FOI REF: FOI-2024-1922
You asked
Please clarify:
a) If the population is living longer, then why should that factor have a bearing on excess deaths? Put another way, isn't it a contradiction in terms?
b) Why, if I have understood correctly, the new method for calculating was applied retrospectively, but not prior to 2020? If 2020 was felt to be an oddity, then surely just remove 2020 from the stats or qualify them?
I would like sight of any emails from other government departments that led to the ONS' decision to initiate the review in the first place. Just to clarify, I am wondering what led the ONS to decide to start the 14 month analysis programme and keen to see any external influence that may have triggered this.
We said
Thank you for your request.
In response to point A, the likelihood of death is highly related to age, with the vast majority of deaths occurring in older people. Therefore, all else being equal, a population with a greater share of older people will tend to experience more deaths in any given week, month or year than a population with a smaller share of older people. This needs to be taken into account when estimating the number of expected deaths in the week, month or year.
To address point B, we have applied the new methodology back to 2011. The estimates can be found in Section 4 of our methodology article from February 2024.
Regarding how we account for periods affected by the direct mortality impacts of the COVID-19 pandemic: to avoid these periods affecting estimates of expected deaths in subsequent periods, they are removed from the dataset when the model is fitted so that they do not contribute to the mortality baseline. This means that estimates of excess deaths in subsequent periods relate to the additional deaths registered in the period, over and above what would be expected from previous periods had they not been extraordinarily affected by the pandemic.
We define periods extraordinarily affected by the direct mortality impacts of the pandemic as being those where COVID-19 was given as the underlying cause of death for at least 15% of all deaths registered in the period across the UK. This threshold gives the greatest coherence between the weekly and monthly data in terms of periods excluded from the model fitting. These periods are April and May 2020, and November 2020 to February 2021 for monthly data; they are Weeks 14 to 22 of 2020, and Week 45 of 2020 to Week 8 of 2021 for weekly data.
Note that these COVID-affected periods are only removed from the data when we fit the models to estimate excess deaths in subsequent periods. We still estimate and publish excess deaths for the COVID-affected periods, using data from previous periods to fit the models.
The 14-month analysis programme referenced was triggered by the need for a unified baseline for excess deaths when looking at how many deaths we would expect in an average year. This requirement was noticed following internal discussions of the differences between ONS’s and Office for Health Improvement and Disparities (OHID)’s methods. A unified methodology would give a clearer indication on the number of excess deaths across the UK for users, noting that there will still need to be more than one method for differing questions (e.g. excess deaths measured by EuroMoMo method will still be needed as this is looking to answer a different question to ours).
In accordance with the UK Statistics Authority’s independence and best practice, the decision was made without external influence. Other government departments were engaged following the decision that a review of the methodology was necessary.