Hacker News Reader

This is incorrect. The sequence test has low false positive rate and higher false negative rate. The positive results alone are significant and informative from a policy perspective. Combined with other measurements like CT, the accuracy can be high.

View on HN

CT for a virus? You can only see that something is inflamed. I mean you can guess the diagnosis, but that doesn't really increases my confidence in testing. If inflammation can be detected, you probably also have symptoms already. That aside, you should limit CTs if possible. And you won't be able to do mass scans anyway. So it is just for specific cases. MRTs could also help and are not as invasive, but maybe they don't show enough.

This Chinese paper compares the sensitivity of diagnosis with chest CT vs RT-PCR:

"In a series of 51 patients with chest CT and RT-PCR assay performed within 3 days, the sensitivity of CT for COVID-19 infection was 98% compared to RT-PCR sensitivity of 71% (p<.001)."

https://pubs.rsna.org/doi/10.1148/radiol.2020200432

Good to know about the positive rate. The media has not been very forthcoming with actual data an usually only quotes wide numbers. What I meant is that is test alone will not do the job perfectly and that is what people expect. Higher false negative rates is what bothers me as people generally think "oh I'm safe now" when that might not be the case. Considering they wait here with testing to after symptoms it would likely make the tests way more useful then to test everyone. It requires labs and those might also be stressed to a point of breaking if you start testing everyone (if that is even possible) for instance.

Reply on news.ycombinator.com