Chek the Channels
Reconciling CoVID-19 with Statistics-101
Today’s go-to conversation with friends, family, and co-workers is about CoVid-19. For people in the investment markets who are well-versed in charting, probabilities, and following trends, the conversation often focuses on statistics, or more specifically, risk analysis. But the current state of this disease makes it even more challenging than calculating the investment markets. With an investment, there is first the decision even to be exposed, and then the only possible outcomes are win, lose, or lost opportunity elsewhere. The coronavirus is different; we can’t just stay away like we can the stock market; it’s a global pandemic. We’re all “involved,” whether we feel we have minimum exposure or greater exposure, we are all somehow impacted, and there are no winners. So, calculating what we expect the results could be, is important to us.
Aside from wanting to figure how long the pandemic will last, the top question is, “How likely am I to get it” and “How likely is it that I will die?” We’re investors, not epidemiologists, so we try to get an answer by calculating a simple ratio. Most use bad outcomes divided by either the full population or the population that has been infected. Of course, we should all come up with precisely the same answer. Then, from this assessment, we can make decisions. Although we would prefer zero risk (no bad outcomes), we’re comfortable with the idea that everything in life has risk, and it is from the level of risk where we make decisions. The problem is, in calculating this simple ratio, we are running into trouble. Even the most analytical among us come up with numbers that vary from .25% death rates to 10%. The reason is that our simple ratio involves a numerator and a denominator. While we have fairly good sources of the numerator, the denominator measure is largely assumed.
Measurements in Public Health
The CDC, WHO, and local public health systems are good at counting incidence when monitoring outbreaks of disease, CoVID-19 is a good example of this. WHO reports on the status of the pandemic with detail by country of current cases, cumulative cases, and total deaths. This information is useful for measuring the stress on medical systems, impact on a region, and trends among regions. However, these data are all numerators in the risk assessment equation. It seems health organizations provide little to calculate the population at risk. The risk ratio becomes even more muddied when you consider one's point of view. People take the total deaths, which is a knowable value, and place that over patients hospitalized, or total tested, number exposed, the entire population, or an extrapolation of other assumed variants. All of these provide a different denominator and broadly different risk ratios.
The wide range of estimates, although not pinpoint, most often conclude that the ratio between the contraction of the disease and death is not high. For most individuals, whether the death rate is 0.5% or 2.5%, the change in our behavior will be the same; these percentages don’t seem that great. However, one of these numbers is five times larger than the other. This lack of imprecision may not alter our personal planning, but spread out over a population; it is the difference between 5,000 deaths and 25,000. For people in government, larger death rates are likely to lead to bigger decisions. Those decisions are where we’re impacted.
Unfortunately, as individuals and for the overall public, we don’t have reliable denominators. This can be helped by large scale testing, which is underway in many countries and regions. This presents other questions, such as the reliability of these tests and those implementing them. One step that will get us closer, if not now, in the future is a step the US Census Bureau just added online reporting to their 2020 effort. For now, the other numbers will still vary greatly. We know how we can lower our personal risk and those around us. Wash our hands and limit contact with others.