TT, Tegnell, and the Lag in the Statistics

In Sweden, 373 people have so far been found dead as a result of the coronavirus. Revised official figures show that the number of deaths day by day has been higher than previously reported, according to TT’s review.

But the backlog is no longer expected to be as large, says state epidemiologist Anders Tegnell at the Swedish Public Health Agency.

– We are now driving the number of infected against the death register. There may be some lag, but not as big as it has been before, says state epidemiologist Anders Tegnell.

Can it really be at 500 or something like that?

– I have a very hard time believing that, We know that the last days are uncertain, but if we go back two or three days, the numbers will be very uncertain.

225 have been reported dead in Stockholm

According to state epidemiologist Anders Tegnell, a lag in the figures has previously been due to the Stockholm Region, among others, reporting deaths in covid-19 with a certain lag. A cumulative total 225 have died in Stockholm, according to yesterday’s figures from the region.

– About a week ago, Stockholm started with a new system to control who died in covid-19 by entering the death register, it was not done before, he says.

– But now Stockholm is very fast in finding deaths and we check all the time and send out questions to the county councils. So in recent days we have a much smaller backlog than we had before.

2020-04-05 16_54_09-CompatWindow

The most recent reported number of deaths from Sweden means that another 28 deaths has been added to the statistics the last 24 hours.  That does not mean that these people necessarily have died during the last 24 hours, as there is a certain lag in the statistics.

Wolf_WTF As I look at the numbers given by ECDC and by the FOHM, and read how Tegnell is explaining them, the more confused I get.  Either he know something that I do not know, or he is purposely downplaying the number of deaths.  Or… he does not know how to read his own numbers.

One would think that after the TT report  about the cumulative number being off by half, he would have gone back to the very same excel sheets FOHM publishes every day, to find what kind of lag there actually is.

..Unless… of course, he is in the same boat as the rest of us.  The report sheet is replaced every day on FOHM web site. There are no archives.  Could it actually be that FOHM is also replacing their own copy, so they have no record on what they have been reporting any day in the past.  That would be funny, if it were so. 

Regardless, I now believe that I cannot trust the way the numbers are being reported.

One clear example is the Lag and how it is being used.  As above, when numbers are being reported, it is pointed out how the number of deaths  the last 24 hours is smaller than reported, because of the Lag.  At the same time, we are not allowed to say, that the number of deaths today is probably greater than the actually reported number, because we are not allowed to apply the future Lag.  Consequently, the numbers reported are no indication at all.

I believe the most accurate ay to report for the public, would be simply to report the numbers as they come in.

BBC does that – simply gives the numbers reported.  On weekends, they do mention that the numbers reported on weekends are always lower, and are likely to raise again in the beginning of the week.  Simple, clear, and a good measure of he epidemic.  And even with the lag during the weekends, you can always compare this Sunday to last Sunday. Without any shifting to the past or to the future. 

At the moment, when the testing is not working, and we really have no way of measuring the epidemic, the most reliable way to estimate the number of infections about two weeks ago, is to look at the number of deaths reported in today.  Then it would not really matter so much, if the actual death took place two days or two weeks ago – the randomness of the Lag would even out the kinks. 

Finally: about “I think it’s about a dozen here or there.”

The proper way to answer that question is : “on day D1 after the report day, we generally have 50% of the deaths reported, on D2, 90%.”  If you are asked, and do not know, you do not guess.  Have one of your people find out and get back with a proper answer.   it is bad enough that hundreds of people had to die because you ppl ‘Thought it was going to be no worse than a common flu.’ I wish you people would have figured this out by now: 

There is a deadly virus spreading in the communities you are supposedly in charge of protecting. Whatever you happen to ‘think’ does not matter


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s