Hacker News Reader

Comment by nl | Hacker News Reader

This seems more likely to be the real reason: https://www.advisory.com/daily-briefing/2020/09/01/covid-tes...

TL;DR: Some (many?) PCR tests are using too high a cycle threshold, which makes them too sensitive. Key quote:

Juliet Morrison, a virologist at the University of California-Riverside, said she believes any test with a cycle threshold over 35 is too sensitive. "I'm shocked that people would think that 40 could represent a positive," she said.

The following is my commentary, and not specifically in the article above:

There were many early tests that were giving false negatives. It looks to me like test manufactures over-compensated to make sure they never give false negatives. There is no data source that I can find that shows what test is used in what area, or what CT is used for what test.

View on HN

This article is basically bullshit derived from a NY Times article, and I'm really sad that it got published. You know how every once in a while the media publishes an article about something you know and it's riddled with errors that drive a particular narrative? That article is my experience.

Research scientists and clinical scientists are worlds apart, and have very different goals in life.

I wouldn't trust a cT coming out of any research lab either, but clinical labs are far far more controlled and have far better processes, and have intense validation that shows that their threshold is repeatable, accurate, and a true representation of their limit of detection.

No research lab does any of this, and they almost never have extremely basic contamination protections like separate rooms for pre- and post-amplification rooms.

I've watched research scientists try to stand up clinical labs during this pandemic, and it has been an education for them to bring their standards up to what is the baseline for clinical testing. Its not unlike giving a butcher a scalpel and expecting brain surgery: there are a lot of the same skills involved but also very different and extremely narrow it's of knowledge.

What does any of that have to do with cT as it relates to accurate COVID-19 detection as discussed in the article?

Is this the NYT article you're talking about?: https://www.nytimes.com/2007/01/22/health/22whoop.html

Can you explain to the layperson why it's "basically bullshit"? The NYT article I linked describes a real-world scenario where a fake outbreak (100% of cases were false positives) actually occurred (note: I am not claiming Covid-19 is fake).

Also, specific false positive rates aside, is it not likely we're now approaching a rate of "cases" which is hard to distinguish from false positive tests, which would help explain the drop in hospitalization / ICU / mortality?

Keep in mind that article is from 13 years ago.

> Many of the new molecular tests are quick but technically demanding, and each laboratory may do them in its own way. These tests, called “home brews,” are not commercially available, and there are no good estimates of their error rates.

The technology, as mentioned, was new at the time and not even commercially available. There have been 13 years of improvement since then.

There's an NYT article from this on the Covid test issues.

Reply on news.ycombinator.com