Forum Topic

Covid-19 death statistics

If, as I occasionally do, you delve into the lower (in both senses) of the comments on the Daily Telegraph website, you'll read many people claiming that all this coronavirus stuff is a fuss about nothing, that these deaths would have occurred anyway and that the whole country is being shut down for no reason at all.This sad lie is exposed in the latest statistics from ONS (http://tinyurl.com/uxw9puwx); the total number of deaths registered in the UK in the week ending 3 April was 16,387; this was the highestweekly total since they started compiling weekly deaths data in 2005; the average of the corresponding week over the past five years was 10,305.

Richard Carter ● 2401d21 Comments

It’s a bit ironic that comments of this sort are coming from some of the most at risk groups. What is the average age of a Telegraph reader by the way?  Older or younger than the average Tory Party member?

Jonathan Callaway ● 2401d

Apparently the average age of Telegraph readers is 61 (but I imagine that's the print version, the internet readership is probably only 60). The comparable figure for Tory party members is 72.The Mail average age is 58, the Times 54 for the print version, 50 for print and online combines, and for the Guardian it's 49.

Richard Carter ● 2401d

The other thing about the figures released today is that they show that the announced daily number of Covid-19 deaths is lower than the real figure because many deaths (how many is uncertain) are not included in the daily figure. OTOH, the daily figures are presumably consistently measured, so they give a real indication of the trajectory over time.

Richard Carter ● 2401d

Richard. My understanding from the daily briefing today is that the figures being given at the briefings are from hospital returns and that's because the hospital figures, although there may sometimes be delays in reporting some deaths if things are busy, are accurate and easily collected and, as you say, give a good trend analysis. The ONS figures which are compiled at a more leisurely pace can include other deaths. The problem is care home deaths cannot be collected in a timely manner to include them in the daily statistics as many care homes still use mainly paper based systems

Michael Ixer ● 2401d

Michael,It is not even mentioned at the briefings that these are only deaths from hospital patients who have tested positive!If they wanted to they could get the ONS daily.It must be deeply distressing for families of care home to think their loved ones lives are not even worth reporting daily never mind the care home testing, ppe

Ed Robinson ● 2401d

I think its mentioned everyday that these are  hospital deaths only reported in a daily basis. Not sure of the logistics of collecting those outside of hospitals is as straightforward as we would like on a timely basis

Andy Pike ● 2401d

Ed. I'm not going to argue with you that the care system is severely underfunded, disjointed and has been treated appallingly during the COVID-19 pandemic. However, because of  that I don't think there is the infrastructure or mechanisms in place to reliably collect accurate death statistics on a daily basis. I'm not saying that's right, just being realistic. My mother was in a care home until her death just over a year ago and while I can't fault the staff or the care they gave may mother I'd say their administration systems were rudimentary and time consuming, and staff struggled to keep records up to date. I don't know, but my experience of comparing labour intensive manual system with well well designed IT systems would suggest the care home records contained some inaccuracies. I'm also not going to defend any of the government ministers and experts for obfuscating the detail in their answers at the briefings, but I feel it's probably better to have reasonably accurate and consistent metrics from hospitals to use for trending.  Yes, I agree it is distressing for families knowing the way their relatives' fates seem to be being ignored and the care services deprived of adequate equipment and testing.

Michael Ixer ● 2401d

The death figures being announced every day are next to meaningless. They exclude care home deaths, that's true. But in addition to that, they are not the deaths which have occurred that day, but the deaths which are reported. These deaths have actually occurred on different days, mainly over the course of the previous week.So all the TV news discussion of "the trend", is utter twaddle. There is no trend to be observed, because it would not be capable of being seen in those daily "totals".The only way to observe any trend is to look only at data a few days old, giving sufficient chance for reports to come in and then look at the accumulation of reported deaths for successive days.As of last week it was still going up. The incubation period is around a week, then it takes, for the unlucky ones, a couple more weeks to die. The effects of the lockdown we are currently engaged in therefore cannot be seen in these numbers. Yet.But do keep at it - it's the right thing to do

Michael Winstanley ● 2401d

Andy - "I think its mentioned everyday that these are  hospital deaths only reported in a daily basis."I played back on iplayer :Today: YesMonday: NoSun: Not clearSat: No

Ed Robinson ● 2401d

Michael W - I completely agree.Michael I - Deaths are all recorded in the UK and have done for centuries - are you telling me a country that has just placed the population under house arrest and cut GDP by a 1/3 can't get the recorded deaths per day?Not everything needs to be automated!

Ed Robinson ● 2401d

Ed. I'm not saying it can't be done, I'm just saying it isn't being done and to my knowledge the mechanisms haven't been put in place to do it. Having worked it IT for nearly five decades I know how much data collected is inaccurate and incomplete, so the situation doesn't surprise me. True, not everything has too be automated but the easier one makes it to collect data the more likely it is to be done correctly. Unfortunately, as Michael W says, there are delays in some data collection (which sort of proves my point) but I guess we have to use what we've got. I'm not defending the position, I'm just being realistic. At least I didn't vote for those in this and previous governments who introduced austerity and voted against health workers pay rises.

Michael Ixer ● 2401d

Michael it can be done - do not confuse process with IT.It's easy to automate once a process is in place but first comes the process.A new process could be set up immediately if the will were there.We are talking about registered deaths which are already recorded but the frequency of collation of data is not currently daily. The sources are already known - not unusual to have daily or even real time reporting with more accurate daily/weekly figures following.I've only worked in IT for less than 40 years.

Ed Robinson ● 2401d

Ed, maybe I probably didn't phrase that to well as I needed sleep. Deaths are  currently registered but that can be any period ip to 5 days in England so I can't see that helping with consistent, next day statistics? If Mr Google is correct there are 11,300 care homes for elderly people with 5,500 different operators. Yes, one could set up a simple manual process, perhaps all care homes phone in figures the next morning? Interesting rollout and operational logistics problem, I think? I'm not sure how well it will be followed in short staffed establishments ... Unfortunately, the lack of testing is also likely to lead to inaccuracies in the cause of death? Hence, figures might still be unreliable. Let's be honest, for whatever reason, the government's on the back foot with this one. PPE and testing for care homes is the  main priority (not to mention more money to fix staffing issues?) while the process to collect and collate next day statistics is being worked out. In the mean time hospital figures will give the basic trend. But I'm not disagreeing with you: it's a mess, just not sure there's an easy fix because, sadly, the issue of care home staff shortages is unlikely to be rectified overnight.

Michael Ixer ● 2401d

Michael you need to think more agile.(Using phone so will be brief)We have reported daily hospital figures where date of death may lag for a few days. We just need the same for non hospital. These can be got when registering. Like hospital data the actual deaths on a given day will be provisional due to the lag.This would give a more complete picture than what is being reported.At the moment the ONS data is ever two weeks.Now either the government is doing this and not telling us or the are going to await the complete ONS data. I'm not sure which of the two is worse.

Ed Robinson ● 2401d

Ed. I'm not sure what your mechanism is for collecting data from 11,000+ care homes who's staff are already stretched, and who have different admin systems?I think we agree:The government should be collecting data daily from care homes as well as hospitals;It isn't currently doing that;It should implement a process to do that asap but figures should be as accurate as possible (but will need to be adjusted retrospectively);It is currently replying on ONS data taken from death certs to get complete datawhich by definition of the process has a time lag - you get up to 5 days to register and then there's a delay collecting and collating data from each local registrar (must admit I'm not too familiar with that)I do think the government, for whatever reason, obfuscated the fact that the figures they were presenting were only from hospitals.Proper agile development is good, using it as an excuse for not doing things properly is bad 🙂

Michael Ixer ● 2401d

Ed. Interesting. Heard on R4 this morning the is already a process.  All care homes have to report deaths to the CQC and that's noe been modified to add where deaths are confirmed COVID or suspected COVID. Didn't catch reporting frequency/time lag. No excuse for inaccurate metrics?

Michael Ixer ● 2397d

Besides the horrific number of deaths I would like to know how many people have actually had/are suffering from Covid19 so we can compare with annual flu rates. But the cases can never be accurately totalled because patients have not been seen by a GP/been tested to verify they have/had the  virus.I am also concerned that some members of the Cabinet wish to lift restrictions sooner rather than later because of concerns about the economy. Surely ensuring that people are safe from contracting Covid19 comes above worrying about the economy?

Lucille Grant ● 2397d

What I dont understand is why so many people who fall ill in care homes are not being taken to hospital. They are care homes not hospices so why dont they warrant the same level of medical care as someone who falls ill at home?

Adam Tripp ● 2397d

Let us await the success or otherwise of the Republican states in the US who are trying to get out of lockdown earlier than expert opinion seems to support.

David Ainsworth ● 2397d

That would still exclude deaths that are not in care homes.Deaths registered would seem to be a logical place to get the data.As I've said they already have this data but are not sharing it!

Ed Robinson ● 2397d

To prevent, not limit the spread may be too late. Unless some one has a plan to share or even can find the governments plan then it will just be about trying to keep the curves from spiking too much.

Ed Robinson ● 2397d