Daily COVID-19 Briefing: 4/16/20

Top news, reports and insights for today:

1. U.S. deaths hit an all-time high again on Wednesday
   Yesterday, an additional 2,438 COVID-19 deaths were reported in the U.S., a new daily high and a 9% increase. This continues a trend of continued escalation of deaths since April 7, albeit with a temporary 2-day decline on Sunday and Monday, which I believe was a lag in reporting due to Easter. Eight states tied or exceeded their previous daily high death totals including 5 in the Northeast (New Hampshire, Maine, Maryland, Massachussetts, and Connecticut), and 3 in the south (West Virginia, Virginia and Texas). Connecticut (+197) and Massachussetts (+151) were especially hard-hit on Wednesday with triple-digit deaths for the first time, suggesting that the epicenter is spreading from metropolitan New York to a broader area of the eastern seaboard. Substantial three-day rises in deaths occurred in Idaho (+51%), Wyoming (first 2 deaths reported), Ohio (+43%), North Carolina (+44%), West Virginia (+47%), Connecticut (+57%), DC (+44%), Massachussetts (+47%), and Maryland (+48%).
   What this means? Unfortunately, this suggests not only that we are not at peak but that daily deaths continue to rise. Yet, the news is full of reports of re-opening. While the rate of increase is slower than it had been through much of March, the pattern we see now suggests that community transmission remains robust but that we continue to flatten-the-curve. In my view, broad loosening of social distancing now will lead to steeper increases in deaths in the short term.

2. Maryland requires masks in public, cases surpass 10,000
   I thought I would check in on the situation in my own state (Maryland), which has been identified as a hot spot in recent days. The figure below shows the daily case curve since the first reported case on March 5. The pattern shows the epidemic took off around the 24th of March, with rapid growth of cases through April 10th, followed by a plateau in new cases of around 600 a day over the last week. It is not clear if this is a true flattening of cases or a peak in the state’s capacity to test. As of Wednesday, Maryland reports 349 deaths from confirmed COVID-19 cases, plus an additional 64 “probable” deaths. That suggests at least a 16% higher death toll compared to official reports and further supports the fact that testing is not keeping pace. Among deaths where age is known, 43% have occurred among those 80 and older, however notably 29 people under age 59 have died (0.3% of deaths). Maryland also reports 278 cases (3%) among children; 66% of cases are below age 60. Women outnumber men slightly among cases (54%) while more deaths were reported in men (52%). Just under half of the state’s cases (45%) were reported in the two counties closest to Washington DC (Montgomery and Prince George’s) and a quarter of the cases are in Baltimore city and county. However, numbers are rising faster outside these areas. African Americans represent a disproportionate percentage of deaths (46%) compared to whites (43%), however the crude fatality ratio (deaths / confirmed cases) is comparable (4.1% in blacks and 5.9% in whites).
   Where we are in testing? Based on data from the COVID Tracking Project, Maryland has done about 56,000 tests, with about 28% positive results. Since April 1, the state has been averaging 2,500 tests a day, but that number hasn’t increased, suggesting we are maxed out on capacity to test. Because 30% of tests are positive in recent days, it tells us we are testing quite selectively. Not until testing is more widespread and available to the general public, will it be clear how far the epidemic has penetrated. Overall, the crude fatality ratio is 4.1% in the state, which suggests that it is highly likely that Maryland has between 4 and 8 times as many infections as are currently known.

3. COVID-19 is filling in the middle of the U.S.
   Last night I was staring at the Johns Hopkins COVID-19 dashboard and thinking about the spatial pattern visible in the map. Then, I went back to older screen grabs of the same map and noticed something interesting (see the two maps below). The top image is from yesterday, the bottom from March 28. Most of us have been focused on big coastal cities. While the bubbles are not entirely comparable here, the big urban centers look quite similar in these two images. What really strikes me about this comparison is how pockets of infection have broken out in the nation’s middle in smaller cities and towns and rural communities. From March 28 to April 15, we went from 101,000 U.S. cases and 600,000 globally to 640,000 U.S. cases and over 2 million global cases. From Texas to North Dakota and Florida to Maine, the dispersion of the epidemic into new territory in all corners of the country is the big story. This has important implications for areas of the country with far fewer resources. It also reinforces the idea that no area can count on remaining untouched by this virus.

Advertisement

Daily COVID-19 Briefing: 4/11/20

Top news, reports and insights for today:

1. Flat growth in deaths does not signal a peak in the epidemic
   Social media, commentators and news outlets are hailing the flattening of growth in deaths as a victory. It is a victory in the sense that we are doing what we set out to do (albeit late): flatting the curve. But a flattening of growth is not the same as a peak as the two figures below show. The top figure (pink/red) is the one I post every couple of days. It shows daily new death reports. Growth in deaths has been flat for the past 4 days, averaging just under 2,000 per day. The lower figure (purple/blue) plots the cumulative number of deaths each day. That shows that the total death toll of the COVID-19 pandemic in the U.S. continues to rise linearly.
   What this means? Numerous pundits are now attacking the modelers, claiming that death projections were clearly wrong, that we have over-reacted and that undo harm has been done to the economy. It is certainly true that modeling and projections are constantly changing as new information becomes available. But what I see is that social distancing is starting to work; in many states, the curves have flattened. That does not mean that death estimates were wrong. We need to keep in mind just what the goal was. Flattening the curve is about slowing the pace of the outbreak to allow the health sector to keep up. That is what we are doing (although the pattern is uneven). The result is that we are prolonging the duration of the outbreak by slowing the transmission cycle. If we continue to see 2,000 deaths a day for 100 days, we will reach 220,000 deaths. Today, the U.S. becomes the nation with the most overall deaths. The peak in deaths won’t come until A) steady and consistent decrease in daily deaths and B) when the rate of recovery is greater than the rate of new infections.

2. Regional pattern of recent hotspots in new cases
   In early March, the pattern of growth in new cases was distinctively regional. In more recent weeks, region had grown less important. Now, I believe we are seeing bigger differences by regions once again. The chart below shows growth in new cases over the last 3 days by state and region. A total of 14 states and DC have seen growth of 30% or more in cases over that time. This includes only 1 of 13 western states (New Mexico), 4 of 13 mid-western states (Iowa, Illinois, Nebraska, and South Dakota), 4 of 13 southern states (Kentucky, Texas, Virginia, West Virginia), and 7 of 12 northeastern states (Connecticut, Washington DC, Delaware, Massachussetts, Maryland, Pennsylvania and Rhode Island). Candidates to become the next major hotspots include New Mexico, South Dakota, Kentucky, Texas, Maryland and Rhode Island. The northeast region continues to be the center of the epidemic in this country with smaller clusters of states heating up around the great lakes, Texas/NM, and much of the eastern seaboard.

Daily COVID-19 Briefing: 4/10/20

Top news, reports and insights for today:

1. U.S. COVID-19 cases now in flat but consistent growth; not to be confused with peak!
   Today’s epidemic curve for daily U.S. cases (below) continues a pattern starting about April 1 of consistent daily growth in total cases. That means about the same number of cases are being added each day. In essence, we have gone from exponential growth in the first 3 weeks of March (cases doubling every 2-3 days), to additive growth until April 1 (doubling every 7-14 days), to flat growth (consistent adding of cases with daily fluctuations around a steady or flat rate of increase). It’s vitally important that people understand that flat growth does not indicate a peak in the epidemic. A peak in the epidemic cycle will occur when two conditions are met: 1) there is a period of sustained decline in daily new cases, and 2) the rate at which patients are recovering is greater than the rate at which new cases are added. Neither of these conditions has been met.
   Why this matters? I see too many people making the leap from flat growth to a peak. A peak in the curves for daily cases is not the same as a peak for cumulative cases. If anything, the last five days have shown a return to gradual increases in daily cases. Several credible models I am monitoring suggest that in most states, we won’t see a peak in cases until anywhere from mid-May till early July.

2. Antibody testing: A game-changer or a potential debacle?
   Up till now, testing for COVID-19 has been done using a complex procedure that looks for telltale genetic fragments of the actual SARS-CoV-2 virus in a sample of fluid gathered on a Q-tip from someone’s nose or throat. That test tells us whether the virus is present in your body. It’s a complicated test that takes hours or days to run and it is far from perfectly accurate. A week ago, the FDA gave approval for a new kind of testing designed to do something different. The new approach looks for presence of antibodies in a sample of blood (that’s why it’s called serologic testing) to determine who has been exposed to the virus. This test can detect infection in people after symptoms have subsided when no active virus is present. More importantly, antibody testing can tell us who might be immune to the illness. This has tremendous potential to be a game-changer for surveillance (improving our ability to track the extent of the epidemic) and mitigation efforts (by determining who is protected). When widely available, this type of testing will be a boost to the health system by allowing us to determine which medical workers are safe to care for patients. The test may be done in the home and provides results in minutes. However, let’s not forget this is a fast-moving crisis and things rarely go exactly as planned. In approving these tests, the FDA has drastically lowered the bar on its evaluation and approval process. That means, lots of companies will rush products to market and some will not be accurate. Questions about accuracy will leave medical workers reluctant to abandon the need for PPE. Dr. Anthony Fauci recently said that antibody testing will be rolled out in a matter of days to weeks. They are already in use in China, Singapore and several other countries. The U.K. is planning to make them available for home use. It’s not yet known how well companies will be able to ramp up production, and distribution will be complicated as it has been with test kits and ventilators so far. Here are some important caveats to keep in mind as the nation gets ready to add this potentially powerful new tool to the fight against COVID-19:

• The test will require a blood sample. It’s not clear whether a finger-prick will be sufficient or whether a sample from a vein in the arm will be needed.
• Just because you detect antibodies, doesn’t mean you are immune.
• Antibody tests will be negative early in the infection. It takes time for the body to produce antibodies. The existing PCR test will still be better early in the disease.
• Some people don’t produce a strong antibody response to a virus, so the accuracy of the test will vary between people.
• Scientists don’t know how long immunity will last or whether re-infection is possible in those with antibodies.
• Developing an antibody test requires a lot of careful research. There hasn’t been time to fully vet the approach that is being relied upon.
• Quality control in the manufacturing of the test will not be optimal, so the accuracy will vary across manufacturer and batch.

3. A tale of two states: My own look at Kentucky vs. Tennessee and what it tells us
   An article by Dan Vergano in Buzzfeed news caught my eye on Wednesday. The article talks about how neighboring states Kentucky and Tennessee offer a natural experiment to compare 2 different responses to the COVID-19 outbreak. Attention to this comparison started with Kentucky resident Stephanie Jolly, whose graph showing the epidemic curves in the two states went viral last week. I decided to create and update my own version of the graph; hers plots cumulative cases, mine plots new cases by day. The graph shows daily lab-confirmed cases in the two states since March 10. As of yesterday, Kentucky reports 1,452 cases and 75 deaths, while Tennessee has tallied 4,634 cases and 94 deaths. Tennessee has a bigger population (6.8 vs 4.5 million) but even accounting for that, Kentucky has seen fewer cases and a flatter curve. The main story I take from this comparison is that early containment really mattered. Kentucky’s governor declared a state-wide emergency on March 6, while Tennessee waited till the 12th. Schools were closed in Kentucky on the 12th, Tennessee waited till the 20th. Restrictions on non-essential businesses and mass gatherings were a week earlier in Kentucky, and Tennessee didn’t finally issue a stay-at-home order until April 2. Basically, Kentucky moved a week to ten days faster to enact social distancing measures compared to Tennessee. It is speculation and not proof, but this early and decisive action by Kentucky governor Andy Beshear may have resulted in half the number of cases and a third fewer deaths. With the exception of April 8, new cases have been essentially flat in Kentucky since March 27. In Tennessee, dramatic rises in cases were seen in the last 2 weeks of March. As disease detectives, we will continue to monitor this and other state comparisons to gather clues about what is working and what is not.