Answering FAQs: Part 3

What can we tell from international comparisons? What is the ‘real’ death toll?

To help the Royal Statistical Society, I wrote answers to frequently asked questions about COVID-19.

Given my frequent posts on this topic, the answers were similar to words already written. The main constraint was a limit of 400 words.

What can we tell from international comparisons?

International comparisons are challenging.

There is no standard definition of a ‘COVID-19 death’. Countries count these figures in different ways.

Before mid-August, there were different definitions in the United Kingdom:

  • England (Public Health England): confirmed deaths in all settings. The person has a positive test result for SARS-CoV-2.
  • Wales (Public Health Wales): deaths in Welsh hospitals and care homes. The deceased person must have tested positive for the virus. A clinician must suspect the COVID-19 disease was a causative factor in the death.
  • Scotland (Public Health Scotland): confirmed deaths in all places. The person must have died within 28 days of their first positive test result. The Public Health Agency in Northern Ireland also uses this definition.

After a review by Public Health England, the changed measure uses the 28-day cut-off.

RIVM in the Netherlands counts “overleden COVID-19 patiënten” (deceased COVID-19 patients). This measure counts deaths in hospital with a positive test result. Sciensano in Belgium includes deaths where doctors suspect the deceased has COVID-19.

Countries can change definitions too. NHS England used to provide the figure for England. That was a count of only confirmed deaths in hospitals.

Interpretations are also difficult. Countries have different testing regimes. Processed tests affect the number of confirmed deaths. Tests can differ: false negative results reduce lab-confirmed deaths.

Travel influences seeds and outbreaks. How many people live close together affects how this virus spreads. Demography, cultures, and health policies differ.

Despite these challenges, countries can learn from each other during this pandemic.

We should avoid precise league tables, and think more in broad tiers of countries.

Our World in Data highlights three ‘success stories’: Vietnam, Germany, and South Korea. The broad conclusion is effective responses need strong action in four areas. These actions cover: prevention, detection, containment, and treatment.

What is the real number of people who have died because of COVID?

It is uncertain.

There are three main ways to count COVID-19 deaths:

  • Confirmed deaths: deaths with a positive test result for SARS-CoV-2;
  • Death certificates: deaths with a medical certificate which mentions COVID-19. Certificates mention a disease either as a cause or contributory factor.
  • Excess deaths: comparing the number of deaths from all causes, to a baseline.

Each measure has different strengths and limitations.

Confirmed deaths

A confirmed death means a death with a positive test result for SARS-CoV-2 . In most countries, this is the daily figure that health departments publish. There are differences in precise definitions between nations.

The main strength of this measure is timeliness.

As it depends on positive tests, this count reflects testing availability and accuracy. This count excludes deaths with COVID-19 that had no test or only a false negative result.

Death certificates

Statistical offices publish death registration figures. These registrations have clinical judgements about how people died.

A Medical Certificate of Cause of Death has two parts:

  • I. The sequence of diseases or conditions that led to the death;
  • II. Other significant factors, which contributed to the death.

Clinicians must certify causes of death “to the best of their knowledge and belief”.

This count is of certificates that mention COVID-19. That means the disease was a cause or contributing factor to the death. Doctors may suspect the deceased person has COVID-19, without a positive test result.

The main strength is the inclusion of suspected deaths. Statistical offices can provide analysis on the underlying causes of death.

One limitation is timeliness: it takes time to register and process these certificates. Clinical suspicions can be wrong.

Excess deaths

Excess deaths is the difference between deaths from all causes and a baseline value. For the Office for National Statistics, that baseline is the average of the past five years.

Image for post
Image for post
This graph is by date of registration, rather than date of death. (Image: Office for National Statistics)

This measure has many strengths:

  • Consistency: we can calculate excess deaths in the same way, across nations.
  • Reporting: this measure is not influenced by reporting practices. Confirmed deaths rely on tests done. Doctors may be hesitant to mention COVID-19 without lab-confirmation.
  • Completeness: pandemics pressurise healthcare services. There may be deaths resulting from disruption to healthcare, or other negative effects.

Interpretation is challenging. It is a statistical construct, dependent on choice of the baseline. Excess deaths are not direct deaths from COVID-19.

This blog looks at the use of statistics in Britain and beyond. It is written by RSS Statistical Ambassador and Chartered Statistician @anthonybmasters.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store