The infectious diseases in an ONS graph are not the ones you might expect

By | March 18, 2021

The Office for National Statistics (ONS) published a graph on 15 March 2021, which claimed to show deaths due to “infectious and parasitic diseases” in England and Wales from 1901 to 2020, along with the deaths due to Covid-19 in 2020. 

The title of the graph stated that “COVID-19 caused more deaths in 2020 than other infectious diseases caused for over a century”, and a line in the article containing the graph states that “COVID-19 was the underlying cause of more deaths in 2020 than any other infectious and parasitic diseases had caused in any year since 1918”. 

While this may be true, the graph is misleading because deaths from some infectious diseases—most notably influenza (or flu) and pneumonia—were not included. At the same time, conditions such as sepsis, which is something that can result from a severe infection rather than a condition that you can catch, was included.

The graph has been widely reported in media outlets, including the Daily Mail and the Mirror, with no context or explanation of which conditions have or haven’t been included. The graph has also been questioned on Twitter.

 

What does this graph show?

The ONS graph has included deaths based on the specific codes given to different causes of death as recorded on a person’s death certificate. These codes come from the International Statistical Classification of Diseases and Related Health Problems 10 (ICD 10).

For its data on deaths from “infectious and parasitic diseases”, the graph uses deaths coded in the first chapter of ICD 10, called “Certain Infectious and Parasitic Diseases”. This includes diseases which are “recognized as communicable or transmissible”, and are recorded using codes A00-B99. This is mentioned in the notes below the graph but isn’t made clear in the headline or subheading, and these notes haven’t been incorporated when the graph has been more widely shared.

Crucially, as the name of the chapter suggests, these are only certain infectious and parasitic diseases. They do not include certain “localised infections”, such as influenza and other acute respiratory infections (codes J00-22). (In the newer version of the coding system—ICD 11—influenza has been included in the first chapter.)

The graph therefore misses some of the spikes in deaths that would be caused by particularly bad years for influenza. It is difficult to know how many influenza deaths this might involve, but it is thought that an outbreak of pandemic influenza in 1969-70, for example, may have caused 30,000-40,000 excess deaths in each wave. 

In 2019, a total of 1,223 deaths were recorded in England and Wales with influenza as the underlying cause. This may underestimate the full impact of the disease, however, which is not always identified by testing, and which appears to be at least “associated” with thousands more deaths each winter.

At the same time, the graph may also overestimate the deaths from other “infectious and parasitic diseases” by including some that might not fit many people’s understanding of the phrase. For instance, about half of the deaths in the ONS graph are due to some form of sepsis, yet, as the NHS says: “you cannot catch sepsis from another person”.

The ONS told Full Fact: “The decision was made to focus on chapter level as this is a group of codes that have been combined by experts at WHO into a group.

“We didn’t include subsections of other chapters as we would then be effectively making the decision as to what we thought was relevant to be grouped together, which is something ONS does not tend to do.”

Overall, while the graph gives an accurate picture of the numbers included in the ONS data, without a clear caption and explanation it could give readers a mistaken impression of what the pattern of infectious diseases has looked like over the last century or so. It could also give a misleading impression of what is encompassed in the term “infectious and parasitic diseases”.

Source