Unhelpful Binaries

I wish a happy New Year and a healthy and less frustrating 2022 to everyone.

 

I had my second bout of Covid in the last week of 2021. The first time was at the start of the pandemic, and I only found out three months later when I took an antibody test. Then I could look back, recall a couple of days of a sore throat, and many possible occasions before the panic started when I could have picked it up from my heavily infested Queens neighbourhood.

 

This time had similarities. This new variant seems to everywhere just now, especially in heavily populated New York. We are smarter about hand washing, distancing and masking now, but the virus still got through despite my three vaccinations, probably at a family gathering. It is still not easy to get a PCR test, but I needed one for my church choir and managed to secure an appointment last Wednesday at a city run site, which duly came back positive, just as mild cold-like symptoms started to emerge. My wife and I had both taken antigen tests earlier in the week that came back negative.

 

My main takeaway from all this is about a series of unhelpful binaries. Whenever we take a binary, yes or no, approach to anything about Covid, we seem to make silly mistakes.

 

The first binary is about whether we have Covid or not. My suspicion is that most of us in densely populated areas where Omicron has started to spread have at least some Covid. But we are not encouraged to think that way; instead we divide the population into those that have it and those that do not. Especially at the beginning, when we thought Covid was almost a death sentence, this binary approach unnecessarily added to our panic. We wanted to call a priest for the last rites if we so much as touched a doorknob.

 

True, some communities did not have any Covid, so it was safe to do more or less anything there. Even now, in places with aggressive policies like China and Australia, Covid can be isolated and eliminated quickly. But most places are not like that. There is some Covid around and probably most of us have a bit of it some of the time. With Omicron the chances have gone up. Our challenge is not to keep it out completely, but to take smart steps to keep the level low enough to restrict the danger, and to further improve our odds by shoring up our immune defences via vaccinations and good hygiene.

 

This logic extends to tests. We tend to view test results as binaries. Indeed most tests come back with a binary positive or negative result. That is not as helpful as it could be. PCR tests are much more sensitive than antigen tests. The likelihood of false positives or false negatives is quite high with antigen tests. PCR tests measure not whether the subject has a current infection but a current or recent infection, so will give positive results for several weeks after somebody is clear of Covid. Even the most sensitive and well-executed tests have to set a cut off level to yield a positive result, and results may cluster around that level, so many positive tests will indicate only slightly higher levels than many negative ones. In practise test execution is mixed, especially when administered at home. All this also requires people to be able to actually find tests and to get results within a time interval where they are of any use.

 

Then there is the reliability of statistics to consider. Different cities and countries use different tests and test different subsets of the population. New York publishes the percentage of PCR tests administered at City sites that come back positive. This indicator is currently about 20%. While a frighteningly high figure, it does not mean that one in five New Yorkers are infected with Covid because the sample is very obviously a biased one. I have been saddened to see statistics used selectively and manipulated shamelessly to promote loaded messages through the pandemic. Some of this is nefarious, but official sources do not help their cause by using similar tactics.

 

The next binary is about safe practices. Of course no set of practises are completely safe, and no practise can guarantee you get Covid. It is all about odds, and should be explained as such. Instead we are offered precise guidelines, and many interpret these as a binary indicator of safety or danger. Last May I attended a Mets game masked and alone in a vast outdoor stand only after following intense protocols to gain admission. A week later I attended another game without any admission protocols and in a packed stand surrounded by unmasked people singing and baying loudly. It would be far better if the guidelines could have more nuance. It still feels very wrong to see vast loud and angry crowds packed in at football games. This is big money wielding an undue influence over public policy and rather an indictment of the intelligence of the attendees, but the binary communication of guidelines does not help. The same big money leads to reduced restrictions restaurants and bars, while I know from personal experience that alcohol is disastrous for self-discipline.

 

The next unhelpful binaries concern general statements about immunity and the pandemic overall. Herd immunity was an unfortunate term introduced in 2020, and still pervades conversation. There is no such thing as herd immunity; instead there is a level where susceptibility becomes less than voraciousness overall so the number of new cases diminishes. But people love to justify reckless behaviour by claiming that herd immunity has been reached. In the same way, no vaccine, booster or prior infection provides complete immunity; they simply improve our odds for a time. Similar logic can easily conclude that most binary claims about the end of the pandemic are rather meaningless. The pandemic will evolve and eventually diminish; it will not end.

A last common binary is about long Covid. We are led to believe that this is a specific additional ailment that afflicts a subset of sufferers. It feels far more likely that a spectrum is a better description, with Covid handicapping most of us to some extent long after we contract it. This weeks analysis of professional footballers whose average performance declined for several months after Covid bears this out. I know myself that walking up stairs had me out of breath for several months in 2020, and that there was more to it than wearing a mask. Yet I can hardly describe that as long Covid. 

 

I struggle to understand the reluctance to use the concept of probability when defining and communicating public policy for Covid. Most people gamble and indeed our screens are currently filled with inducements to gamble recklessly. So authorities must trust us to understand at least something about how odds work. Yet they cannot seem to trust us when it comes to health guidelines.

 

Occasionally something more helpful surfaces amid the cacophony. I still like the image of Covid as cigarette smoke, with contagious people equated to smokers. Being close to many heavy smokers in an unventilated space will very likely lead to some smoke entering my lungs, especially if the smokers are breathing all over me. But a few light smokers on the other side of a park won’t do me much harm, and any harm will vanish quickly unless my lungs are in really poor condition already. This feels an excellent image to understand Covid. Yet most of the time I am served up instead with platitudes are vaccination immunity and six feet being a universally safe distance. We can do better.

 

Covid is only an example of where binary thinking inhibits smart responses. For centuries sexuality was viewed in the same way, with predictably harmful results. What about foreigners? Communists? Christians? In practise many of us might share some features of all these categories but feel forced into binary positions.

 

The same simplistic guff applies to most discussion of criminality, dominated by a narrative of catching and neutralising the bad guys. Are most of us easily described as good or bad? If we are bad at a certain moment, why might that be? What is the best way to make us less bad more often? It is not difficult to find effective answers to these questions, but the lazy binary approach makes it much harder.