March 19, 2020
I had clinic today which was so enjoyable. The SCCA was a bit of a ghost town, but still less different from baseline than anywhere else I have been in the last 10 days. Given the immune-compromised nature of our patients, we are always hyper aware of germ prevention, cleanliness and overall precautions. While elective stem cell transplants are being deferred, for our patients with acute leukemias, their process cannot stop for COVID. Today I met some young people who have already suffered an enormous amount and must be aware that this virus is a potential mortal threat. Yet, they handled themselves with grace, humor and hope. It was a good day.
I only have time for a few articles today. Hopefully more tomorrow:
My dad sent me this presentation from Xihong Lin at Harvard & the Broad Institute. I do not think it is an exaggeration to say that this is a document of historical importance: https://drive.google.com/file/d/14tGJF9tdv4osPhY1-fswLcSlWZJ9zx45/view
I recommend going through it slide by slide. No scientific training is necessary to get the gist.
Slide 36 describes centralized quarantine measures which decreased R0 from 1.25 to 0.32. The effect over 2 weeks was to virtually eliminate the local epidemic. As long as books and laptops are permitted, I personally would be more than happy to sleep in a gym for a couple of weeks to slow things down, save lives, save local businesses and get our lives back to somewhat normal more quickly. I wonder if this is even being considered as a back-up plan in the US. It certainly should be.
Of the 42,000 health care workers flown in from elsewhere in China to Wuhan, zero were infected. Zero. So if you want to avoid getting COVID-19, then one great strategy is to fly into the world’s most intense focus of infection and then care for incredibly sick infected patients for 16 hours each day. Of course, that assumes you live in a country with adequate supplies of PPE. The fact that PPE are running low for our front-line providers is a disgrace. It makes my blood boil.
A vital component of China’s success was early diagnosis and treatment. They treated their mild cases too. I wish I knew more about which drugs were used and which drugs worked.
China’s role in all of this is obviously extremely complicated. I think it is completely fair to take China (and many other countries in Asia) to significant task for not eliminating wild animal sales in markets after SARS v1. Even without SARS CoV-2, this is a despicable practice. The lies and cover ups during the initial weeks of the Wuhan epidemic also reveal weaknesses of all authoritarian systems.
However, the heroism of the Chinese people, and the decisive decision making of its government to contain the epidemic, must also be applauded. Had these early and dramatic interventions not been done, we all would have been hit much earlier and more frequently. The remarkable story described in this slide stack saved 1000s of lives in Wuhan and China, and also gave the rest of the world time to prepare. Many countries in Asia used this time valuably (see https://www.jtsid.com/post/march-12-2020). Europe and the US did not.
We live in a banana republic. When you vote for an openly cruel and autocratic bully, who flouts his top-tier idiocy on a daily basis, eventually the whole country is left to pay the bill. This is what happens when expertise is dismissed, and when obviously uncaring people are given the reigns. Trump’s plan to make states compete for tests has resulted in complete chaos that is evident if you look at all published data on positive and negative tests by state. New Mexico has 2762 negatives and 35 positives. South Dakota has 551 negatives & 11 positives. These are the types of positive to negative ratios one would hope to see to have any hope of containing a local epidemic.
But. New Jersey has 210 negatives and 742 positives. Ohio has 119 positives & 140 negatives. Thanks to the Atlantic, we have these numbers as the CDC is not providing them: https://covidtracking.com/data/Clearly, these 2 states are only testing hospitalized patients and those with a very high probability of infection. They have no idea who is infected in the general population and are not tracking the full extent of local epidemics at all. Expect the worst: nursing home outbreaks, overwhelmed hospitals, tragedy in health care workers and delayed interventions.
Lack of testing is only part of the problem:
Washington epidemic. More not great news. I have been restraining myself from looking because I would not have expected any signal until at least a week after large scale physical distancing. I finally peeked today on covidtracking.com which is data compiled from the Atlantic. The data is very much imperfect because testing is still not widespread, is probably biased against more mild cases and may vary from day to day. With those caveats in place, I see no evidence of social distancing impacting the local epidemic as of yet. If these trends continue a week from now, then we are really facing the worst. Here are the take home points.
1. Cumulative cases are still in an exponential trend. Doubling time ~3 days:
2. New cases appear to be increasing without any sign of deceleration towards the desired bell curve shape:
3. The proportion of positive tests started higher (probably biased by those with higher likelihood of infection getting tested but also appears to be leveling off at ~6% or maybe even increasing:
Few of the people I have spoken in Seattle who have a decent to high chance of infection (fever / cough) have an easy option to get tested. Still, it may be a bit too early to see an effect of social distancing. I am going to try to just chill out and not check the #s again for at least another 4 days.
Hydroxychloroquine. This is an anti-malarial drug that is used also for rheumatologic diseases. It is not without issues in respect to side effect profile. It has been shown to have broad antiviral activity in plates of cells in the past but has never demonstrated enough oomph for it to be used as a standard of care. In general, drugs that are selected from a massive library of compounds specifically for their direct antiviral activity have a much better track record than those which are repurposed such as hydroxychloroquine. Nevertheless, at this point any data is welcome and I think this agent should be front and center among those being considered as treatment and prevention options for COVID.
But…. today’s new data does not meet the standard required for informed clinical decision making. They showed lower viral loads with the drug, particularly when it was combined with azithromycin (of Z pack fame). The fact that the control group was selected out of convenience and that the study was non-blinded, introduces massive bias into the results. The treated and untreated are not matched at all according to age and the numbers are small. The data is reported in a pretty haphazard fashion with no a priori hypotheses and undefined p-values scattered all over the paper.
I do like that they used nasal shedding as an endpoint and as a potential correlate of potential for severe disease & transmission: our group has recognized that this will be necessary to power studies (show statistical significance) with a smaller number of participants.
Overall, this study is enough to justify testing this agent prospectively in a proper clinical trial (which can be performed relatively quickly) and it is worth considering its use in real world situations. However, there is no certainty whatsoever at this point, that hydroxychloroquine has any clinical benefit for SARS CoV-2. My hunch is that there are better options out there.
This drug is giving me hope. I cannot wait to see the data and hope it is not oversold. It is bizarre that results are in the newspaper before being in any medical journal at al. This would never happen in ordinary times. Hoping to see a New England Journal of Medicine or Lancet paper any day now if this is really true: https://www.theguardian.com/world/2020/mar/18/japanese-flu-drug-clearly-effective-in-treating-coronavirus-says-china
On being a scientist vs a clinician. I am both, though far from perfect at either. I love having a foot in both pools because the thought processes feel so different. The fun part of science is the rigor (doing things properly), but also the creativity. The latter component is the feature of working in science that is missed by the general public. Great experiments or cool new hypotheses often arrive unexpectedly and at weird times, like during a run or a few drinks with colleagues. These moments of inspiration are really exciting.
There is certainly extreme rigor in clinical practice as well. However, the aspects that I enjoy the most are the patient interactions (patient care is a natural way to get exposed to a wide breadth of society) and the pattern recognition, which I would define as the process of almost unconsciously tying together ~10 pieces of data to arrive at a diagnosis or a treatment plan. The latter aspect is the crux of training that allows a good clinician to make better bedside decisions than detailed textbook knowledge alone would allow. Pattern recognition is built from experience.
Pattern recognition is not the same as hard data and is fraught with mental terrain traps. Clinician judgement is easily colored by numerous biases including the Dunning-Kruger effect, and confirmation bias. Younger physicians are now trained about these potential biases which is nice to see and extremely helpful.
I have yet to see a person infected with SARS CoV-2 (other than friends who have it and seem to be doing just fine). I think for most infected people there is not much to see.
However, I am starting to read personal reports from ICU docs about the clinical course of the sickest patients. Already they are seeing patterns unique to this infection. With the huge caveats of bias and this being very early in the epidemic applied, I have heard (mostly on facebook threads, so again take this with a huge grain of salt) that COVID can be a cruel disease. Patients decompensate rather quickly and require lots of oxygen. They then often get better on the ventilator and appear to be on a slow road to recovery, but then may develop irreversible distributive shock and die. The presentation is apparently quite distinguishable from severe flu. More data and experience are needed, but I appreciate the posts of those in the trenches which will help many providers who will see their first severe cases in the coming week.