A Positive Test Means You Have the Virus, or You Don't

“But I don’t want to go among mad people,” Alice remarked.
“Oh, you can’t help that,” said the Cat: “We’re all mad here. I’m mad. You’re mad.”
“How do you know I’m mad?” said Alice.
“You must be,” said the Cat, “or you wouldn’t have come here.”

― Lewis Carroll, Alice in Wonderland

Imagine your refrigerator breaks down, and you go to a respected appliance store to find a replacement. You see a new model of a fridge you have had your eyes on and ask the salesperson what he thinks of this particular model. “Yes, it’s the newer model,” he says. “It is possible it can keep the food cold, but it’s also possible it won’t keep the food cold.”

At that point, you might not only question whether you want to buy that fridge, but also how you got dropped into the middle of an Alice in Wonderland performance.

This past Sunday in our local newspaper, the San Francisco Chronicle, the lead story was titled “Antibody Test Hopes, Doubts.” Here are two quotes from the article. The first is a “disclaimer” from the FDA required on all the new Coronavirus antibody tests: “Negative tests do not rule out SARS-CoV-2 and positive results may be false.” In other words, the fridge might work or it might not.

Also quoted was Marin County’s public health officer. He said, “I tell them you will likely be negative” if tested. “And it either means you were not infected or you were and it’s wrong.” So, the test either means you were infected or you weren’t infected, which I guess covers all the bases.

I want to be clear about this point: This confusion is not a matter of “we need to improve the tests” or “we need to understand the immune system, virus, mutations, etc., better.” This circumstance is a failure of the very concept of surrogate testing, which is what’s being used with Covid-19. You simply can’t use a surrogate test EVER for diagnosis unless you have a definitive (called “gold standard”) test to compare it to.

I have addressed this topic many times during this crisis, but I’ll try again to be as clear as I can about the requirements for an accurate surrogate test. Take the example of the testing to diagnose a pulmonary embolism, otherwise known as a blood clot in an artery to the lung. In a patient with chest pain and suspicion of a clot in one of the arteries in a lung, one can do an angiogram, in which dye is injected into the artery. If a clot is present, the dye will show the presence of this clot 100 percent of the time. This test becomes the gold standard, meaning, you can use it to detect the presence or absence of a clot with 100 percent certainty. You can then test a surrogate examination called a V/Q scan against the angiogram results. The V/Q scan looks at the perfusion or flow of the blood into the lung (this is a somewhat simplistic explanation but it demonstrates the point). A clot will affect the flow of blood.

Here’s how it works. If you take100 patients who have a blood clot present on angiogram, you can determine the number of patients who have a positive V/Q scan. If 99 out of the 100 proven patients with a clot have a positive V/Q scan, then you know with certainty that the false negative rate is 1 percent. At that point, the doctor can forgo the more expensive and dangerous angiogram and use the V/Q scan, knowing it will pick up 99 percent of the positives.

The next step is to do the V/Q scan on 100 patients with chest pain whose angiogram is normal, meaning, you are certain they have no blood clots. If the positive rate is 2 out of 100, then you know with certainty that the false positive rate is 2 percent. Again, this allows you from then on to forgo the angiogram and do only the V/Q scan on people with characteristic signs and symptoms of a pulmonary embolism because you know for certain the error rate of the test.

With Covid-19 testing, the gold-standard test has never been done. Therefore, it’s impossible to know how accurate the test is. The only conclusion that any rational person can draw is that the entire testing for Covid-19 is completely unstandardized and therefore meaningless. What happens in a situation like this is you get reports of wildly divergent false-positive and false-negative rates.

In addition, I’m frequently asked about antibody tests, as in, “doesn’t having antibodies mean you are immune to the Coronavirus and therefore must have had an infection at some point?” According to the Marin public health director, it turns out we have no clue what an antibody test means. Years ago, we were taught antibodies meant you were immune and therefore safe. Then they came out with HIV tests and found antibodies and told us that meant you had the disease. Then they said that sometimes having antibodies meant you were immune, sometimes it meant you had the disease, and sometimes it meant nothing. Antibody tests are another surrogate test that means nothing and can never be improved to mean something. The whole thing is a house of cards.

To finish this point, I was interested to read recently that the president of Tanzania was skeptical of the testing. He decided to test a goat, a sheep and a paw paw fruit to see what happened. He gave the samples names and dates of birth and sent them in to be analyzed. The goat and the piece of fruit tested positive. I hope this doesn’t trigger the FDA to ban goat-milk products and paw paws. But now that we live in Alice’s world, anything is possible

Stay well and keep questioning.

“Millicent Sowerby illustration” by sofi01 is licensed under CC BY-NC 2.0