Covid-19 contact tracing in
England and Wales
Dr Ian Brown
Tracingappv1
 Govt’s “NHSX” unit outsourced production of a
“centralized” contact tracing app, with limited
transparency and stakeholder feedback
 Data protection regulator provided detailed advice to
NHSX, but often felt ignored; later criticized for not
taking regulatory action
 Frequent expert and media criticism once Apple/Google
Exposure Notification API announced, becoming louder
as app release delayed for months
 As predicted, problems arose with keeping apps “awake”
– govt tried and failed to persuade Apple to change iOS
(as in France – bitter complaints about sovereignty)
 Large-scale trials showed many iPhone users were
therefore not sending/receiving notifications
 Govt created then scrapped ethics board after
sustained criticism
Tracingappv2
 Govt adopted A+G model and quickly
released app, with additional QR code
“check-in” functionality (10m downloads
in three days)
 Media discovered close to zero
notifications sent based on
hundreds of millions of check-ins
 NHSX forced to release update
reducing “risk sensitivity” due to
earlier configuration mistake
 November – govt announced
interoperability of apps from E+W,
Scotland, N Ireland, Gibraltar & Jersey,
but excluded from EU E-Health Network
Non-NHS
accessto
Covid-related
data
 Police given access to test and trace data to check those
ordered to isolate are doing so
 Public fears about police accessing tracing app data – but
use of Apple/Google API allowed a wide range of influencers
to explain this isn’t possible
 GCHQ/NCSC involvement in designing tracing app, and
broad access to NHS information systems
 Controversial involvement of Palantir to operate
infrastructure
 Controversial involvement of Serco to operate national
tracing service, with significantly lower success than local
public health units handling complex cases
Openquestions
 To what extent have exposure notification apps succeeded
in notifying pre/asymptomatic cases, leading to testing,
isolation, and further notifications?
 Difficult to measure until Apple/Google have added
statistical tools to their system
 Has aerosol transmission been a key oversight?
 How can governments better prepare technologies for
future emergencies, and processes for science/technical
advice?
 A pandemic was top of the UK’s national risk register for
over a decade, and there was already computer science
research on exposure notification. Yet there was utter chaos
when Covid-19 hit
 Should governments create stronger legal powers of
compulsion over tech companies? But if so, how to
maintain “separation of powers” that can support public
trust?
 Is stronger public institutional control important, esp. with
highly-trusted institutions such as the NHS?

Covid exposure apps in England and Wales

  • 1.
    Covid-19 contact tracingin England and Wales Dr Ian Brown
  • 2.
    Tracingappv1  Govt’s “NHSX”unit outsourced production of a “centralized” contact tracing app, with limited transparency and stakeholder feedback  Data protection regulator provided detailed advice to NHSX, but often felt ignored; later criticized for not taking regulatory action  Frequent expert and media criticism once Apple/Google Exposure Notification API announced, becoming louder as app release delayed for months  As predicted, problems arose with keeping apps “awake” – govt tried and failed to persuade Apple to change iOS (as in France – bitter complaints about sovereignty)  Large-scale trials showed many iPhone users were therefore not sending/receiving notifications  Govt created then scrapped ethics board after sustained criticism
  • 4.
    Tracingappv2  Govt adoptedA+G model and quickly released app, with additional QR code “check-in” functionality (10m downloads in three days)  Media discovered close to zero notifications sent based on hundreds of millions of check-ins  NHSX forced to release update reducing “risk sensitivity” due to earlier configuration mistake  November – govt announced interoperability of apps from E+W, Scotland, N Ireland, Gibraltar & Jersey, but excluded from EU E-Health Network
  • 5.
    Non-NHS accessto Covid-related data  Police givenaccess to test and trace data to check those ordered to isolate are doing so  Public fears about police accessing tracing app data – but use of Apple/Google API allowed a wide range of influencers to explain this isn’t possible  GCHQ/NCSC involvement in designing tracing app, and broad access to NHS information systems  Controversial involvement of Palantir to operate infrastructure  Controversial involvement of Serco to operate national tracing service, with significantly lower success than local public health units handling complex cases
  • 6.
    Openquestions  To whatextent have exposure notification apps succeeded in notifying pre/asymptomatic cases, leading to testing, isolation, and further notifications?  Difficult to measure until Apple/Google have added statistical tools to their system  Has aerosol transmission been a key oversight?  How can governments better prepare technologies for future emergencies, and processes for science/technical advice?  A pandemic was top of the UK’s national risk register for over a decade, and there was already computer science research on exposure notification. Yet there was utter chaos when Covid-19 hit  Should governments create stronger legal powers of compulsion over tech companies? But if so, how to maintain “separation of powers” that can support public trust?  Is stronger public institutional control important, esp. with highly-trusted institutions such as the NHS?

Editor's Notes

  • #2 NB public trust in NHS, and continued application of GDPR in UK, including its DPbD&D Art 25 requirement
  • #3 NB Scotland and NI took Apple/Google approach from the start, latter for interoperability purposes with RoI
  • #5 https://www.youtube.com/watch?v=LliPufBQHEo&feature=youtu.be