End of March US COVID Update

The first full week of spring is in the books. This week’s data can probably be spun either way. Michigan continues to be a problem and there are 7 or 8 other states where the downward pattern has slowed or stopped. But we are still far away from November/December in terms of numbers of cases and hospitalizations.

Jumping in to the overall numbers, for the week ending on March 27th, there were 9.6 million tests recorded which is essentially flat from last week. Positive cases rose from 388,475 to 425,825 and the percentage of positive results rose from 4.0% to 4.4%. But both of those figures are lower than any full week in January, December or November and are similar to pre-surge numbers back in mid October. Naturally, we’d prefer numbers to continue to trend down but it is important to keep the raw numbers in context when discussing weekly increases.

The CDC website does have data on hospitalizations. The seven day average of new admissions (March 20 through March 26) was 4816 which was up from the 4621 in the previous seven days. But at peak, the average was 16,522 and the level now is back to mid October levels. They also look at trends in emergency department visits (not all states have data though) and the percentage of visits due to corona like illness or with a COVID-19 diagnostic code is under 3% and has been flat since late February.

The good news out of this is that we haven’t seen a real upturn in hospitalizations which would be expected during a surge. The bad news is that hospitalizations may lag cases a bit so it may be too early for it to show up.

Michigan is again the exception to the rule as the percentage of ED visits related to COVID-19 has been increasing. Some other states where things are not looking as good are Vermont, New Jersey, New York, Pennsylvania, New Hampshire, Connecticut and Massachusetts. Among the states that are in the best shape right now are Arizona, New Mexico, California, Oklahoma, Arkansas and Kansas. There’s some geography at play here.

In an attempt to quantify things, I’ve looked again at the total number of newly reported cases last week vs the peak total. I also look at a doubling rate which is just the number of cases divided by the average of the last ten days. In short, it calculates how long it will take a state to double the number of cases assuming that the number of new cases per day remains at the average. I arbitrarily assigned points based on where they were and then gave a bonus point to any state that saw an increase in cases from the previous week.

Enough words, let’s get to the graphs. The first two are California which I just mentioned and Texas which is also in great shape but just not quite as far off its peak as California. Both saw cases decrease the last week so Texas gets 1 point and California gets 0.

In this first group, we had five states see a increase but really one that looks like it might be going in the same direction as Michigan.

• New York (10 points) – While cases were up, New York gets points because it is only about half off the January peak and cases appear to have flattened so we still aren’t seeing things drop like other states. Because of the high average, New York is one of a few states where the doubling time is less than a year.
• Pennsylvania (9 points) – Big jump last week breaking above early February levels and now at 40% of peak value. Doubling time is also less than a year. A state to watch this week.
• Florida (5 points) – Small jump in cases last week but Florida ranks here because their current total is about 32% of peak which is among the highest so cases haven’t dropped as much here as in other states.
• Illinois (3 points) – While the total for last week was the highest since first week in February, it is still only 18% of the peak back in November. If it turns upward more sharply then it becomes a problem.
• Ohio (3 points) – Really similar to Illinois which a small increase in cases last week but still just at 16% of the peak.
• Georgia (2 points) – Cases dropped last week although they are still at 16% of peak. Still the trend continues to be down although at a slower rate.
• Arizona (0 points) – Cases dropped again last week and the total is now 5% of the peak. Plus the doubling time at the current rate is about 4.5 years so there is no problem here at all.

There were five states with increasing totals last week. Michigan and Massachusetts are the worst and Minnesota is not trending in the right direction but hasn’t really yet reached problematic levels.

• Michigan (14 points) – This last week is 70% of the peak back in November so they aren’t that far off of the peak number of cases observed in the fall surge. The graph continues to show big climbs and it really seems to be a new surge. The doubling time is 167 days which is almost the lowest in the country. It still seems odd that we’d have a single state surge but each week seems to bring more evidence that is what is going on.
• Massachusetts (7 points) – did see an increase last week and the current total is 34% of the peak but they haven’t broken back over early February levels. The doubling time is 296 days or less than a year so that is why they rank so high.
• Minnesota (4 points) – Cases have been on the rise here for several weeks in a row but it is a more gradual rise and they currently sit at 20% of the fall peak. The doubling time is still over a year so they rate a small level of concern as things are going in the wrong direction.
• Indiana (2 points) – Saw the first increase in a couple of weeks last week but it is still just 13% of peak and doubling time is over two years. Not yet a problem despite having Michigan on the border.
• North Carolina (2 points) – Cases were down again last week and are now 18% of peak.
• Tennessee (1 point) – Cases were down last week and are now at 11% of peak with a doubling time of over two years. Looks to be in good shape.
• Wisconsin (1 points) – While cases were up last week, the level is only 7% of peak and the doubling time is more than three years. So they only get the point because cases were up a bit last week. No problems here even surrounded by Michigan and Minnesota.

Another group where five of the states saw an increase from last week but only one seems to be a problem.

• New Jersey (12 points) – cases have increased for five weeks in a row and are now above early February levels. They currently are 69% of the peak and the doubling time is 206 days which is again one of the few states where that is less than a year. Still, the increase has not been as dramatic as what is being observed in Michigan and not as fast as the surge in the fall. A strange spike but it does make them a state to be concerned about.
• Colorado (4 points) – Cases were up a bit last week but have really settled into kind of a plateau. Right now, they are at 22% of peak so things haven’t dropped as much as other states. With the doubling time just over a year, they get a middle amount of points. But I like the graph better than Minnesota which also had four points.
• Virginia (4 points) – Very small increase in cases last week. The number is 26% of peak which is above average for the country and the doubling time is just over a year. Still, like Colorado, the shape of the curve is better than Minnesota.
• Iowa (3 points) – Saw an increase last week but the number is 13% of peak and they haven’t broken back above mid February levels so it really looks good here.
• South Carolina (3 points ) – Cases dropped a bit last week but they are still at 24% of peak and they won’t rank lower until cases drop more. Still things are moving in the right direction and South Carolina was one of the last states to peak so it may take longer to get to lower levels.
• Missouri (2 points ) – Cases up last week but they are 15% of the peak. Really similar to Iowa where cases haven’t broken back above mid February levels so things continue to be good here.
• Alabama (1 point) – Cases continue to drop and the current week is 10% of the peak. No problems here.

In this group, we had two states where cases went up last week. I wouldn’t call either one a problem right now but they are not trending in the right direction and need to be watched.

• Maryland (5 points) – Cases up again last week for the third week in a row and now they sit at 37% of the peak. The doubling time is right at a year so they rank just a tier below the problem states I mentioned at the beginning. Like New Jersey, the upward trend is kind of gradual but an upward trend is not where you want to be.
• Washington (4 points) – Cases also up last week and are now at 27% of peak. A notch behind Maryland in terms of concern so the points seem to fit here.
• Kentucky (2 points) – Cases continue to drop and are at 16% of peak with a doubling time approaching two years. They haven’t hit a bottom yet here but the trend is in the right direction.
• Louisiana (1 point) – Saw another drop last week and are sitting at 11% of peak with a doubling time of about three years. Haven’t hit a bottom yet but in really good shape.
• Utah (1 point) – Cases also down last week and at 13% of peak. Also in good shape.
• Arkansas (0 points) – Cases down again last week and are now at 6% of peak with a doubling time of over four years. Would expect things to start flattening here because you can’t drop to zero but there is no evidence of a problem here.
• Oklahoma (0 points) – Pretty much a match to Arkansas. Cases are 8% of the peak and the doubling time is about 3.5 years. Like Arkansas, they haven’t hit a bottom yet but I’d expect to see that. But there are no problems here.

Two states saw an increase in cases but there is really only one that is a concern.

• Connecticut (8 points) – Cases were up last week and the jump was pretty big although they haven’t broken beyond early February levels. Still the current number is 45% of peak. It is probably fair to point out that I don’t have data for April and states like Connecticut (and New York, New Jersey and Massachusetts) had big peaks then so they might look better if I had that data. Still, they never dropped as much as other states and now things here are headed in the wrong direction.
• Idaho (2 points) – Cases were down last week but they currently sit at 18% of peak which is part of the reason they get to 2 points. The graph and data don’t show any real problems though.
• Nevada (2 points ) – Cases were up slightly last week but are still just 11% of peak and the doubling time is about 3 years. A good place to be.
• Mississippi (1 point) – Cases dropped again last week and they currently site at 11% of peak with a doubling time of about 3 years. In very good shape.
• Nebraska (1 point) – Small drop in cases last week but they’ve been at a bottom since the week of February 20th. The current number is 12% of peak so another place where things look really good.
• Kansas (0 points) – I have to put an asterisk by this one. Cases have dropped sharply but they’ve also had several days where they didn’t report anything. Kansas has always been spotty on reporting so I don’t know if they’ve got a bolus built up. Still the raw numbers are great.
• New Mexico (0 points) – There was a small drop in cases and they are currently at 8% of peak. Also in very good shape.

Second to the last group and six of these states saw an increase. One is a problem with three others in a range to watch for a bit.

• New Hampshire (9 points) – Like a few other states, cases have been rising for a few weeks and they now sit at 40% of peak. They fall into the category where the increase is more gradual and not like the earlier surge. Still, the doubling time is less than a year and cases are moving in the wrong direction.
• Oregon (4 points) – While cases went up last week, they’ve been bouncing for four week and they haven’t broken back above late February levels. They currently sit at 22% of peak but the increase last week doesn’t really look like a problem.
• Rhode Island (4 points) – Did have the highest total in five weeks but they still didn’t break back above mid February levels. They currently sit at 29% of peak with a doubling time right at a year. Not going to make too much out of a one week increase but it is one to watch.
• West Virginia (4 points ) – Similar to New Hampshire with cases going up for several weeks in a row but the increase is kind of slow and gradual. They are at 27% of peak with a doubling time right at a year so they are a tier below New Hampshire in terms of concern. But not moving in the right direction.
• South Dakota (3 points) – A small increase last week and the highest total since the week of January 23. Still they are only at 13% of peak with a doubling time of 1.5 years so I’m not concerned about the small upward trend.
• Montana (2 points ) – The only state in this group where cases went down. They are at 13% of peak and seem to be in a bottom so no problems here.
• North Dakota (2 points ) – They got a point because cases did go up last week. Last South Dakota, this was the highest total since the week of January 16th. But they are just at 10% of peak and with a doubling time of over two years. Not yet a concern.

The final group. Again, six of the seven states in this group saw an increase from last week. One is a problem and one isn’t and the rest are in the middle but need to be watched to see where they go.

• Vermont (18 points) – they actually get more points than Michigan. The trend had been slightly downward but they saw a pretty big jump last week and are now at 95% of the peak. Vermont has had the most unusual pattern without a clear exponential phase or a big drop. They’ve been flat at what are high levels for them with a doubling time of 123 days. So they actually top the list in this ranking.
• Maine (7 points) – After slowly rising for a few weeks, cases dropped a bit last week and now sit at 32% of peak. Maybe Maine is the model for the other states that are seeing a slow rise. But the current level is high enough that their doubling time is 241 days or less than a year and they get a lot of points. Still a concern as we don’t want to overreact to a single data point. Another state to really watch next week.
• DC (6 points) – I know they aren’t a state but I include them here just to give the complete picture. They really didn’t have a high peak so they are currently at 42% of that peak and cases did rise last week although it didn’t really break out of the flat part they’ve been in. Would still say things are at a plateau and not really a big concern.
• Hawaii (6 points) – For a lot of reasons, Hawaii isn’t really in sync with the rest of the country. They didn’t go above the August peak this fall and, like Vermont, the overall case count is low. Still, cases have gone up for a few weeks in a row and they sit at 34% of peak with a doubling time right around a year. Wouldn’t say it is a big concern but just trending slightly in the wrong direction.
• Delaware (5 points) – While cases were up last week, they’ve been bouncing up and down for five weeks in a row. The current total is 30% of the peak and they doubling time is right around a year so they get a few points for each. It could be they are just going to plateau at a relatively high level for a bit but the trends here aren’t alarming.
• Alaska (4 points) – hard to see because they had basically the same number of cases last week as Vermont. For Alaska, it was the highest total since the week of January 23rd but you could still argue it is more in the plateau than a real upward trend. The current number is 22% of peak but we can’t make a real statement based on a single data point. We need to watch and see what happens.
• Wyoming (1 point) – A very small increase last week but they are still just at 8% of peak so it is more accurate to say they are in a bottom. The doubling time is about 2.5 years so no problems here.

Michigan remains the most obvious problem. Pennsylvania and Connecticut had big jumps but one point does not make a trend so more data is needed. We’ve got a few states showing increases but more of a slow gradual increase that needs to be monitored. And Vermont remains the most unusual in terms of the pattern but it is also true that cases just aren’t dropping there.

Overall, we had a slight increase in cases last week but are still at pre surge levels. Nowhere near enough evidence to say that there is a new surge going on. Wish we could figure out what is going on in Michigan though.