FOI Ref: FOI/2023/4817
You asked
I would like information on some major changes in the Public Health Data Asset (PHDA) unvaccinated 70-79 population in March 2021, from the two datasets that relate to the ONS publication 'Deaths by vaccination status England' of 6 July 2022.
Dataset 1: Deaths occurring between 1 January and 31 October 2021 edition of the dataset and;
Dataset 2: Death occurring between 1 January 2021 and 31 May 2022 edition of this dataset.
In dataset 1, in the tab of table 7 in cell E95 the Person Years figure for March 2021 for the unvaccinated 70-79 age group is 27,837.
In dataset 2, in the tab of table 2 in cell G131 the Person Years figure for March 2021 for the unvaccinated in the 70-79 age group is 14,554.
This implies a reduction of 47.7% of the (average) March 2021 unvaccinated population (since the average March population is approximately 365/31 x the Person Years figure) between these two publications (1- 14,554/27,837) when bar some minor data adjustments the unvaccinated in the two publications should essentially be the same people. At the same time the *total* PHDA population (unvaccinated and vaccinated) for the 70-79 age group for March 2021 stayed the same to within 0.1% as measured by the equivalent Person Years change
Please can you tell me what material changes the ONS made that resulted in this major unvaccinated population change that occurred between dataset 1 and dataset 2 (albeit most of the change appears to relate to changes from the 'to 31 October 2021' dataset and the 'to 31 December 2021' dataset)?
We said
Thank you for your request.
We frequently receive updated data from National Immunisation Managmenet System (NIMS), which includes updates to previously inputted vaccination records and new vaccination records that were not recorded previously. This can cause changes to the populations in each category in the publication. There will also be changes due to deaths being registered.
Looking specifically at the populations in the 70 to 79 age group in March 2021 in the different versions of the publication, there is a drop between the publication going up to October 2021 and that going up to December 2021.
In our December 2021 publication (the first which included third dose or booster doses), we omitted people who had a first and third dose or booster recorded, but not a second dose as we cannot correctly assign these people to the vaccination status groups. In this publication, this corresponded to 76,068 people, as mentioned in the measuring the data section of the release, and the equivalent numbers for subsequent publications are also presented in the respective measuring the data sections of the releases. These people are more likely to be elderly (due to having received a booster). Removing them reduces the populations of earlier vaccination status groups earlier in the year. It is likely that this contributes a large part to the drop in the populations that you observe here.
Please note, as the mortality rates are calculated from the same population, if there is a fall in the population due to the sample included, this will equivalently affect the populations and deaths so will not cause a large variation in the mortality rates.