Please give an account of what happened including roles of persons involved, dates and times:

The statistical news release relating to the publication of the weekly ‘Deaths involving COVID-19′ data on
Wednesday 7 October 2020 was mistakenly live on the NRS website prior to the scheduled publication time.

NRS have been publishing the COVID-19 death statistics on a weekly basis from early April 2020.  In August we moved to a more comprehensive monthly publication, while still making the headline weekly
numbers available via the Scottish Government (SG) COVID-19 dashboard. However, the figures for this
week (week 40, 28 September to 4 October) showed an increase that we considered warranted providing
additional awareness to users.

A proportionate and quick solution was to do a statistical news release that briefly explained this week’s rise and highlight again to users where the data was available (SG dashboard). We had previously made users aware that the weekly data would be made available on the SG dashboard, Given the tight turnaround time, we did not pre-announce the fact that NRS would issue a statistical news release to supplement the publication of the data on the SG dashboard.

This news release was supposed to be published at 12 noon on 7 October 2020, however due to a
misunderstanding between the NRS webmasters office and our statistical web team it was published around 5pm on 6 October 2020.

NRS were alerted to this error around 10.40 am 7 October 2020, by a Scottish Government statistician who was part of the Pre-Release Access (PRA) list. The web team were notified and the news release was
removed from the site shortly before 10.50am. It was then published at the scheduled time of release at 12 noon. The SG dashboard was updated at the usual time of 12 noon.

NRS statisticians alerted all those on our PRA list immediately of this issue so that they were prepared for any potential early use of the unpublished figures, including communications colleagues perhaps having to deal with any unexpected requests or queries. We also advised our colleagues to continue to treat the statistics as unpublished and make no comment on them prior to the official scheduled publication at 12 noon.

Impact of the breach

Provide details of the impact of the breach both inside the producer body and externally:

NRS have investigated this issue, and although the web analytics show that 47 different users accessed the news release while it was online before its scheduled publication time, NRS believe that the impact has been very minimal and are not aware of any of the figures or commentary from the news release having been used in the public domain prior to release at 12 noon.

Following a review of news outlets and social media, there was no evidence of any early reporting of the new weekly total for COVID-19 deaths in Scotland. We also received no customer requests and no media
requests relating to these figures ahead of publication.

Corrective actions (taken or planned) to prevent re-occurrence

Describe the short-term actions made to redress the situation and the longer-term changes to procedures:

We have held discussions with the statistics web team and colleagues in the NRS webmasters team (IT) to review what happened, and have updated the web publication process guidance to ensure there is clarity on the process for issuing statistical news releases.

The decision to publish a statistical news release was taken at fairly short notice in relatively unusual
circumstances and this was a major factor in the miscommunication. We were looking to provide additional context and awareness to users, at a time when the key trend was changing after a sustained period of very low numbers.

The relevant colleagues in the NRS statistical and web teams have reviewed the web approvals process to
ensure clarity and awareness of the process for quality assurance checks made on the timings of publications and associated outputs (like statistical news releases).

NRS continue to monitor the COVID-19 death numbers closely and are putting plans together to be ready to reinstate more comprehensive weekly reporting in the near future. This would also mitigate against any potential for miscommunication over publication of outputs.