Saturday, February 29, 2020

Finally real COVID-19 testing to begin

From today's (2/29/20) NYT:
The U.S. Food and Drug Administration announced Saturday that it was authorizing American laboratories to develop their own coronavirus tests, which should significantly increase the country’s testing capacity.
The effect could be rapid. About 80 labs and private companies have applied for emergency approval for tests they have already created. If they have submitted evidence that the tests work, the labs and companies will be able to use them immediately, rather than wait for the F.D.A. to complete reviews and issue approvals.
“This action today reflects our public health commitment to addressing critical public health needs and rapidly responding and adapting to this dynamic and evolving situation,” the F.D.A.’s commissioner, Stephen M. Hahn, said in a statement.
Experts have been frustrated with the limited availability of coronavirus tests in the U.S., which until now could only be provided by the Centers for Disease Control and Prevention. Broader testing will enable more rapid detection and isolation of people who have the coronavirus to help contain the spread of disease.

Why did this take so long?

What about antibody tests?

From the Jerusalem Post:
The US Centers for Disease Control and Prevention (CDC) may have harmed the country's ability to track and detect the spread of the COVID-19 disease due to a series of missteps, including refusing to use the tests recommended by the World Health Organization (WHO), according to ProPublica. 
As the virus began to spread, the CDC decided to start creating its own, more complicated test instead of using the test guidelines provided by the WHO. The test was made to check for a variety of different viruses. When the test was sent to labs across the country, it didn't work and falsely flagged the presence of other viruses in harmless samples.
Update:  From the WaPo of April 5, a detailed chronology of how our two public health agencies totally screwed up testing for Coronavirus in the US.

Thursday, February 27, 2020

Good advice on infection prevention from the WHO, and a great deal of caution

I think all the advice provided here (dated today 2/27/20) is sound, and is not alarmist. This virus has spread remarkably quickly, frequently triggers lethal illness, and there has not been anything like it since the 1918 flu.   COVID-19 must be spreading within the US in ways we don't know yet.  The lack of available testing massively exacerbates this problem.

WHO has issued simple but practical guidelines for how to do the best we can given the current situation, even if people lack access to masks, gloves, etc.

The only WHO advice that I am not sure is correct is the advice to use hand sanitizer (generally 70% alcohol).  I have not seen any country guidance that says alcohol sanitizer has been tested on COVID-19 and works.  The only decontaminant that US and UK authorities recommended is bleach, at a specified dilution.  Be aware that viral killing might require more than a quick swipe with a "wipe" that has been wet with dilute bleach.  So far we have not been told the details on how to effectively decontaminate surfaces, although testing should have been done by now.

I should note that while China has been widely criticized for hiding or underreporting COVID cases, the US is doing the same thing by withholding testing from all but a tiny number of potential patients.  Furthermore, lack of clear identification of potential cases opens the door to massive spread, as noted in this piece today by Helen Branswell of STAT:

... Eventually, more than 10 days after she went into hospital, the CDC agreed she could be tested. Dozens of health workers who may have come into contact with her at NorthBay VacaValley Hospital, in Vacaville, Calif., are now being monitored.California Gov. Gavin Newsom was critical of the testing debacle in a press conference on Thursday. His state has only 200 kits to test for the new coronavirus, he said.
“Testing protocols have been a point of frustration for many of us,” Newsom said. He added that, based on conversations with the CDC, states have been informed new protocols are coming and they have been promised an “exponentially” increased capacity to test.
Indeed, on Thursday the CDC announced a new testing protocol that will greatly expand the number of people who should be tested.
Requests for comment from the CDC on Thursday went unanswered... 

Here's what we need to know about COVID-19 testing

The problem of highly restricted testing in the US

The COVID-19 virus is quite deadly (mortality appears to be between 2 and 10% as discussed in my last post) though since we don't know how many asymptomatic or mildly symptomatic cases there are, we still can't be sure of this.

Why don't we know the number of cases?  In the US, so far you can't order the PCR (or any other test) without CDC approval.  Approval is contingent upon meeting a high bar, including known exposure.

Thus we are missing lots of cases, who are not being asked to self-quarantine if the cases are mild, and are thus likely to spread disease.

Why aren't we testing broadly to get a sense of actual case numbers?  The best guess is that CDC does not trust its own test, and that by restricting testing to those who are almost certainly positive to start with, they will make fewer diagnostic errors.  

Tests kits were sent to state labs by CDC, then recalled, with the excuse that there was a problem with a "reagent".  A reagent could be any component of the test. CDC has not replaced the reagent or the test kits.  States are begging for them.

Dr. Messonier (head of Vaccines and Respiratory Diseases at CDC) says CDC is keeping up with testing, there is no backlog, they can test 400 samples together, and turnaround is overnight.  That should clue you in that relatively few are being tested.

Some other countries are reporting large numbers of new cases daily.  Do they have better tests?  Has any journalist looked into the relative sensitivity and specificity of tests being used around the world?

PCR is a rapid screening test, which when positive almost always connotes a true positive.  Negatives may actually be positive if the sample does not contain enough virus or is collected or transported incorrectly.

What is also needed are viral culture tests, which could theoretically pick up cases missed by PCR.  Such cultures might also be used to test the effects of drugs on the virus.  What work is being done to develop COVID  viral culture techniques?  Might such tests already exist, since they would probably be very similar to those for SARS and MERS?

Update March 30:  FDA says it is not their fault that tests were not available for novel coronavirus.  Between the lines, I am reading, "It was the CDC, not us, and we accept none of their blame."

Update Feb. 27:  Australia developed viral cultures; so has CDC:

Akst J. Australian Lab Cultures New Coronavirus as Infections Climb. The Scientist.
CDC has grown the COVID-19 virus in cell culture, which is necessary for further studies, including for additional genetic characterization. The cell-grown virus was sent to NIH’s BEI Resources Repository for use by the broad scientific community.
If there is a bat or other animal model for this virus, so much the better for testing drugs that might show efficacy against COVID-19.  What work is being done on developing or using animal models?

I assume one gains immunity after infection with COVID-19, which usually occurs after infections, but not always.  The next test needed is an antibody or T-cell test that will accurately identify immunity to the virus.  People who have recovered should be studied to identify markers of immunity.  Then, we will be able to identify those who can safely care for infected individuals without catching the disease themselves.

Let me reiterate:  with the limited testing now being performed in the US, we are definitely missing many cases, and missing the opportunity to appropriately use quarantines to limit the rapidity of spread.  Slowing spread is critical to producing enough masks, gowns, ventilators, ECMO machines, training hospital staff, building hospitals just for COVID-19 cases, etc.   

We need an accurate test, probably a PCR test, NOW.  What is holding it up?

Update Feb. 27, from WHO:

Ensuring ongoing test availability. WHO has procured a commercial assay (manufactured under ISO:13485) with strong performance data and shipped to over 150 laboratories globally as an interim measure for Member States requesting support. The main goal is to strengthen global diagnostic capacity for detection of the COVID-19 virus. Support is now also provided to ensure the quality of testing through the implementation of an External Quality Assurance mechanism.
Update March 27:  China has a monkey animal model of disease as well as antibody tests
Update:  New CDC Testing Criteria, issued 2/27
Criteria to Guide Evaluation of PUI for COVID-19
Local health departments, in consultation with clinicians, should determine whether a patient is a PUI for COVID-2019. The CDC clinical criteria for COVID-19 PUIs have been developed based on available information about this novel virus, as well as what is known about Severe Acute Respiratory Syndrome (SARS) and Middle East Respiratory Syndrome (MERS). These criteria are subject to change as additional information becomes available.

Clinical features and epidemiologic risk
Clinical Features&Epidemiologic Risk
Fever1 or signs/symptoms of lower respiratory illness (e.g. cough or shortness of breath)ANDAny person, including health care workers2, who has had close contact3with a laboratory-confirmed4 COVID-19 patient within 14 days of symptom onset
Fever1 and signs/symptoms of a lower respiratory illness (e.g., cough or shortness of breath) requiring hospitalizationANDA history of travel from affected geographic areas5 (see below) within 14 days of symptom onset
Fever1 with severe acute lower respiratory illness (e.g., pneumonia, ARDS) requiring hospitalization4 and without alternative explanatory diagnosis (e.g., influenza)6ANDNo source of exposure has been identified
Update 2/28:  Initial CDC testing info and interpretation protocol:

2019-nCoV rRT-PCR Diagnostic Panel Results Interpretation
2019 nCoV_N12019 nCoV_N22019 nCoV_N3RPResult Interpretationa
+++±2019-nCoV detected
If only one, or two, of three targets is positive±Inconclusive Result
+2019-nCoV not detected
Invalid Result

Assay Limitations

  • Analysts should be trained and familiar with testing procedures and interpretation of results prior to performing the assay.
  • A false negative result may occur if inadequate numbers of organisms are present in the specimen due to improper collection, transport or handling.
  • RNA viruses in particular show substantial genetic variability. Although efforts were made to design rRT-PCR assays to conserved regions of the viral genomes, variability resulting in mis-matches between the primers and probes and the target sequences can result in diminished assay performance and possible false negative results.
Update 2/28:  NPR reports that China and Singapore have an antibody test they are using to detect recovered or asymptomatic cases.  ZeroHedge reports that CDC has only tested a total of 466 people in the US, as of past 2 days, while UK has tested 6,000.  If so, CDC needs to get off its behind and get a reliable test and make it available to all labs in the US with rtPCR capability.  Who knows how much community spread is ongoing when CDC refused to test people unless they had a close association with China and were sick with an illness consistent with COVID.

Is Trump aware that the head of the COVID response at CDC, Lisa Messonier, is the sister of Rod Rosenstein, the Deputy AG who suggested secretly recording Trump?

Friday, February 21, 2020

CORONA VIRUS update: Fomites, PPE, Mortality


These are inanimate objects like doorknobs, shared computer keyboards, surfaces where droplets from coughs and sneezes land, or where hands have touched, which may transmit the virus.

Questions that the authorities should be making every attempt to answer include the following.  

a)  What % of infections are transmitted via fomites?  (In other words, how careful should you be when touching things, including objects you cannot avoid touching, like subway turnstiles, where virus may be present?)  The danger is touching a contaminated surface, then touching your eye, nose or mouth (mucous membranes) with enough live virus to cause infection.  Good patient histories and testing of COVID survival under different environmental conditions should shed more light on this.

b)  How many viral particles are needed to cause infection?  (The authorities claim COVID was transmitted from animals.  If so, there should be an animal model of infection, enabling lab testing of infectious dose, modes of transmission, etc.)

c)  How long does this virus live on objects, under different conditions of temperature, humidity, and surface type?  This can be easily tested, and CDC should stop guessing..

d)  What kills this virus on surfaces?  Alcohol hand rub (~70%) works to destroy some viruses but not others.  It is unreliable for Norovirus, for example.  Is coronavirus killed when alcohol rub is used for, say, 20 seconds on hands?  Why hasn't CDC given us this information?  

How are hospital rooms being decontaminated effectively in China?  

Personal protective equipment 

The photos of Chinese doctors in full gear worry me more than anything.  They are dressed in maximal containment suits--identical to what would be used if the patients had Ebola.  While it is said that COVID is less contagious than SARS, which China had a lot of experience with, the PPE suits suggest they believe it is extremely contagious.


According to the South China Post, where the photo below also came from, the mortality rate is actually 10%.  This article explains why estimates of mortality rates for COVID-19 could be too high or too low at this point. 

Let me explain.  There are 20,833 recovered cases, and 2360 deaths.  Another 55,000 people are still being treated.  So adding the number of recovered plus number of deaths you have 23, 193 people whose illness is over.  Of this total, 90% recovered and 10% died.  I don't know if these numbers are accurate, but they are what has been reported from China.  They are not encouraging.  This looks to me like the repeat of the 1918 flu (except it is a coronavirus, not influenza virus).  Hopefully I will be proven wrong as we learn more, but the degree of rapid spread and the potential for prolonged incubation periods (suggesting quarantines are of limited value), the lack of effective drug treatment, and the mortality rate suggest otherwise.

We are unlikely to have enough masks and gear to deal with it.  Best thing to do right now is to practice not touching your face with your hands, and to become aware of what your hands touch, and how to touch things in ways that place yourself at less risk.  In other words, by a) using gloves, b) handwashing before touching your mucus membranes, c) perhaps using napkins or paper towels or other disposable items to shield your hands in public areas like buses, subways, toilets to avoid contamination.  Whether it will be possible to maintain a minimum 6 feet of distance between yourself and someone who is coughing or sneezing is another question.