For many of us, the coronavirus has added something to the daily routine. We get up, make coffee, and then check how many new COVID-19 cases there are. Was yesterday better than the day before? This week better than last week? Are we flattening the curve? Is there a second surge?

But those daily counts may not mean what we think they mean, and the answers to the questions were all interested in are much more nuanced.

Last week, Dr. Daniel Podolsky, president of UT Southwestern Medical Center, told us there is wide variance in the amount of time between someone contracting COVID-19 and the case being reported. Sometimes the lag is a day or two. Sometimes its more than two weeks. Whats more, theres no way to tell how many of each days new cases were recently contracted and how many are weeks old. Because of that variance, day-to-day reports of new cases dont accurately reflect real-time public health dynamics.

Podolsky suggested another metric for measuring the viruss spread: ICU beds. On Monday, 25 Dallas hospitals reported 3,447 of 5,713 beds occupied. Thats just over 60%, which is down from a peak of nearly 70% last week but still higher than last month.

Still another metric, and one that would tell us exactly how successful weve been at flattening the curve, is something called an R0 score (pronounced R naught). The R0 is a measure of how many people are being infected by each infected person. If the R0 score is 1, then every infected person infects exactly one more, and the outbreak will plateau. If the R0 is greater than 1, then the outbreak is growing with each cycle, or what experts call serial interval. Thats the average time between each successive infection. Taken together, the R0 and the serial interval can predict how quickly an outbreak will grow or recede.

Without precautions like social distancing, doctors say, the coronavirus will have an R0 as high as 3 (thats very contagious; more contagious than flu or SARS1) and a serial interval of about four days. So based on those numbers, if there were one infected person in Dallas on March 1, then on March 5, there would have been four infected people (the original case plus the three he infected). On March 9, there would have been 13 infected, and on March 13, there would have been 40 infected. The number grows exponentially with each successive generation so that by the middle of April every person in North Texas would have been infected.

Of course, that is not what happened. The virus has not spread at that rate. And, whats more, Texas aggressive measures to slow the spread have spared us from the fate of hot spots like New York, according to Dr. James Cutrell, the director of the infectious disease fellowship program at UT Southwestern.

But as useful as R0 is in theory, it is frustratingly hard to determine. The actual R0 of the coronavirus in Dallas is affected by myriad unknown variables like weather, ZIP code (some neighborhoods are naturally more isolated than others), public policy and human behavior. The number is such a murky composite of medicine, mathematics and social science that Dallas County Health and Human Services doesnt report it.

But others are willing to guess. The founders of Instagram have coded a resource called rt.live that reports estimated R0 values for each state in the U.S., and though its not an academic or scientific resource, it is based on data from The Atlantics COVID Tracking Project and is used by institutions such as the White House and Johns Hopkins University. As of May 5, rt.live put Texas score at 0.85, down from a high of 1.2 in March, meaning that if it is right, the outbreak is receding.

All of this is fascinating guesswork but doesnt give us a lot of confidence in those daily reports of new cases. And that, said Cutrell, is the point. He compared day-to-day or even week-to-week case counts to stock market day trading: When you chase daily numbers you can miss the long-term trend. He said we should look at longer time horizons when judging our success at flattening the curve. Which means that our evaluations of public policy changes should also be tempered.

Last week, Dallas County reported more cases than any week since the outbreak began, and that on the heels of Gov. Greg Abbotts relaxing shelter-at-home rules. The temptation is to assume that the latter caused the former. And while more interpersonal interactions will certainly lead to more infection, Cutrell said its too simple to couple numbers from last week with policy from the week before.

There are two lessons to learn from all this. First, as World Health Organizations emergencies chief Dr. Michael Ryan said Monday, without adequate testing and contact tracing measures we are essentially shutting our eyes and trying to drive though this blind. But second, and more comforting, perhaps we can take a break from fretting over up-to-the-minute reports on this pandemic. As we said Monday, the psychological trauma of this time cannot be ignored. Theres no reason to feed that trauma with daily reports, so wed encourage putting the day-to-day numbers in context of the longer trend and reviewing other data points to see where we are headed.

View original post here:
Daily COVID-19 reported cases dont say what we think they say - The Dallas Morning News