you have the disease is low. I think many people don’t realize that and that’s the point I try to make assuming that the sensitivity and the specificity of these tests are similar in each setting. However, quite often in screening settings the sensitivity and specificity are even worse so you should be very, very careful with the interpretation of test results from screening settings. People might say - why don’t you use a better test that tests up to 99% sensitivity and 99% specificity and those tests are indeed available in diagnostic settings, however, these tests are quite expensive and when you go to asymptomatic women you have a very large population because most of the population doesn’t have any symptoms. The disease prevalence and the ones you see is a smaller number so what happens if you use those expensive tests for screening purposes then the costs per case detected goes up substantially, and I think that’s also a very important message.”

Dr. Kees Jansen: “So basically it all boils down to the amount of money we as a society would be willing to spend on a test and the other thing, of course, is the percentage of unrest that you may create if you cut costs. If you have a cheaper test, a number of patients will be alarmed by a false-positive test, which is very much independent on the sensitivity of the tests as such. Are there some other remarks you wanted to talk about?”

Dr. Hajo Wildschut: “I think you’re right. I didn’t talk about that during my lecture, but you have to follow the WHO criteria of Wilson and Jungner. One of the first points they make in order to provide money for tests like this is the importance of the health problem, and when is a health problem important? I think that depends on two things. One is it’s important when you have a severe disease which affects the health of the particular person and, secondly, it’s important when it’s highly prevalent. We often talk here at these settings about rare diseases so, indeed, the Wilson and Jungner criteria for importance are fairly subjective. I think we have to discuss in the community and assign money to those health outcomes which we consider as important not only in terms of severity of the disease outcome but also in terms of the likelihood of the disease occurring.”

Dr. Kees Jansen: “Thank you very much.”

Dr. Hajo Wildschut: “You’re welcome.”

Dr. Wildschut's presentation is available for reading.