I think you mentioned that your partner and family work as scientists in relationship to the pandemic? I know you have posted helpful data, so I'll pose this question to you or whoever else might have input.
What do you make of the numbers of those tested or died of COVID-19? I saw a
research that said the numbers confirmed may be 6% of total infections. Can we tell if numbers are going up or down? They say they are going down in my state because the percentage of positive cases within all tested are going down. That would be hopeful in light of all these gatherings, but I feel uneasy about the numbers. Are all demographics getting tested? Are people who refuse to believe in the virus unlikely to go for testing unless they ne ced a ventilator? Is the number of people dying from COVID-19 accurate? There is
another article that says the overall numbers of deaths are up and some of those could actually be COVID-19.
I can see how easily those numbers could be adjusted based on contextual factors, so there is a way I feel like I can't tell what is actually going on with numbers infected. What is your take on that?
I'm not an expert, so I'm not comfortable providing any definitive guidance. I can provide you with my thoughts as a statistically literate professional. Also, I'll be happy to forward along any non-privileged information shared with me by my family. My fiancée is a scientist studying the prevalence of COVID-19 in the San Francisco Bay Area (see below), so that might prove interesting.
Stanford teams with UCSF, Chan Zuckerberg Biohub to study prevalence of COVID-19 in San Francisco Bay Area | Stanford News
What do you make of the numbers of those tested or died of COVID-19?
Both the number of confirmed COVID-19 cases and number of confirmed COVID-19 deaths are certainly lower than the the actual number of infections and deaths. The former is self-evident. The latter can be substantiated by comparing the absolute number of deaths year-to-date in 2020 with prior-year numbers (deaths with incidental COVID-19 infection would not skew this comparison).
The interesting metrics derived from these numbers, e.g. prevalence and infection-fatality-rate, are going to be preliminary estimates at best. But even with preliminary estimates one can derive a range of uncertainty around them which will tell us
something. I would encourage you to find a few primary sources and periodically follow up with them going forward. Forget media outlets and secondary reports.
I saw a research that said the numbers confirmed may be 6% of total infections.
Yes, as of the date of the underlying analysis (March 17th) it was reasonable to assume that a large number of infections were not yet identified. Testing regimes were in their infancy in most parts of the world. But that 6% figure would not be applicable in the United States. It was a relatively simple estimate that presupposed the underlying fatality rate (from Wuhan data, 1.38%) and applied it to the number of actual observed deaths in various regions. Using that same logic today, by dividing the current US death count of 120K by the assumed fatality rate of 1.38% we would estimate the actual number of cumulative infections to be 8.7M. Using that study's average time from onset to death of ~18 days, we can look to the total number of confirmed cases from 18 days ago to derive the detection rate: 1.8M / 8.7M ~ 21%. This would suggest that the official number of cases reported in the United States represents only 21% of the total number of infection, i.e. 4 in 5 infections go undetected.
Ultimately, this illustrates the relationship between the prevalence and the underlying infection-fatality rate, but does not provide us with any greater certainty as to the actual values of either number.
Can we tell if numbers are going up or down?
We can discern the trends in infections before the last week by looking at changes in the positive test rate, assuming that the number of tests performed and relevant protocols remain unchanged. We can more accurately discern the same trends from several weeks ago by looking at changes in the number of reported deaths. However, both are lagging indicators. Again, I would encourage you to identify a few trustworthy
primary sources and stick with them. Many statistically literate
and well-informed professionals are providing analyses of the data.
They say they are going down in my state because the percentage of positive cases within all tested are going down. That would be hopeful in light of all these gatherings, but I feel uneasy about the numbers.
I would look toward long-term public health officials (not political appointees) in your state to assess your risk.
Are all demographics getting tested? Are people who refuse to believe in the virus unlikely to go for testing unless they need a ventilator?
I don't know the answer to this. I was recently tested via nasal swab (wouldn't recommend) because I had to undergo a medical procedure and it was mandatory. There are certain activities that require testing (medical procedures, certain travel, etc.) and so anybody wishing to engage in those activities would have to consent to being tested. I can also personally attest to several acquaintances experiencing symptoms who did not get tested or who were refused testing, though in each case those people self-quarantined.
Is the number of people dying from COVID-19 accurate? There is another article that says the overall numbers of deaths are up and some of those could actually be COVID-19
It depends on what you mean by accurate. As I said previously and as your linked article suggests, the number of confirmed COVID-19 deaths is likely lower than the actual number. But the conclusions you or I would generally draw from the available information shouldn't substantially change. What is beyond doubt is the substantially higher risk the virus poses to the elderly and otherwise infirm. People who are young and fit and healthy have less to worry about (in terms of risk of their own death), but could still get very sick and would pose a risk to the lives of those around them.
I personally would not participate in large group gatherings. I'm continuing to work from home, and I have my groceries delivered to my home. I wear a mask every time I leave the house. I avoid touching my face when I'm out, and wash my hands when I return home.
I can see how easily those numbers could be adjusted based on contextual factors, so there is a way I feel like I can't tell what is actually going on with numbers infected. What is your take on that?
I think your degree of uncertainty is appropriate given the circumstances. The data is very contextual. Again, I would encourage you look to a few trustworthy professionals to help you interpret the relevant information. Ignore media reports.
I hope this is helpful.