ONS Death Data and the Disparity with PHE Daily Updates
July 14, 2020
Jason Oke, Carl Heneghan
The Office for National Statistics (ONS) today published deaths registered in England and Wales in the week ending 3rd July, 2020 (Week 27). Of 9,140 deaths registered in this week, 532 (6%) mentioned “novel coronavirus (COVID-19)”, on the death certificate, the lowest number of fatalities involving COVID-19 in the last 15 weeks.
The ONS also release deaths by date of occurrence, which allows us to calculate the change in the number of deaths occurring and chart the trend over time.
The figure shows that from a peak of 1,272 deaths on the 8th April the daily number of deaths has fallen considerably, with 64 occurring on the 30th June and 37 on the 1st July. Although these latter counts could be added to in the next week, the trend is still for fewer deaths albeit at a slower rate than seen in April and May.
What has become apparent in recent weeks is the growing disparity between the numbers released by ONS and those reported by Public Health England which are widely disseminated in the media.
The total number of COVID-19 deaths in England according to ONS for 3rd July is 48,154 whereas the equivalent total announced on the 4th by PHE was 39,626.
The PHE figures also vary substantially from day to day. For example, 16 new deaths were announced on 6th July, but the following day, 152 were reported.
This variation is most likely due to the appearance of ‘historic’ deaths that have occurred weeks before, but for some reason unknown to us, get reported in batches on particular days. To counter this variation, a moving average smooths the trend, but even this is at odds with the ONS data.
The PHE moving average for 30th June is 103, which is more than the ONS numbers for 30th June and 1st July combined.
The moving average is overlaid on to the ONS figures to show the PHE figures average has been consistently higher than ONS for some time.
A similar analysis of the English hospital data by date of occurrence shows 25 deaths for 30th June. As this is 60% of COVID deaths in hospitals, we can surmise that the total number of deaths in all settings is approximately 40 per day, much closer to the ONS numbers.
Because of the inaccuracies in PHE data we recommend using ONS data and the NHS England data to understand the trends in deaths over time. To reduce confusion we require all deaths reported by PHE to include when they occurred as opposed to the day of reporting.
Jason Oke is a Senior Statistician at the Nuffield Department of Primary Care Health Sciences and Module Coordinator for Statistical Computing with R and Stata (EBHC Med Stats), and Introduction to Statistics for Health Care Research (EBHC), as part of the Evidence-Based Health Care Programme.
Carl Heneghan is Professor of Evidence-Based Medicine and Director of Studies for the Evidence-Based Health Care Programmes (Full bio and disclosure statement here)
Disclaimer: the article has not been peer-reviewed; it should not replace individual clinical judgement, and the sources cited should be checked. The views expressed in this commentary represent the views of the authors and not necessarily those of the host institution, the NHS, the NIHR, or the Department of Health and Social Care. The views are not a substitute