Are data owners risk averse? Learning from the Covid-19 pandemic in Suffolk

​

Coda Coastal Health Data Workshop

​

Anna Crispe, 13^th May 2024

Today’s conversation

• What we did with data to support the pandemic response in Suffolk – Coronawatch; the SODA Vulnerable Persons Dataset – sharing some examples that everyone can use if you don’t have examples from your area / practice�
• What did we learn from this about the possible barriers to data sharing – consider your own experience or make some notes as I am talking and we will share our thoughts

​

• What can help us get through these barriers – consider your own experience or make some notes as I am talking and we will share our thoughts�
• Summary discussion and commitment to action!�

At the start of the COVID-19 pandemic we faced many challenges:

• No local data for Suffolk (761,000 people, 1,450 square miles, relatively elderly population, rural and coastal, some urban and coastal deprivation particularly in Ipswich and Lowestoft)
• No effective data sharing with PHE on infectious disease in place (methods or governance)
• No clarity from PHE on what our role was
• No experience of the data needed to handle a global pandemic (although we had the technical skills and knowledge)
• We had some innovative local data sharing arrangements in place through SODA – but not systematically across whole system and not with key partners including Environmental Health, Schools, Care Homes etc

​

From May 2020 onwards, we started getting access to detailed local data…

• New data sources became available as the pandemic progressed – we were one of only a handful of Councils nationally to ever fully automate data sharing from PHE

​

• Individuals had different ‘unique’ identifiers in different datasets – required complex linking and matching across huge volumes of data – our usual methods couldn’t cope!

​

• Constant need to evolve to meet ever expanding demands for data right across the Suffolk system led to ‘CoronaWatch’ which enabled the entire C19 response including:�
• System wide collaboration to deliver required interventions to prevent further transmission
• 24/7 local data on ‘hotspots’ which drove rapid action in high risk settings including schools, care homes and workplaces
• Rapid feedback loops on effectiveness and outcomes e.g outbreak progression, deaths, vaccine take-up

​

CoronaWatch - A three step approach

CoronaWatch – Level 1

A thirst for more…

• As any analyst will say, the more analysis is provided, the more questions get asked!
• As the need for intelligence grew in Suffolk, so did CoronaWatch….
• What started as an eight page Power BI report increased 12-fold over the following months.

​

​

• Thanks to colleagues at NCC for the initial code

​

• In the public domain right from the word go�
• Single source of the truth�
• Ran on automated data sharing and linking from gov.uk and our local data – always provides the most up to date data�
• A key source for everyone in Suffolk, 24/7/365

7

The SODA COVID-19 Vulnerable Persons Dataset - Project Background

• Following the outbreak of COVID-19, SODA was tasked by the Collaborative Communities Board (CCB) to identify individuals and households in Suffolk who may be particularly vulnerable to the impact of the coronavirus either clinically, financially, or socially.
• The CCB drew up a list of vulnerable groups, beyond those identified centrally through the ‘NHS shielded list’:

• A team of analytical, data, IG, and IT specialists was pulled together to support the CCB in identifying individuals falling into these groups across Suffolk.

8

Data Collection and Collation

• Over 40 datasets were encrypted before being sent to SODA and uploaded to a secure data warehouse. Suffolk County Council data, UK government data including the Shielding list, and food parcel delivery information were also uploaded.
• We linked data on 145,000 individuals in 109,000 households and analysed the depth and width of vulnerability in the population – this was then given to our innovative ‘Home But Not Alone’ Collaboration who used it to make proactive contact with people and families

The Impact of the Vulnerable Persons dataset – numbers, money and funding

Enabled 24,283 proactive outbound calls to be made by ‘Home But Not Alone’ to the most vulnerable people and families in Suffolk – Suffolk Collaborative Communities Board

Recognition of the validity and importance of using data to find individual people in need of support – East Suffolk Community Partnership Forum

Led directly to the creation of a specific funding stream for ‘vulnerable people and places’ - East Suffolk Community Partnership Board funding allocation

Part of the data used to make proactive contact with 30,572 people in Suffolk for Winter Grant Scheme (Dec 20 – Sep 21) which led to the distribution of £5.2 m to people in need of support – Suffolk Collaborative Communities Board

The Vulnerable Persons’ dataset directly helped us mitigate the negative effects of managing C19

​

• It clearly identified individuals with both wide and deep vulnerability allowing immediate proactive contact and support through HBNA, building on other linked data and finding real and urgent need�
• Identified people for ongoing support - ‘Summer in a Box’, winter grant scheme, PHM�
• Has helped the whole Suffolk ‘system’ to understand vulnerabilities in our population in new ways, which is now directly influencing resource allocation�
• Highlighted as international best practice by the Kings’ Fund, and as national best practice by Local Government Chronicle�
• Will help us to mitigate the long term effects of the pandemic, including inequalities – if we can keep it going past the pandemic and link it into our new population health management data

​

• Made us think much harder about data ethics and put a framework and process in place to �support

But it probably shouldn’t have taken a pandemic for this to happen, or for it to gain serious momentum!���

Discussion time – from what I have said, and from your own experience in recent years, what would you identify as key barriers to data sharing in your area / organisation?

Not enough data

No analysts / no tech

Can’t link the data

Don’t know what data we have

It’s illegal

Data we can’t ‘get at’

Too much data

Data infrastructures cost money

The data’s all wrong

No business case

Scary for senior leaders

Once you know, you have to act…

Not in my organisation’s interests

Not enough resources

Not my priority right now…

Some ‘reasons’ for not sharing data I have witnessed/heard over the years…

Further discussion- from your own experience, or from what I have said today, what can help us to get through those barriers?

Tell (lots of) stories – LOUDLY!

Find your tribe – IT, IG, academia and practitioners

What makes the business case – safety? Quality? Money?

Be able to prove your impact

Make it useful for the ‘front line’

Use statistical testing

Present it properly!

Take small steps if you have to

#demandyourdata – mobilise subjects

Data leadership – who?

Invest in analysts and training

Be opportunistic

Relationships matter – not just when you are in a crisis

Take the time to really understand the detail of who holds what in your area

Get the foundations right asap – intent, IG documents, governance

Think about all the data – the places and the people

Use national policy when it helps you

Think laterally – proxy indicators may be the best place to start

Use routine data to understand vulnerability now

It is not impossible that something will go wrong – think ahead about how to manage risks and challenges

What action are you going to commit to doing today following this discussion to enable better data sharing in the future?

Read guidance or policy I’ve learnt about today

Talk to someone more regularly – build a relationship

Find out more about my organisations’ IG policies and those of our key stakeholders

Think more about data ethics and how to tackle those issues upfront

Make a detailed list of where all the key coastal data sits and who holds it

Think about if something went wrong – how can you future proof?

Thank you for listening….

Final comments