March 21, 2020
A new source of anxiety for me is that this is not a healthy pace for science. Twice in the last few days, I got faked out by new data because it seemed exciting and landed in very reputable journals. In this age of pre-print articles having great influence, I at least thought that the peer review process could help filter out some of the noise. However, on closer reading I found myself confused with some very basic stuff like figure legends, definitions, study designs and selection of statistical tests. In one case, because the data had me excited about my test and treat early idea (https://www.jtsid.com/post/lower-the-death-rate), I emailed the corresponding author who graciously emailed back almost immediately. He told me that the data in question had been gathered under some duress and that many of my most basic questions could not be answered. This was a distinguished scientist with a substantial legacy of contribution: there was a sadness in the tone of his words.
This story is understandable. When hospitals are running low on PPE, starting to lose health care workers to sick days, and stressed beyond belief, research must seem like a secondary priority to those on the front line. Plus, many of our labs are almost completely shut down to ensure physical distancing. Even computational groups like mine are feeling pressed to get relevant data published at literally ten times the speed we are accustomed to. It leaves too much space for oversights and cutting corners.
Today our president tweeted about the hydroxychloroquine + azithro results and sounded like a carnival barker. As I described yesterday, these results are questionable https://www.jtsid.com/post/march-19-2020). Under normal times, they would not make it past an editor at any major journal.
There are niches within science that are critically important to this pandemic, for which I have little insight whatsoever. Perhaps the most relevant is vaccine development and production. I truly hope that the dedicated scientists in this space are feeling the urgency but not letting it impact the care with which they do their work. I would rather a have a great vaccine in 2 years than nothing at all, because of hasty decisions that were made along the way.
Despite these unprecedented challenges, it still falls upon editors to not simply accept an article because it contains the word COVID and comes from a hot zone. It is completely appropriate that scientists are being pushed to act rapidly during this crisis. The all hands on-deck attitude is wonderful to witness. However, at the end of an experiment, the data is the data, and there is always an upper limit to its usefulness. The results of the hydroxychloroquine + azithro study are the most obvious example: from that study, we learned only one thing, that we need to do more and better studies of this drug combination. Sorry to say, but this will take months even if things go perfectly.
Even small overstatements or misinterpretations can propagate into massively, misguided investment decisions about how to allocate and optimize funds towards diagnosis, treatment and prevention. Therefore, I hope that we can all take a break from our overwhelming degree of busyness and continue to ensure scientific excellence and integrity.
As far as interpreting new information in the press, any result in the next month claiming one of the following: superinfection, increasing virulence of the virus or successful new treatments, have to be considered as preliminary, at best.
I did not read as much as I would have liked the last few days but I really thought this was a fantastic article. Most of it has been covered in this blog in form or another already. However, I loved the concept of “the dance” detailed in Charts 13, 15 & 16. We are in the midst of the hammer and it is terrifying because we know that a substantial dose of pain is coming for all of us. However, how our society handles the dance is really what dictates our future, The concept of ranking distancing measures according to utility in lowering R0 but also inflicting economic pain is brilliant:
Rural COVID. I loved this piece. We travel to small town Northern and Eastern Washington frequently for soccer, and I am fond of these areas, so I worry and wonder how rural areas will fare. Epidemiologically, this article made me think about the spread of HIV in parts of sub-Saharan Africa along truck routes. Emotionally, it brought out the intimacy of these unique communities. I caught up with a great friend yesterday who lives in the rural southeast. He and I both wondered if small town America will cope or not. A lot of the articles on this subject have been on the red / blue divide in response to the crisis. They articles are not hard to find but I will spare you. Instead read this for a dose of shared humanity:
Cruise ship epi. From the Diamond Princess. Very important data. These investigators report a remarkably high percentage (32%) of virtually asymptomatic infected people, most of whom were middle aged or higher. Yet 27% were severe (required oxygen), 8% “deteriorated” but none died. More infected people did not have fever than did. More infected people did not have cough than did.
Over half of non-severe cases (asymptomatic + mild) had abnormalities on chest CT. Ordinarily I would frown upon ordering a chest CT on a person who is barely ill. However, I thank the investigators for doing this. We now know that this is a lung disease, even in people who barely feel it: bummer. It is also an upper airway disease which is why it transmits so well:
Yesterday, I described what I am hearing about critical illness for COVID. How about the more common scenario of a self-resolving illness in the common man? A very close friend of mine with a great sense of humor was diagnosed a few days ago and I have been checking in on him. His report: “I think it is more sleepy than actually feeling sick. It’s kind of a strange sensation. It is kind of like how I feel every single day of my life except more extreme.”
Most personal inquiries I have received over the last 48 hours pertain to trying to track the epidemic in different countries. For this reason, I am going to repost a couple of key links to follow and understand:
This is a critical and extremely cool document. You can follow the real time R0 (the effective reproductive number) in many countries to see which direction things are going, and also how certain we are about these estimates (notice the wide confidence bounds for the US due to undertesting):
This is the state by state data but be careful. As I mentioned the data gathering is reliable. However, the discrepancy in numbers of tests per state is appalling so these are not real case counts:
Reasonable modeling estimates for the State of Washington:
Be safe everyone.