More than 30,000 Hungarians died of coronavirus - What about excess mortality?

New COVID-19 cases, serious cases, all kinds of exciting/alarming ratios

Over the past 24 hours, Hungarian authorities diagnosed 55 people with coronavirus infection, and 5 COVID-19 patients have died, taking the death toll to 30,004. The number of new daily cases has not been this high for two weeks.

An extremely low share (0.2%) of active cases are in hospital, but an increasing percentage of them require artificial ventilation. The 7-day average of the ratio of those on ventilators to the total number of COVID-19 patients in hospitals is up at nearly 22%.

Here are two other charts showing the same changes. Both are for a three-week period up to 7 July. The one on the left shows the number of people with coronavirus infection in hospital and on ventilator, and also their 3-day averages, which continue to show a decline. The left-hand chart shows the same ratio as the one above, only not a 7-day average but the daily readings and their 3-day averages. It shows that the ratio of those on Covid patients on ventilator to the number of those in hospital spiked to 24% and then corrected back to around 18%. The 3-day average also came lower but remains around 21%.

The calculations used for the following charts are based on 7-day averages.

The first chart shows ratios of 7-day averages: [the 7-day average of a given day – 7-day average of the previous day] / 7-day average of the previous day.

What we see is that the decline in the number of active has been gradually shrinking for about two weeks to around 0.5% by now, while the change in the number of hospitalised Covid patients shows a diminishing rate.

The following chart shows the ratios of the changes (curves) you see above. The blue line: orange/green, i.e. change in the number of hospitalised COVID-19 patients per change in active cases (with 7-day averages). Red line: blue/orange (on ventilator, 7-day average) / in hospital (7-day average). Wherever the ratios are over 100%, the numerator is larger than the denominator, i.e. the rate at which they discharge Covid patients from hospitals (including those that are reported as recovered by GPs) is higher than the rate at which the number of active cases is dropping. This means that an ever smaller percentage of active cases are in hospital. The red line dropped below 100% on 27 June and has stayed there ever since, hitting rock bottom at 8.4% on 5 July. 

Alarming change in test positivity

 The change in test positivity also gives reason for concern.

The 3/21-day average of test positivity has jumped to 129.61% from around 93%, and the leap did not occur even earlier because of an extremely low reading. However, the positivity rate was almost 1% already yesterday (0.95%), and the 21-day average is 0.67%. The latter will go even lower, which means we can say goodbye to the improvement in the positivity rate for some time.

The number of tests remains extremely low (5,765) and only 55 came back positive yesterday. Let's put this into perspective!

The last time 5,000 or so tests yielded an 0.95% positivity rate was on 20 August 2020 (5,245 tests, 52 positive). More than 55 new cases were registered on 25 August (73 new cases). Two days ago, authorities reported 21 new confirmed COVID-19 cases which we could compare to 5 August (33 cases). It took merely two days this time to achieve a similar rise in the number of new cases that Hungary had over a two-week period in August last year. The 3-day/21-day ratio is likely to stay north of 100% for a while.

And the same figures on a longer horizon, between 1 March and 7 July:

Excess mortality 60% over 2019 in April 2021

The spotlight remains on March and April 2021, as the peak of the third wave in the coronavirus pandemic, as well as on the first 20 weeks of the year, as the stats office has revised data up to the 20th week for three times already.

As you can see in the tables below, excess mortality was around 47% in March both over 2020 (47.7%) and 2019 (46.9), while the April figures turned out ever worse (as expected). Excess mortality on the 28 days between 29 March and 25 April was 60.2% over the same period of 2019, and nealry 50% over 2020. In March-April this year, nealry 49% more people died in Hungary than in the same two months of 2020 (+48.7%) and 2019 (+53.1%).

Excess mortality on the first 21 weeks was nearly 24%, KSH data show (see more detailed tables and charts below).

The Delta variant (B.1.617.2) keeps spreading at an increasing rate globally, while the vaccination campaign in Hungary, with a population of about 9.8 million fizzled out at around 5.5 million people vaccinated against SARS-CoV-2 with at least one dose.

Excess mortality on the first 20 weeks

The upper section of the following tables (green header) shows data for separate weeks: the number of deaths in 2021; the weekly differences over 2020, 2019 and the average of 2015-19; and the ratios of the differences where the annual difference (for 2020, 2019 and 2015-19) is divided by the total weekly deaths of the given (base) year, not 2021. (It would have made no sense to compare 2021 to 2016-2020 because last year’s data would cause quite a distortion due to the coronavirus pandemic. Hence the comparison to 2015-19.)

The lower section (blue header) shows accumulated figures, i.e. the second week shows deaths for the second AND first weeks, the third for the 3rd, 2nd and 1st, etc. This is why the total on week 13 is the same in both tables.

Excess mortality for weeks 12 and 13 were 62.5% and nearly 74%% compared to the same weeks of 2020, over 84% and 87% compared to the same weeks of 2019 and 73-74% compared to the 2015-2019 average of the same weeks. The stats office has revised data up to the 20th week three times. (In fact, data were revised 10 times for the 13th week, 9 times for the 14th week, 8 times for the 15th, etc.)

Excess mortality in the first 20 weeks (up to 23 May) reached 27.4% over the same period of 2020 and 24.8% in the first 23 weeks.

The following tables show revisions up to week 20 between 16 June and 7 July. As you can see, serious revisions are continued to be made for weeks in March and April, mostly April (13-15).

Here’s a more concise version of the above tables with a couple more charts to boot. (Click to enlarge.)

The sharpest revisions between 16 June and 7 July affected the latest figures (weeks 20, 19, 18), but major adjustments were also made for April (week 14, 15, 16).

Black March and April

Excess mortality data for weeks 9 to 16, i.e. from 1 March to 25 April, show that the period between mid-March and mid-April was the deadliest relative to the same weeks of 2020, 2019 and to the 2015-19 average. This is shown also on a chart.

EXCESS MORTALITY BETWEEN 1 MARCH AND 25 APRIL 2021 WAS 48.7% OVER 2020, 53.1% OVER 2019 AND 48.1% OVER THE 2015-19 AVERAGE.

In this period, nearly 31,192 people have died in 2021, against cc. 21,000 last year, 20, 370 in 2019 and 21,000 on average between 2015 and 2019.

65,596 deaths in the first 20 weeks, more than 72,000 in the first 23 weeks

The tables below contain all weekly KSH revisions since 27 April. The colouring is per row, except for the total revision and total revision ratio columns to show when the stats office carried out the sharpest revision for the given week.

The numbers of the revisions refer to the 13th week, i.e. its first revision of the 13th week mortality data was done on 5 May, the 2nd on 12 May, etc. As the red cells show you, the sharpest adjustments are generally made on the first, second or third revisions, but there are oddities too. For instance, mortality date for the 13th week was left unchanged on the fourth revision, but then upward adjustments were made to the tune of 45, 63, 64, and 62. This is strange but has nothing to do with the KSH, as it merely processes data provided by registry offices. Two 0 adjustments for the 17th week were also followed by upward revisions of 39 and 13. What’s more peculiar about this phenomenon is that registry offices report keep ‘finding’ a lot of dead people even three months after they had died (e.g. in late March, early April).

Mortality stats for the 13th week of the year were revised upwardly the most (by 565 people or 14.34% over ten revisions).

A total of 65,595 people died in Hungary in the first 20 weeks of 2021, and their number has been revised upwardly by 3,859 or 5.88% since 27 April.

Coronavirus has been the deadliest in Hungary for months, as the nation of 9.8 million people topped the global rankings in terms of Covid deaths per one million population. After an unexpected and drastic revision of mortality statistics in Peru (currently 5,794), Hungary is now second with 3,114 deaths per one million people.

Cover photo: Getty Images