Covid-19: The false positive problem – using tests that specifically err on the side of producing more false positives

NICOLA Sturgeon said it was better to “err on the side of caution” on Covid testing even if it means generating more false positives.

Source: Sturgeon says she would rather ‘catch as many Covid cases as possible’ – even if it means more false positives | HeraldScotland

Note below “Whenever you do screening tests, you actually want a lot of false positives because you want to capture everybody with the disease”. For that reason, they recommend that those testing positive with a rapid test then be tested with a PCR test. Except that is not happening in my state.

Below is a description of the test my state is rolling out en masse – without the follow up PCR testing. In fact, the CDC says not to do subsequent tests – and if you do and you test negative after a positive test, the CDC (and my state) say you still must quarantine for 11 days. We are using a test that generates false positives and insisting that the result be the definitive result.

How many false positives? According to this, up to 5%. 95% sounds reliable doesn’t it? In my county, the actual prevalence of the disease has been estimated at 1 in 1000 people. Test all 1,000 people and you will likely find 1 actual case, plus 5% or about 50 false positive results. We have no way of identifying which of those 51 cases is the real one – so all 51 will go into quarantine. Our official case count will go up by 51 cases, even though statistically, 50 of those are false positives. In fact, 98% of those cases are false positives, not the 5% you were probably expecting. This is a well known issue for screening mass populations for low incidence diseases. I have repeatedly asked OHA, CDC and OHSU how they are dealing with this – and all I get is silence.

UPDATE: Nevada has stopped using certain rapid tests saying 60% of those testing positive then tested negative using a PCR test. Depending on the prevalence in the Nevada population, that results sounds plausible with a 98%-99% accurate test.

Not surprisingly, our state’s “test” numbers are poised to explode, which will be followed by new restrictions on the population. I expect announcements of new rules within the next 2-3 weeks. 

This – right here – is the test my state will be using to double and triple weekly testing:


Another oddity – my state suggests everyone get tested if you have any cold symptoms, even just a common runny nose, or allergieis, or if you may have been in contact with a later confirmed case. The price charged for the test is $60 to $300 each (actual cost from manufacturer is about $5?). For some people, the test is free, paid for by someone else.

The result of the test will decide if you are forced into quarantine or not.

For almost everyone, the test result will have no effect on treatment or outcome.

In other words, you are paying for a test to provide statistical data to public health – and rarely will it be used to improve your health. Due to the false positive problem, many will be unnecessarily forced out of work for an extended period.

Children tend to get common colds about 6 to 12 times per year. The cost of testing each time could range from $360 to $3,600 per year. If you have 3 kids, do the math on your annual expenses.

What am I missing? I have repeatedly asked public health officials about this but they do not answer.

Update: On another related post I added an update – other tests in the BinaxNOW series have 3% to 6% false positive rates. Due to how the statistics work out this can lead to a majority of “positive tests” being false positive results.

Remember, I am an idiot. I need to disclaim that I have no training in health care and anything I write about health is FOR ENTERTAINMENT ONLY.