Today we talk about ELISA tests, and of the logistical challenge that massive testing represents.
ELISA technology has been with us for decades. It is mostly used to detect antibodies, although it is also applied in agriculture and other commercial ventures. It requires specialized machines, which can be set to tell us if antibody is present, and how much. It can test for IgG and IgM, but separate plates and chemicals are needed for each one. It is more than 90% accurate, both in terms of specificity and sensitivity (are we picking up most of the cases; are we sure that these are Covid cases).
The problems are expense and the need for a blood draw (which involves a blood drawer and adequate protective equipment).
The reaction between a viral particle and an antibody is much like a three-dimensional jigsaw puzzle. Some pieces give you a tight fit; these are likely to give us accurate information. Sometimes the fit is not quite as tight. Think about that giant jigsaw puzzle that you get stuck on as the remaining pieces dwindle. Almost certainly you have placed a piece in the wrong place. Since we have hundreds of antibodies circulating, occasionally there will be one that looks like anti-Covid, but really isn’t. This is what we call a false positive.
All tests known to man have false positives associated with them. It can’t be helped; it’s like Newton’s principle that says that no machine can be 100% efficient. We try to minimize these, but as we will see, they are often a thorn in our sides.
The reaction between a viral particle and an antibody is much like a three-dimensional jigsaw puzzle.
Why can’t we test everybody?
If we are talking about testing for current virus infection, existing technology requires a trained technician with adequate PPE, plus thousands of machines geared to work 24/7 for many months. If we are looking for accurate evidence of past infection, we need many blood drawers and thousands of (different) machines running 24/7.
Then come the false positives. Suppose we have machines that are 95% accurate, which is considered excellent performance. That means that five of every hundred people tested will show positive, when they are disease-free, and 5% will be told that they are OK, when they are or have been infected.
Take HIV as an example. Imagine that we wanted to test everyone. We come to a cloistered convent where no one has had sex in decades, or blood transfusions ever. Out of 100 people tested, five will be positive with the best machines available. Consider the panic that we would bring to these five people, plus the expense of doing additional tests to rule out disease. Multiply those five by a million, and this is the situation we have when we indiscriminately test 100 million people.
Therefore, at least at first, we need to target the testing to people that are more likely to be infected, such as health care workers and everyone that deals with the public face to face. Then we will need a robust public health force that will span out to check contacts, and quarantine those who harbor live virus.
There is also a need for targeted random testing, meaning people picked by lottery in different neighborhoods and professions, also accompanied by good public health follow-up.
Ideally, as Bill Gates has said, we should develop the technology where we can self-swab our nose to detect active virus, and self-spit into a jar, to detect antibodies. This is nowhere near fruition, but if we partnered with every country, every university, and every biotech firm in the world, it is a very attainable goal. Unfortunately, the international leadership has not been there, although the level of cooperation between individual scientists of different nationalities for this crisis has been extraordinary. We just need the people at the top to step up with the money and technical infrastructure.
In the next day or two we will discuss vaccines and potential treatments.