5 ms·
‘How Could the CDC Make That Mistake?’
- timr 6y agoThis is a misleading article. While it's true that an antibody test is capable of detecting prior infections and a PCR test is not, this has little bearing on the overall specificity/sensitivity profile of any particular test. The main argument here appears to be that antibody tests have false negatives, therefore antibody tests will make test numbers look artificially lower. But there are antibody tests with very high specificity and sensitivity (relative to PCR testing), and there are less-reliable PCR tests. Moreover, we know there are error bars on these positive rates and case counts, and other sources of error (e.g. sampling bias) likely play a much larger role in uncertainty than false negatives in antibody tests. If a clinic is using an IgG/IgM test (particularly one with a good sensitivity/specificity profile), there's no reason not to include the results with PCR data in aggregate stats. South Korea expanded testing primarily through widespread antibody testing, so we're hardly alone here.
- redis_mlc 6y agoThe real question is why states care about inaccurate corona test results in the first place. Oh, silly me. Some governors plan a Whitehouse run. "What good are the test results if you can't trust them?" https://www.cnn.com/2020/05/20/health/testing-coronavirus-cidrap-report/index.html https://www.cnn.com/2020/05/20/health/testing-coronavirus-ci...