Guidance

Reproduction number (R) and growth rate: statement of voluntary compliance with the Code of Practice for Statistics

Updated 1 April 2022

Introduction

The UK Health Security Agency (UKHSA) publishes the reproduction number (R) and the growth rate of coronavirus (SARS-CoV-2) for England and the 7 NHS England regions.

UKHSA is committed to applying the 3 core principles (referred to as ‘pillars’) of the Code of Practice for Statistics, on a voluntary basis, in its publication of the R value and the growth rate. These pillars are:

1. Trustworthiness: confidence in the people and organisations that produce statistics and data

2. Quality: data and methods that produce assured statistics

3. Value: statistics that support society’s need for information

The following describes how these principles have been applied in a proportionate way.

Trustworthiness

R value and growth rate estimates are produced by experienced academic and civil service scientists and statisticians who specialise in epidemiology. Both experts and independent figures are included in the approval process for the estimates, to ensure they are impartial and objective. The Epidemiology Modelling Review Group (EMRG), which sits within the UK Health Security Agency (UKHSA), challenges, considers, and approves these estimates with oversight from the DHSC Head of Profession for Statistics before public release. To aid transparency, we have published the method, including the approval process, in this methodology document.

Before 1 April 2022, the government published this statistical release every Friday at 1:30pm. From 1 April 2022, the statistical release occurs fortnightly, on every other Friday at 1:30pm. The 1:30pm release time differs to the conventional release time of 9.30am due to the time needed for quality assurance and approval post-production.

Going forward we aim to formally pre-announce the publication date to ensure that all users are aware of it. If the estimates cannot be published or publication is delayed, this will be announced on the R value and growth rate publication page as soon as possible.

To support the government’s response to the coronavirus (COVID-19) pandemic, it may be necessary to release the publication earlier to facilitate emergency press releases or public announcements. Changes in publication dates and timing may also be needed to accommodate bank holidays. When this happens, it will be clearly flagged on the publication page. Due to its use in the pandemic response and the method of construction, a wide range of policymakers see the estimates before publication. Policymakers, however, have no influence over the estimates.

Any revisions to the estimates take place in line with DHSC’s revision policy. Unscheduled corrections will take place as soon as practicable and will be prominently displayed alongside the relevant figures on the publication webpage or time series document as appropriate. As the R values and growth rates are contemporary estimates, calculated using the data available at the time, there are no scheduled revisions of the estimates.

Quality

The methods used to calculate the R value and growth rate are available in the methodology document. Briefly, several academic groups have produced individual estimates using different model structures, approaches and data streams, including:

• hospital admissions
• intensive care unit (ICU) admissions
• death data
• testing data

These models are reviewed regularly to ensure they use the most appropriate methods and data streams. Several models are now routinely run on the UKHSA’s quality assured systems. Any changes are included on the publication webpage and the methodology document is updated accordingly. The individual estimates are then combined into a single estimate.

The models are peer reviewed and have gone through the Rapid Assistance in Modelling the Pandemic (RAMP) initiative. EMRG provide challenge and quality assurance to the individual and combined estimates using the procedure outlined in the methodology. This process also supports the R value and growth rate estimates for Scotland and Wales, as well as England.

Due to the changeable nature of the COVID-19 situation, it is vital that modellers use the most appropriate and current tools available. Due to the timeliness of the publication of the estimate (one day after approval), it cannot be guaranteed that every method will be published at the time of release. The methods document will be updated at the earliest opportunity to ensure it accurately reflects the methods used.

The estimates are calculated using the most recent data available. The limitations of the data (for example, some data are time lagged by 2 to 3 weeks) are clearly explained and their impact on the estimates within the publication.

Commentary within the publication gives definitions and explains what the estimates mean in practice. The uncertainty of the estimates is given, explaining why a range is given instead of a single number. Less reliable estimates are highlighted when they do not meet the criteria to ensure appropriate use. The commentary is under constant review to ensure it is relevant and accurate.

A time series of all historic estimates is available for the user on the publication page.

Value

Over the course of the COVID-19 pandemic, the R value has become a frequently used statistic to help the UK understand the transmission of the disease. Collaboration with other departments and organisations ensures the R value and growth rate estimates are used correctly when informing policy decisions, supporting an accurate and consistent narrative around COVID-19.

To ensure that the R value and growth rate estimates are accessible to the public, commentary is included that simply explains what the statistics mean, provides definitions of specialised terms, and sets out the limitations of the statistics. The commentary also gives the statistics context and explains how they relate to other COVID-19 statistics. Links to other key COVID-19 statistics from other government organisations are also provided.

The publication has evolved over time to include new content, such as:

• national and regional breakdowns
• links to other statistics
• more detailed commentary

Government began publishing the R value from 29 May 2020 and the growth rate from 12 June 2020 on a weekly basis. These estimates have wide uncertainty ranges and producing them more frequently may not show meaningful change and could be potentially confusing, while any less frequently could be uninformative, given the potential pace of change of the epidemic. User feedback will be sought to continuously improve the publication.

Key user feedback will also be considered so that estimates are communicated effectively to decision-makers. Depending on the future course of the epidemic, use of the R value and growth rate will continue to be reviewed.

The statistical release is published in an open format and a historic time series of the estimates is available on the publication page to users to access previously released data.