As I sit here at home for going on what seems like forever, and as I watch daily press briefings where he said/she said goes on for hours at a time, it occurs to me that we are not doing what needs to be done because we don’t know what we need to know. What we’re lacking is data. I know this sounds boring, but boring would be a refreshing change from what’s been happening. I would like to offer a recommendation for the future to assure that actions are proportionate to the magnitude of the health threat. It’s actually a simple recommendation: statistical process control, or SPC.
SPC has been used successfully in industry for over 100 years by managers and engineers to guide them in decision-making. The principle is simple:
All processes vary, and the variation must be analyzed using statistical methods.
Without statistical methods and good data we can’t know if the variation we are seeing is due to special causes, or whether it is normal for the process. And if decision makers don’t know that, they won’t know what to do. What policy makers need is a reliable estimate of the magnitude of the threat.
- Is this virus the worst thing since the black death? Or just another flu like illness?
- How deadly is it?
- How contagious is it?
- Does it affect all age groups equally?
- All races and ethnic groups?
- What is the breakdown between women and men?
- How are people with different medical conditions are affected?
- What is the geographic distribution?
- Patterns of spread? Speed of spread?
There are ways to answer all of these questions. Statistical methods exist to provide the answers. But the statistical methods must be applied to representative random samples. And that we do not have.
To understand a process you need data that are representative of the process. The process we are interested in is the contagious disease process. For the current pandemic of the novel coronavirus covid-19 we need random samples of the population of the entire world, a perfect job for a reputable version of the WHO. If that isn’t available a new organization should be formed with the mission of providing honest, reliable data. The strategy would be as follows:
- A panel of public health experts, statisticians and epidemiologists would develop a plan for obtaining representative random samples of the global population.
- Samples would be taken at frequent intervals, perhaps weekly.
- Sample data would be loaded into a database open to the public. The data would be subject to regularly scheduled audits, plus ongoing examination by members of the public. Adherence to sampling procedures would also be audited. The purpose is to assure absolute confidence in the raw data. Data is the currency of knowledge and without good data you have the classic garbage in, garbage out situation.
- Statistical analysis would be ongoing and dashboards created showing the results. The analysis would provide answers to the above questions. Dashboards would be public.
The emphasis would be on describing the baseline condition. What are the answers to the above questions during normal times? You can’t understand crises if you don’t know what things look like normally. On an average there are 56 million deaths that take place in a year. As of today (April 15, 2020) there have been 16,251,265 deaths year-to-date, 126,761 due to covid-19. Is this good? Bad? Normal? We simply don’t know. But we’ve been social distancing and our economies have been shut down for nearly a month at a cost of trillions of dollars. The collateral damage is immense and the potential damage is nothing short of a possible end to civilization as we know it. At the very least we risk the end to our privacy.
All of these are a consequence of our political leaders making decisions without good data. The time for this to stop is now.
Take a tour! In this interactive video you will discover what you will learn in our Lean Six Sigma Black Belt training and certification program.