Antibody Testing, Some Basics
This post is to highlight and direct you to a superb article by Caroline Chen, a health and science writer for ProPublica. It is entitled What Antibody Studies Can Tell You — and More Importantly, What They Can’t. I encourage you to click that link and read the article itself. This is some of the clearest writing on a complex topic that I've seen. To whet your appetite, here are the sub-headings of article:
Antibody studies can be used to answer more questions than you might think.
Setting up a sero-survey correctly means you need to test a random population — easier said than done.
Test accuracy can skew results in some pretty surprising ways.
Forget the headlines, your city is nowhere near herd immunity.
There are two types of death rates. Most people are mixing them up.
Stop comparing this to the flu. Without a coronavirus vaccine, we are far more vulnerable.
Antibody tests aren’t ready to be used to issue “immunity passports.”
Caroline Chen deserves a prize for the clarity and readability of her explanations of complex, nuanced material. You would not be wasting your time if you read it carefully--several times. A clear understanding of this material will equip you to cut through a whole lot of muddled, and sometimes needlessly sensationalist reporting and headlines.
When you're done absorbing the first article I encourage you to carefully read an earlier work of Ms. Chen's from April 2: What We Need to Understand About Asymptomatic Carriers if We’re Going to Beat Coronavirus. In that article she clearly addresses how scientific data is sifted and formed into public health recommendations.
Keep to the high ground,
P.S. A few musings on the topic of science and statistics:
People (and stock markets) crave certainty. People often rally around a leader who offers a certain future, even if, for example, the predicted future is a fixed date and description for the end of the world. In contrast, science might start with something that's pretty certain, "Covid-19 kills some people" and then strive, by asking questions and analyzing data, to come up with the most reliable prediction of what's going to happen. Science is messy. Watching the process science uses to seek truth can be confusing--but is a far better guide than a would-be leader blustering inanely about "bouncing back," a "V-shaped recovery," or touting unproven or dangerous treatments based on the last thing random thing he heard.
Scientists, and effective governments. run on numbers: gathering numbers, understanding numbers, analyzing numbers. (The Founders knew this. That's why we have a Constitutionally mandated decadal census.) In this pandemic well-meaning (but sometimes not-so-bright) media folk bombard us with a bewildering array of statistics and ill-defined terms, e.g. case fatality ratios, infection fatality ratios, antibody prevalence, and even just the simple term "cases," often without clarifying what the number means and how it was determined.
Questioning the importance and accuracy of any number is essential. Does the number actually represent a cross section of the population we're trying to draw conclusions about, i.e. was enough data gathered or did the method of gathering of the sample bias the result? Questioning the data (formally that's called "peer review")--and then improving that data with more gathering and analysis--is an important way science works. Absolute certainty is always a little out of reach, but each step adds to what we know and the clarity with which we can look into the future.
Read Caroline Chen's articles. Arm yourself we a clearer understanding of the basics. Read the news with greater clarity.