{"id":2520,"date":"2020-07-25T18:20:59","date_gmt":"2020-07-25T17:20:59","guid":{"rendered":"https:\/\/marriott-stats.com\/nigels-blog\/?p=2520"},"modified":"2020-07-25T22:41:21","modified_gmt":"2020-07-25T21:41:21","slug":"covid19-deaths-latest-data-england","status":"publish","type":"post","link":"https:\/\/marriott-stats.com\/nigels-blog\/covid19-deaths-latest-data-england\/","title":{"rendered":"COVID19 Deaths #1 – Latest Data and Trends for England"},"content":{"rendered":"

Last updated on 25th July 2020 – future updates will be infrequent.<\/em><\/p>\n

The latest data for deaths due to COVID19 (Coronavirus) in England as of Friday 24th July 2020 show that the first wave of the pandemic is now over when one looks as excess deaths.\u00a0 People will still be dying of COVID19 for weeks yet but the overall number of excess deaths is now negative.<\/p>\n

<\/p>\n

The 6 time series for COVID19 related deaths in England<\/span><\/strong><\/h4>\n

\"\"<\/p>\n

For latest trends in number of COVID19 infections in England, please click here.<\/a><\/em><\/p>\n

Each time series for deaths is denoted with a 4 letter code which I will use throughout.\u00a0 Clicking on the 4 letter code will take you to the source data.\u00a0 I have only extracted data for England from these sources but some also cover Scotland, Wales & Northern Ireland.<\/p>\n

    \n
  1. PHEr<\/span><\/a> – Public Health England COVID19 Registrations – <\/span><\/b>Daily number of deaths by date of registration with COVID19 on the death certificate and confirmed with a positive test for COVID19<\/span><\/span>.\u00a0 Published everyday, this is the most common headline figure.<\/span><\/li>\n
  2. NHSo<\/strong><\/span><\/a> – NHS England COVID19 Occurrences<\/strong><\/span> – Daily number of deaths by date of occurrence with COVID19 on the death certificate.\u00a0 This is also published daily.<\/li>\n
  3. CQCn<\/span><\/strong><\/a> – Care Quality Commission COVID19 Notifications<\/strong> <\/span>\u00a0– All care home are required to notify the CQC of any death in their home within a short period.\u00a0 Since the outbreak, care homes are now able to say if they suspect the death was COVID19 related without a test.\u00a0 The data is passed onto the ONS who published the data weekly.<\/li>\n
  4. ONSr<\/a>\u00a0<\/span><\/strong>– ONS COVID19 Registrations<\/strong><\/span> – Daily number of deaths by date of registration with COVID19 on the death certificate from all locations.\u00a0 This is published weekly on a Tuesday but the daily data can be found on the COVID19-ENGLAND tab of the downloaded spreadsheet.<\/span><\/span><\/li>\n
  5. ONSo<\/a>\u00a0<\/span><\/strong>– ONS COVID19 Occurrences<\/strong><\/span>\u00a0– Daily number of deaths by date of occurrence with COVID19 on the death certificate from all locations.\u00a0 This is published weekly on a Tuesday but the daily data can be found on the COVID19-ENGLAND tab of the downloaded spreadsheet.\u00a0 Note two columns are shown with different cutoff dates and I take the data from the column with the latest cutoff date.<\/span><\/span><\/li>\n
  6. ONSx<\/a>\u00a0<\/strong>– ONS Excess Death Registrations<\/strong><\/span> – Daily number of deaths by date of registration with COVID19 on the death certificate from all locations.\u00a0 This is published weekly on a Tuesday and can be extracted from the WEEKLY DATA tab of the downloaded spreadsheet.\u00a0 I use the day of week pattern of the ONSr series to convert the ONSx weekly data into ONSx daily data.<\/span><\/span><\/li>\n<\/ol>\n

    To summarise, 4 time series (NHSo, ONSo, ONSr, PHEr) measure deaths where COVID19 is stated somewhere on the death certificate with PHEr requiring a positive COVID19 test in addition.\u00a0 The other two time series (ONSx & CQCn) do not require COVID19 to be mentioned on the death certificate.<\/p>\n

    The latest data and trends for all time series are shown below using the same chart format.\u00a0 The format is explained in the first section PHEr but all charts use the ONS definition of a week running from Saturday to Friday.\u00a0 All charts also begin with the week commencing Saturday 14th March 2020 which is recorded by the ONS as week number 12 which makes the latest week ending 24th July week number 30.\u00a0 Where necessary, the same week number convention is used throughout this post.<\/p>\n

     <\/p>\n

    1 – PHEr – Public Health England COVID19 Registrations<\/strong><\/span><\/h4>\n

    IMPORTANT – A significant discrepancy in the PHEr series has been uncovered which could create confusion.\u00a0 In this link, I explain why the daily PHEr figure is on average 42 too high<\/a>.<\/em><\/p>\n

    As of 24th July 2020, a total of 40,213 death registrations in England have been recorded in the PHEr series.\u00a0 The breakdown by day is shown in the chart below as bars.<\/p>\n

    \"\"<\/p>\n

    Note on 23rd May, a change was made to the way data is collected for this series.\u00a0 I explain this change at the end of this section but it causes a kink in the data on that weekend.<\/p>\n

    A solid line showing the 7 day Centred Moving Average (CMA) is also shown.\u00a0 A 7 day moving average is used since a week has 7 days and registrations are affected by weekend & bank holiday working patterns.\u00a0 The moving average is centred which means that an average for day T is the average number of daily deaths recorded between day T-3 and day T+3.\u00a0 CMA’s are the best way of spotting turning points in the data and the chart shows that the CMA for PHEr peaked on 9th April 2020 and turned downwards on 11th April 2020.<\/p>\n

    A dashed purple line shows a simple extrapolation model I have used to project the 7 day CMA into the future.\u00a0 A number of steps are needed to get to this point.\u00a0 First, I calculate the daily growth rate<\/strong> for each day T as follows:-<\/p>\n

    Growth Rate (Geometric) = dcPHEr(T) = Log[ cPHEr (T)\/cPHEr(T-1) ]<\/strong><\/p>\n

    where cPHEr(T)<\/strong> is the total number of death registrations as of day T.\u00a0 The data is plotted as diamonds on the chart with a scale that looks like a percentage scale.\u00a0 That is because this formulation of the growth rate will give more or less the same answer as the standard arithmetic % growth rate<\/p>\n

    Growth Rate (Arithmetic) = %DcPHEr(T) = [ cPHEr (T)\/cPHEr(T-1) ] – 1<\/strong><\/p>\n

    when the growth rate is around 10% or less.\u00a0 It starts to diverge above 10% but I use the convention of treating geometric growth as a percentage.<\/p>\n

    The advantage of geometric growth rates is that they can be properly averaged over time whereas arithmetic growth rates can’t be averaged.\u00a0 For example, if you have a quantity Q that starts at 100, grows by 20% the next day and then falls 20% the day after, the average growth rate is 0% but you don’t end up back at 100 but 96 instead.\u00a0 Geometric rates get around this issue so +20% geometric followed by -20% geometric get you back to where you started i.e. 100.<\/p>\n

    This is why I can calculate and plot a dashed black line showing the 7 day CMA of the geometric growth rate.\u00a0 From the end of March, this fell fairly rapidly but not in a straight line.\u00a0 Instead the 7 day CMA is curving which is indicative of cumulative growth that gradually slows down.\u00a0 It is this 7 day CMA of the geometric growth rate that I have extrapolated and results in the solid black added to the chart below.<\/p>\n

    \"\"<\/p>\n

    It is then straightforward arithmetic to convert that solid black line into an estimate of the 7 day CMA of the daily number deaths as shown by the dashed purple line.<\/p>\n

    The extrapolation itself is a straightforward statistical model which is built automatically in my spreadsheet.\u00a0 \"\"First I have to identify the time period I want to use to build my model.\u00a0 For this time series I’ve used the period 17th June up to the present day.\u00a0 I denote the 17th June as day 0, 18th June as day 1, etc and then plot this scatterplot with day number on the horizontal axis.\u00a0 On the vertical axis, I plot the logarithm of the 7 CMA of the growth rate.<\/p>\n

    If you’ve followed the calculations so far, you will spot the vertical axis is in fact a log of logarithm but it does result in a near straight line fit in the chart.\u00a0 Those who know about these things will spot there is some autocorrelation of the residual errors in the chart but I have not bothered to include this in my formula since I just wanted something that was simple to code automatically in a spreadsheet.\u00a0 With this model, it is then straightforward to apply it to day numbers beyond the end of the data set used to build the model and give the black line shown on the chart above.\u00a0 The 95% confidence interval (not prediction interval) for the extrapolated black line is +\/-0.1%.<\/p>\n

    Change of definition on 23rd May<\/span><\/strong><\/em> – From this date, PHE started to include deaths with a confirmed COVID19 test where the test was undertaken from a definition known as Pillar 2.\u00a0 I won’t explain what this means but when the change was made, PHE did not go back and revise the historical data beforehand.\u00a0 Instead, the additional 445 deaths that had occurred before 23rd May were simply added to the data for the week 23-29 May.\u00a0 The effect was to introduce an artificial trend that distorted the scatter plot above which meant I had to adjust for this effect.\u00a0 This is why my extrapolation model begins well after 23rd May i.e. when the centred moving average is no longer distorted by this effect.<\/p>\n

     <\/p>\n

    2 – NHSo – NHS England COVID19 Occurrences<\/strong><\/span><\/h4>\n

    As of 24th July 2020, a total of 29,117 deaths in English hospitals have been recorded in the NHSo series.\u00a0 The breakdown by day of death is shown in the chart below as bars.<\/p>\n

    \"\"<\/p>\n

    The format of the chart and the extrapolation into the future uses the same format and method as for the PHEr chart in section 1 above.\u00a0 The major difference with the NHSo series is that the historical data is continually updated as deaths registered on dates in the future are allocated back to the date death occurred.\u00a0 On average, death occurrence takes place 5 days before death registration but it can be both quicker and slower than that .\u00a0 Some deaths don’t get registered until a month after occurrence, perhaps because a complicated autopsy or coroner investigation, which then results in data over a month old being revised upwards.<\/p>\n

    Because of this, the last 5 days of the NHSo series is ignored when calculating trends and extrapolating the 7 day CMA into the future.\u00a0 If you’ve followed my Twitter feed, you will know that from the middle of April, I realised the 7 day CMA in the geometric growth rate stabilised if you did this.\u00a0 The scale of deaths can still change but the trend is stable and at this point in the pandemic, it is the trend that is most important.\u00a0 Certainly the peak in daily number of death occurrences in hospitals is clearly marked on 8th April 2020.<\/p>\n

    The 95% confidence interval in the extrapolated 7 day CMA of daily deaths is +\/-0.1% but this does not take into account future revisions to the historical data.\u00a0 You might notice from the chart that the extrapolated dashed line does not appear to join with the known 7 day CMA.\u00a0 It is because of the anticipated future historical revisions that explains this.\u00a0 My current method of allowing for this is a crude one and I haven’t as yet tested it properly but it seems to be semi-reasonable for now.<\/p>\n

    A final point to note.\u00a0 From the 24th April, NHSo started to include suspected COVID19 deaths in NHS commissioned services where no COVID19 test was carried out.\u00a0 Between 24th April and 24th July, a total of 1,209 such deaths in England have been recorded.\u00a0 Over the same period a total of 10,086 confirmed COVID19 deaths occurred in the NHSo series.\u00a0 The additional suspected deaths are not included in NHSo but if they were, they would increase the death toll by 12% though this percentage has increased to 20% in recent weeks as shown in the chart below.\u00a0 This is almost certainly due to the much lower number of daily deaths inducing volatility in this ratio.<\/p>\n

    \"\"<\/p>\n

    It should be noted that these additional NHSo deaths are included in the ONSr & ONSo series about to be discussed in sections 4 & 5 below.<\/p>\n

     <\/p>\n

    3 – CQCn – Care Quality Commission COVID19 Notifications<\/strong><\/span><\/h4>\n

    Since 10th April 2020, the Care Quality Commission has been collecting daily notifications of deaths in care homes where COVID19 was suspected to be the cause of death.\u00a0 It has always been a requirement for care homes to notify the CQC as soon as possible, if a death occurs but up to the end of March, it would appear doctors were not putting COVID19 on the death certificate.\u00a0 The ONS tracks deaths by location in their excess deaths data (ONSx) and any spike in care homes would be detected there but care home deaths were not prominent in the ONSr & ONSo series to begin with due to the lack of registration.\u00a0 Since April, this has changed but the CQC decided to start tracking notifications where COVID19 was suspected even if the doctor did not put COVID19 on the death certificate.\u00a0 That decision generated the CQCn series shown below which although is daily, it is in fact published weekly every Tuesday by the ONS rather than CQC.<\/p>\n

    \"\"<\/p>\n

    One point to be aware of is the potential for overlap with NHSo when looking at CQC data.\u00a0 Care homes are required to notify the CQC if any of their residents die regardless of the location of the death.\u00a0 That means CQC also record deaths outside of care homes but who were residents of care homes.\u00a0 Given the age of care home residents it should be no surprise that some will die in hospital and will be recorded in NHSo if the death is thought to include COVID19.\u00a0 The data shown in the chart is for deaths of care home residents in care homes & unstated locations only i.e. not including deaths in hospitals & elsewhere (such as outside) so this potential duplication is avoided here.<\/p>\n

    Since data was not collected before 10th April, we do not know the true cumulative count which means the underlying growth rate cannot be calculated.\u00a0 This explains why you see no diamonds on the chart.\u00a0 The 7 day CMA for daily notifications can be calculated and this shows a peak of 18th April.\u00a0 A visual inspection of the chart though suggests that the Easter bank holidays might have distorted the moving average and that the real peak was a few days later.<\/p>\n

    Before I proceed to section 4, I just want to point out that in section 7, I will show that the data in the CQCn chart is confirmed by the ONSr time series for Care Homes only.\u00a0 In other words, the CQCn series is not giving a different picture to other time series for care homes.<\/p>\n

     <\/p>\n

    4 – ONSr – Office of National Statistics COVID19 Registrations<\/strong><\/span><\/h4>\n

    As of 10th July 2020, a total of 48,388 death registrations involving COVID19 from all locations in England have been recorded in the ONSr series.\u00a0 The breakdown by day is shown in the chart below as bars.<\/p>\n

    \"\"<\/p>\n

    Again the chart format and extrapolation is exactly the same as for PHEr.\u00a0 Since this a series of death registrations, weekend working patterns have an effect.\u00a0 Compared to PHEr, the weekend and bank holiday effect is far more pronounced for the ONS who appear to work standard office working hours.\u00a0 The 7 day CMA of both the daily deaths and the geometric growth rate is therefore less stable hence why I have removed the daily diamonds from the chart to make it easier to read.\u00a0 I have still extrapolated the time series but this time the 95% confidence interval for the extrapolated 7 day CMA of the growth rate is +\/-0.2% which is higher than that of the PHEr series.<\/p>\n

    Unlike PHEr which has a peak date of 11th April, ONSr daily deaths peak is 17th April.\u00a0 ONSr includes deaths from all locations and it become clear that deaths in care homes were on a different timeline to hospitals.<\/p>\n

     <\/p>\n

    5 – ONSo – Office of National Statistics COVID19 Occurrences<\/strong><\/span><\/h4>\n

    As of 10th July 2020, a total of 48,532 death occurrences involving COVID19 from all locations in England have been recorded in the ONSo series.\u00a0 The breakdown by day of death is shown in the chart below as bars.<\/p>\n

    \"\"<\/p>\n

    As for the NHSo time series, the historical data is continually revised upwards due to future registrations.\u00a0 This time, the trend is stabilised if you ignore the last 4 days which are shown by the checkered green bars.\u00a0 The peak for ONSo is 11th April compared to 7th April for NHSo.\u00a0 As for NHSo, the 7 day CMA does not join up with the actual CMA since future historical revisions are anticipated.<\/p>\n

    I need to point out one wrinkle if you are reading the ONS spreadsheet itself.\u00a0 They provide two ONSo series.\u00a0 The first one only includes the total number of death registrations recorded in the ONSr series above in section 4.\u00a0 As of 10th July, there were 48,388 registrations which are then reallocated to date of death occurrence.\u00a0 However, the publication date for this time series was 21st July June and ONS were already receiving registrations in the week of 11th-17th July but had not as yet compiled the full dataset.\u00a0 Many of those registrations can be allocated to a date of death occurrence for dates up to 10th July and it is this data series that is shown in the chart above.<\/p>\n

    It is instructive to plot both ONSr & ONSo on the same chart.\u00a0 This gives a visual feel of the time lags between death occurrence and death registrations.\u00a0 The current gap between the peaks is 7 days but that is probably larger than it appears due to Easter falling in between the peaks which would have made some registrations later than usual and thus widening the time lag between the peaks.\u00a0 Prior to the peak, occurrences outnumbered registrations but since Easter, on average, registrations have been 20% higher than occurrences.\u00a0 This is a function of the time lag between occurrences and registrations and does not indicate any errors in either series.<\/p>\n

    \"\"<\/p>\n

     <\/p>\n

    6 – ONSx – Office of National Statistics Excess Death Registrations<\/strong><\/span><\/h4>\n

    If you are not familiar with the concept of excess deaths, this article by Anthony Masters<\/a>, an ambassador of the Royal Statistical Society, gives an excellent explanation.<\/p>\n

    I am using Excess Deaths as a proxy for all direct and indirect effects of COVID19 hence why the chart title still includes the word COVID19.\u00a0 Unlike the other 5 time series, ONSx can be negative since Excess Deaths is derived from this chart as the difference between total number of death registrations per week and a baseline number of death registrations.\u00a0 This is what happened in the latest week ending 19th June when excess deaths were -88.\u00a0 I have previously stated that I would regard the first wave of the COVID19 pandemic as being over once excess deaths went negative.\u00a0 Using this definition, the first wave lasted 13 weeks from 14th March to 12th June and resulted in 57,467 excess deaths in England.<\/p>\n

    \"\"<\/p>\n

    ONSx is published as a weekly time series but since the data is death registrations like ONSr, I have assumed that pattern of registration by day of week is identical to ONSr.\u00a0 That allows me to generate an ONSx daily series as below.\u00a0 The extrapolation method I used in the other time series cannot be used if the variable being predicted can be negative which is the case for ONSx hence why no extrapolation is shown.<\/p>\n

    \"\"<\/p>\n

     <\/p>\n

    7. Do the time series show similar trends?<\/span><\/h4>\n

    The chart below plots the 7 day CMA of the geometric growth rates for the 5 time series presented in this post.\u00a0 Death registration series (PHEr, ONSr & ONSx) are plotted as dashed lines whilst death occurrences series (NHSo & ONSo) are plotted as solid lines.<\/p>\n

    \"\"<\/p>\n

    ONSr used to lag PHE by a few days due to deaths in locations outside hospitals, such as care homes, following a later trend than hospitals.\u00a0 The series converged in mid-May but have now diverged again because of the one-off adjustment made to the PHEr series on May 23rd as explained in section 1.\u00a0 \u00a0On 25th July I analysed this discrepancy in more detail and concluded the PHEr figures have been overstated on average by 42 per day since 23rd May<\/a>.<\/p>\n

    The two occurrence series are now much closer together after being apart by a few days reflecting different timings of deaths in and out of hospitals.<\/p>\n

    The two series that are published every day, PHEr & NHSo, were 5 days apart but this has widened noticeably over the last few weeks.\u00a0 Again this must be due to change of definition for PHEr at the end of May.<\/p>\n

    To finish off, I want to compare the CQCn series with ONSr & ONSx data for care homes only.\u00a0 Both ONSx (Chart 2A below) and ONSr (chart 2B below) can be broken down by location of death.\u00a0 4 locations are used, Home, Hospital, Care Home & Other.\u00a0 The difference between ONSx & ONSr is denoted as ONSn (chart 2C below) and represents the number of excess deaths without COVID19 on the death certificate.<\/p>\n

    \"\"<\/p>\n

    In the middle chart 2B, I have added the CQCn line for care home residents.\u00a0 As you can see it is a near perfect fit to\"\"\u00a0the ONSr line for care homes as verified by the table shown here.<\/p>\n

     <\/p>\n

    – More posts about COVID19 –<\/strong><\/span><\/h4>\n
      \n
    1. A very useful guidance to interpreting statistics of COVID19<\/a> published by the Royal Statistical Society.<\/li>\n
    2. My collection of links to all kinds of material<\/a> related to the statistics of COVID19, epidemiological modelling and testing.<\/li>\n
    3. How large a sample is needed<\/a> in order to decide whether COVID19 restrictions can be lifted?\u00a0 A lot, lot less than you think!<\/li>\n
    4. How many excess deaths will there be as of 19th June?<\/a> This is my estimate of excess deaths using a statistical model.<\/li>\n
    5. Latest trends in number of cases of COVID19 in England<\/a><\/li>\n
    6. Are Public Health England’s COVID19 Death Counts misleading?<\/a><\/li>\n<\/ol>\n","protected":false},"excerpt":{"rendered":"

      Last updated on 25th July 2020 – future updates will be infrequent. The latest data for deaths due to COVID19 (Coronavirus) in England as of Friday 24th July 2020 show that the first wave of the pandemic is now over when one looks as excess deaths.\u00a0 People will still be dying of COVID19 for weeks […]<\/p>\n","protected":false},"author":3,"featured_media":2997,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_genesis_hide_title":false,"_genesis_hide_breadcrumbs":false,"_genesis_hide_singular_image":false,"_genesis_hide_footer_widgets":false,"_genesis_custom_body_class":"","_genesis_custom_post_class":"","_genesis_layout":"","footnotes":""},"categories":[6],"tags":[164,163,184,185,180,179,169,182,40,168,46,183],"class_list":{"0":"post-2520","1":"post","2":"type-post","3":"status-publish","4":"format-standard","5":"has-post-thumbnail","7":"category-forecasting","8":"tag-coronavirus","9":"tag-covid19","10":"tag-cqc","11":"tag-england","12":"tag-nhs","13":"tag-ons","14":"tag-pandemic","15":"tag-phe","16":"tag-presenting-data","17":"tag-sars-cov-2","18":"tag-trend-analysis","19":"tag-trend-extrapolation","20":"entry","21":"override"},"_links":{"self":[{"href":"https:\/\/marriott-stats.com\/nigels-blog\/wp-json\/wp\/v2\/posts\/2520","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/marriott-stats.com\/nigels-blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/marriott-stats.com\/nigels-blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/marriott-stats.com\/nigels-blog\/wp-json\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/marriott-stats.com\/nigels-blog\/wp-json\/wp\/v2\/comments?post=2520"}],"version-history":[{"count":35,"href":"https:\/\/marriott-stats.com\/nigels-blog\/wp-json\/wp\/v2\/posts\/2520\/revisions"}],"predecessor-version":[{"id":3010,"href":"https:\/\/marriott-stats.com\/nigels-blog\/wp-json\/wp\/v2\/posts\/2520\/revisions\/3010"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/marriott-stats.com\/nigels-blog\/wp-json\/wp\/v2\/media\/2997"}],"wp:attachment":[{"href":"https:\/\/marriott-stats.com\/nigels-blog\/wp-json\/wp\/v2\/media?parent=2520"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/marriott-stats.com\/nigels-blog\/wp-json\/wp\/v2\/categories?post=2520"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/marriott-stats.com\/nigels-blog\/wp-json\/wp\/v2\/tags?post=2520"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}