Coronavirus: Latest Official Data Looking Hopeful - but the Numbers Don't Stack Up
All across Budapest, and I assume the country, government-sponsored posters, columns and advertising hoardings proclaim: "Let's look after each other - Every Life Counts". Sounds reassuring, certainly, and nobody doubts the hard-pressed healthcare staff are doing their very best in these difficult times. It's also true the latest statistics on the government website are encouraging, indicating that we are past the worst in daily infections, and that deaths have almost plateaued. However, data compiled by the Central Statistical Office raise serious suspicions that the government numbers fail to tell the whole story. But let's start with a summary of the week, according to the government website, https://koronavirus.gov.hu/#hirek. The headline news is that new infections detected are down by 7,500, or 25%, on the previous seven-day total, that is from just over 30,000 to 22,500. (Numbers rounded for simplicity. See summary for precise figures.) True, the total number of active infections is up to 197,500, 4,800 more (2.5%) on the previous week, so Hungary could well have 200,000 suffering from the virus over Christmas (a lot will depend on the numbers who recover). However patients being treated in hospital are down around 8% to just over 7,000. Deaths for the week attributed to the virus totalled 1,134, up by 37, or just 3.3% on the previous seven days.
This brought the total Covid-19 deceased past the 8,000 mark, and constitutes the second worst week on record, but as deaths inevitably lag infections, this increase is arguably only a reflection of the worst ever week for new cases a fortnight back. So far, so hopeful then. Except for the Central Statistical Office (KSH) and its compilation of deaths, which are published for each week of the year with about a month's time lag. (These are in Hungarian, but anyone can understand them with a little help.) https://www.ksh.hu/docs/hun/xstadat/xstadat_evkozi/e_wnh004f.html?down=4468 I began watching these in May, suspicious that deaths might have been misrecorded in the first wave of the pandemic. Astonishingly, the death numbers in April this year were actually lower than last year's, so nothing appeared untoward. However, things appeared to be going awry as the KSH October numbers came out, and with November, the divergence between what would be expected and reality as per the KSH has grown yet wider. Now there has been a lot of talk about 'excess deaths' in relation to Covid victims, but I have found that different people take this term to mean different things, so I shall avoid it here. Put simply, because of Covid-19 victims this year, you would expect more deaths, and you should be able to estimate the final numbers fairly accurately by taking the average deaths for a given period in recent years, and adding to this the deaths attributed to the pandemic over the same period this year. So, if we take the KSH deaths for weeks 44-47 (from the last week in October, to the third week in November – the dates vary slightly year by year) for 2017 – 2019, we find the average number of deaths is 9,578 for the four weeks, ie a little under 2,400 per week. In that same four-week period, the government figures for deaths attributable to Covid-19 this year is: Total deaths reported on November 23 – Total deaths reported on October 25. ie 3,891 – 1,472 = 2,419, (or 605 deaths per week). So, we should expect to find the total deaths for this year for weeks 44 – 47 should be Deaths in 'normal' year + Additional official Covid deaths that is: 9,578 + 2419 = 11,997. Over the four weeks, this equates to 3,000 per week. But according to KSH, actual deaths in weeks 44 – 47 this year total 13,954. This is way above what we would expect from the earlier estimation, the difference being 13,954 – 11,997 = 1,957 deaths. In fact, it is 489 additional deaths per week than would be expected. This means the actual deaths per week in weeks 44 -47 this year has been 3.489 rather than the 3,000 predicted using the government figures. While I was working on these numbers on Friday, website hvg.hu found the KSH numbers for the month of November. These not only back up my calculations, but indeed paint an ever worse picture of what we may call 'unexplained' deaths, as reported here. (Sorry, Hungarian only)
https://www.ksh.hu/docs/hun/xftp/gyor/nep/nep2011.html?utm_source=kshhu&utm_medium=banner&utm_campaign=theme-nepesseg-es-nepmozgalom In summary, there were 15,366 deaths in November this year, that is 5,279, or half as many again as last year. It marks the worst month for deaths in Hungary for more than two decades, since January 2000, when the country suffered from a serious flu epidemic. Hvg.hu calculates that there were more than 2,000 deaths in November than would be expected from the norm plus the government Covid figures (again, similar, though slightly worse than my own calculations, at roughly 500 unexplained extra deaths per week).
It puts this divergence down to a number of possible reasons, including Coronavirus victims who were not properly diagnosed, some who may have died as a result of not receiving proper treatment because of the pandemic along with the possibility of what the writer terms “a communication error in the epidemiological information”. As far as I can tell, the government has given no explanation for this additional 500+ weekly deaths so far.
It begs the question: Every life counts, but is every death accounted for? If we take weeks 44-47 or the entire month of November, the answer is simple: so far, it seems not. SUMMARY of Numbers according to the government website this morning. New infections this week = 22,589 (was 30,122 in previous week) Deaths this week = 1,134 (was 1,097)
Total since the start of the pandemic attributed to Coronavirus = 8,099
Total active cases of Covid-19 = 197,447 (was 192,683 in previous week)
Patients undergoing hospital treatment = 7,022 (was 7, 646)
Patients on respirators = 520 (was 606)