On tap this week, we conduct our usual stroll through the current numbers. In addition, we will introduce another wrinkle this week, our old friend influenza. Other topics for today include a bit more discussion about the COVID vaccines that are likely to be available within the next couple of weeks, testing options for COVID-19, and some of the recent changes in CDC recommendations. Cannot wait for some better things on tap. Some Prairie Bomb in particular would be quite nice. Man, oh, man, oh friend of mine, all good things in all good time…….
Since I am at a loss for words with how to reach the general population as to the severity of the situation we are in, I have selected instrumental jazz numbers to fine tune your minds today. Aptly named All Blues, this tune from Miles Davis is off of his Kind of Blue album from 1959. This album is considered one of his finest by most (I prefer his early 1970’s work), and is often cited as the best-selling jazz record in history. If inspired to ever check out a legend at his finest, A Different Kind of Blue includes a show at the Isle of Wight.
Fun With Numbers
As we have the last few weeks, we will start with a quick European update before presenting this nations numbers followed by those of Suffolk County.
The European Status Report
In Europe, the second wave reached their pain point prior to us reaching ours in the United States. As a result, they decided to shut things down again around five weeks ago. What you can see from their chart is that it took a week for the increase in new cases to begin to taper, with a subsequent weekly drop in new cases since that time. Last week we reported that there appeared to be a flattening of the curve with respect to European COVID related deaths. This week, it appears that they are finally seeing a small decline in weekly death tolls. The take home point here is that any change in behavior can have an effect on cases in about one weeks time, but the effect on fatalities will trail by about five weeks. Perhaps I should distill this message a bit further:
You can in fact make a difference in how bad this pandemic affects our collective ability to remain above ground through the choices you make!
A double dose of Jerry today. This is some Dawg style for you. Garcia and David Grisman performing “So What” by Miles Davis, another track from that classic Kind of Blue record. Selected here for the Fun With Numbers session, primarily because the “So What” attitude of this nation is the driving force behind the disturbing numbers you are about to be presented.
So where were we five weeks ago? That’s right, we were gearing up for Halloween. So how did that work out for us? We already knew that the number of cases began to increase about one week following Halloween. And as expected, that effect is now beginning to manifest in markedly increased rates of death. We are now approaching 3,000 deaths on a daily basis in this country. This represents three times the rate we were seeing a month ago.
Entering the Thanksgiving holiday, the number of daily cases was twice that of the week leading up to Halloween. This means that five weeks from then we anticipate a doubling in fatalities. That places you at roughly 5,000-6,000 deaths daily. But alas, Thanksgiving was no normal holiday. Folks traveled in abundance and gathered in mass. And we are starting this week to see the effect of that. It is not pretty. The daily case rates are now nearly triple that leading into Halloween and likely to go up considerably again this week. Do not be surprised if there are 10,000 deaths per day to ring in the New Year in this nation.
Above image is from El Paso, Texas a couple of weeks ago where they are using refrigerated trailers to manage their ghastly situation.
While for the most part we have not felt the effects of the “second wave” much locally, this has begun to change drastically for the worse. Pre-Halloween we were seeing roughly 100 cases per day in Suffolk County. Post Halloween this started to increase to about 500 cases per day leading into Thanksgiving. The effects of Thanksgiving are starting to become evident, and we are now seeing over 1,100 cases per day the last few days. I would expect this upcoming week to be significantly worse as the full Thanksgiving effect starts to rear its head.
Hospitalization rates locally are also on the rise and are reflecting that post-Halloween surge in cases. Where we were seeing roughly 35 hospitalizations per day leading into Halloween, we are now at 10 times that number.
Not surprisingly we are starting to see fatalities significantly increase as well. For months the death tally looked like a nice horizontal line on the chart with few new deaths in Suffolk County. There were six deaths in October in Suffolk County. That number was matched in the first two weeks of November. There have been 38 deaths since that time.
Peds First Pediatrics Data
Our experience here is no different than the rest of the locality with respect to case numbers increasing. Our silver lining is that children continue to show far lesser symptoms. We have not had needed to admit any children to the hospital to date, nor have there been any severe outcomes we are aware of in any of our patients. As you can see in the table below, not only are our case numbers increasing, the percent positive rate for those tested has been rising quickly. Feeling ill looking at these numbers yet? The next section may leave you a Whiter Shade of Pale.
A little more Miles Davis today. Entitled Double Image, this track is from the Complete Bitches Brew Sessions. So selected as we are about to have two serious viral infections circulating.
Lest we forget our usual wintertime buddy, influenza. This year thus far, things have been quiet regarding influenza. That is likely to be changing soon. This situation has the potential to really throw kerosene on this fire. Influenza itself can be lethal. With the healthcare system functioning at its limits with COVID alone, this situation needs to be monitored closely. Among the greatest challenges will be setting up systems in a manner to minimize COVID patients and influenza patients from cross pollinating one another, especially in the hospital setting. While at this time the percent positive rate of those tested has not increased significantly, the overall number of patients with influenza is beginning to rise locally. How many more weeks before we start feeling its effects and to what degree is unclear at this time.
A Little Bit About COVID Testing
While I have mentioned predictive values previously, I am going to provide a bit more detail on it today to provide you with a hopefully better understanding of the types of tests we use and their respective limitations. This issue has been raised often of late as PCR testing is now being required for some clearances to school or other activity. Concerns about the antigen test were raised in a recent New York Times article as well. By the time you are done with this section you should understand why we implement the testing strategies we do.
Definitions By Example
Sensitivity: To test the sensitivity of a new pregnancy test, it is administered to 100 women who already had their pregnancy confirmed by sonogram. The test was positive in 96 of them. The sensitivity of this test would be 96%. The false negative rate of this test would be 4%.
Specificity: To test the specificity of a new pregnancy test, it was administered to 100 women not named Mary, and who had never engaged in any behavior that might lead to pregnancy. Of the 100 women, 3 tested positive. The false positive rate of this test would be 3%. The specificity of this test is 97%.
*For both sensitivity and specificity calculations you are comparing a test with a known population of subjects.
Positive Predictive Value: “The test is positive? Doc, is that true?” If the test is a flu swab in July, a positive test has a low predictive value for being a true positive in New York. That same test in January has a much higher predictive value in a person with fever, cough, sore throat, and body aches.
Negative Predictive Value: “The test is negative? Are you sure, doc?” If you are asking that question because the cardiologist doing the ECHO on your heart says that you either lack a heart beat or are altogether heartless, the test has a low negative predictive value because you are actually asking the question. Unless you are the Tin Man that is.
*For both Positive and Negative Predictive Values the calculations involve testing a subject for which you do not know the answer. Whether that subject is likely to actually have the condition impacts these calculations.
If you are interested in a great review with nice pictures on this topic as it pertains to COVID testing, click here.
Went for the trifecta with Jerry today. An acoustic instrumental interpretation of a Bob Weir song.
Antigen vs PCR Testing for COVID-19: The Numbers
As most are aware by now, there are two methods to test for current infection with COVID-19, antigen testing and PCR testing. Antibody testing is to assess for past infection and therefore not within the scope of discussion.
PCR testing is considered the “Gold Standard” for testing. It is important to note that PCR testing is not perfect. It has a sensitivity of roughly 80% in people with symptoms. That means it returns a false negative result roughly one in five times in patients that are known to be infected. The specificity reported in the same review article for COVID PCR testing was 95%. This means that for every 100 people that are known not to have COVID, 5 will test positive anyways. When the prevalence in the community is high, as it is currently, the positive predictive value is very high. The negative predictive value worsens however with increased prevalence.
The PCR test does not produce an immediate result, and when there are lots of tests being performed, it may take up to one week to obtain results. This is not ideal if your goal is to identify and quarantine positive cases before they have a chance to spread the love. However, it is a very good test in other ways and will circle back in a moment to discuss.
Antigen testing has the benefits of being both cheaper and quick to perform. The company’s data on sensitivity and specificity of the antigen test were determined based on comparison with PCR testing in symptomatic people. This is comparing a test to an imperfect test. That being said, the company’s data showed a sensitivity of 95% and a specificity of 100%. Its reported sensitivity is higher than PCR because it is being compared to PCR rather than the same measures that PCR was compared to. This is why comparing apples with apples is important. Suffice it to say, while the article in the New York Times discussed that there were false positives and negatives, these validity of those results is limited by the means of comparison.
In the study discussed in the New York Times article, nearly 2500 people were tested by antigen and PCR testing. Of those 885 had either symptoms or an exposure history. Key again is to remember the antigen test was designed for symptomatic detection of COVID. Detecting a positive among a population with exposure history but without symptoms is not what it was intended to be used for. The other 1551 subjects were randomly selected and without symptoms. Again, not the intended use of the antigen test.
Of the 885 people of the higher risk arm of the study, 305 tested positive by PCR testing. Nearly one in five of these was missed by the antigen test. What was not disclosed was how many of those “misses” were asymptomatic exposed patients as opposed to symptomatic patients. Of the 1551 in the lower risk arm of the study, 19 were positive by PCR and only 6 on antigen testing. The article also mentions that 7 people were “falsely” identified as positive by the antigen test. Taken at face value it means the positive predictive value of the test was 97.3%. That is not shabby for an immediate answer in addressing a disease for which delays in quarantine and contact tracing are of prime importance. The number may be slightly better when you consider the comparison is with PCR testing. It is possible the antigen test identified something the PCR missed!
Antigen vs PCR Testing for COVID-19: Strategy
When someone requires testing, starting with an antigen test has benefits, especially in the symptomatic person. Quick and less expensive testing to identify and mitigate risk of a person with COVID-19 is extremely advantageous. The positive predictive value is sufficient to diagnose and quarantine. The few percent chance of a false positive test is outshone by the mitigation potential of not waiting days for an answer. And that is again assuming that those are truly false positives (as opposed to a PCR providing a false negative).
When an antigen test is negative, the interpretation of the result must be viewed carefully, especially in a person without symptoms who is being tested for exposure concerns. In these cases, a back up PCR test is recommended to ensure maximum identification of those infected with COVID-19. This is perhaps the best take home point for the study discussed in the New York Times. The negative predictive value of the antigen test, in asymptomatic people especially, is not sufficient in times of higher prevalence of disease to base decisions.
As it is asked often, we do not send backup PCR tests on positive antigen tests. Due to the high positive predictive value of the antigen test and the significant false negative rate on a PCR test, it is not safe to assume the negative PCR test is the “true” answer.
To address another question that is frequently asked about testing. You do not require at test at the end of a full 14 day quarantine to return to any activity if you remain symptom free after a positive exposure. Per the new CDC recommendations, there is the option for 10 days symptom free without testing, but they acknowledge that there could be a few contagious folks reentering society with that recommendation. Additionally, you do not require a test after testing positive for COVID if you have quarantined for 10 days after the onset of symptoms to prove you are now negative. It is possible to test positive by either antigen or PCR testing for quite awhile past the point of being a concern for contagion. If your job or school is requiring testing for these reasons, it is due to either a lack of knowledge of the guidelines or pure ignorance of the science that underlies the testing.
Remember the goal: identification and quarantine of infected individuals in as quick a manner as possible to prevent continued spread in the community.
Changes in CDC Recommendations
Last week the CDC released some alternatives to the now standard 14 days quarantine for directly exposed individuals. The rationale for these alternatives is almost like throwing in the towel. Specifically, because people refuse to follow directions, they are offering options they acknowledge are not ideal in the hope that enough additional people follow directions to offset the folks leaving quarantine earlier while still contagious. This is like giving your four year old a cookie after eating two bites of their dinner because you are happy they ate something and don’t want to deal with a tantrum. North Korea is the opposite extreme of course. They have a very different way of dealing with non-compliance.
The new options are to:
- End quarantine after 10 days without any testing if no symptoms have been present throughout those 10 days
- End quarantine after 7 days with a negative test within 48 hours prior to ending quarantine
These options are not currently considered acceptable by New York State. As you can likely tell from my tone, I disagree with the new options.
This video is Wynton Marsalis performing the John Coltrane classic Giant Steps. Hopefully, we too will be taking some giant steps forward shortly.
The FDA has a meeting set on December 10th to address Pfizer’s COVID vaccine application for emergency use authorization. A similar meeting is scheduled the following week on December 17th to discuss Moderna’s candidate vaccine. Unless there is something negative that has yet to be disclosed, expecting both of these vaccines to receive permission for administration during the pandemic, while they continue to monitor how well they work and with what side effects. To date, no serious side effects have been attributed to these vaccines.
Not surprisingly, the CDC, and decision makers in general, have made the recommendation to immunize the highest risk first. This would include nursing home residents and workers along with frontline medical staff. Not surprising, because when you are managing with short supply of any remedy, you select the best risk/reward ratio you can. Some may argue that elderly do not always respond as well as young people to vaccination, and that they may be “wasting” doses as a result. Fact is, the elderly in nursing homes are such high risk for poor COVID outcomes, that there can be no waste in the effort to protect them. Expect that this group of individuals will be receiving their doses over the next two months.
Once the high risk are immunized, the subsequent recommendations will likely follow the same logic as highest risk first. That would leave children last in line for the vaccine as they are the lowest risk population. Trials of the vaccine in children are only first beginning at this time as well.
While the expectation is that the highest risk populations will be fairly apt to accept the vaccine, there is some concern that when available there may be difficulty getting the rest of the population to accept the vaccine. It is my hope that the mere fact that by the time they are eligible to receive it, the couple of million doses administered to the high risk populations will provide enough data to prove effectiveness and reassure that no serious side effects are to be expected. For the sake of returning to normalcy, let us collectively hope this poke works as well as the trial data released indicates, and that enough people voluntarily take it to end the pandemic.
Just as a tease for future discussion….Once the vaccine is available for the general population, how do you manage the safety of those who elect not to be immunized? The vaccine data released to date shows that the vaccines are nearly perfect at preventing serious COVID related disease, but still permit infection to occur in a significant subset (5-10%). This means some vaccinated people can get infected and possibly still transmit infection. They would be protected from serious disease, but the person they infect may not.
Last cut for the day. Wes Montgomery covering the Beatles “A Day in the Life.”
Over the last week, some parts of the country are beginning to succumb to the realization that shutting down is their last option to try to tame this surge. As always, this is not an easy decision to make because of the collateral damage it causes. That being said, there has to be a pain point by which you cannot stand by and watch. My pain point passed. We will be seeing 5,000 deaths every day soon due to where we stand today. You ok with that? 10,000 each day make you queasy? It won’t take much more crowding in the shops prepping for a holiday gathering. I hope your pain points are reached before large gatherings in a few weeks. If not, I sincerely hope you don’t need any basic type of emergency care or hospitalization. There won’t be a bed for you. Seriously folks, stop messing around, the ending to this story is getting worse every day. There is very promising hope for an end to this over a couple of months. You made it this far already. Finish the marathon. Please.
Post authored by Jason Halegoua PhD, MD, MBA, FAAP. Jason is the founder of Peds First Pediatrics in 2009, and has been a practicing general pediatrician since completing residency at Schneider Children’s Hospital in 2004. In addition to earning his medical degree from the Medical College of Pennsylvania, Jason earned a PhD in Molecular Pathobiology for his work contributing to the understanding of the genetic regulation of immune responses to murine leukemia viruses from Hahnemann University in Philadelphia and an MBA in Finance from Hofstra University.