Outline

• Who we are
  • Data Science in PHE
  • Our Vision
• What we do
  • Consolidate
  • Automate indicators
  • Reporting and presentation
  • Package examples
• Outputs

Who we are

PHE is an executive agency of the Department of Health and Social Care. We provide government, local government, the NHS, Parliament, industry and the public with evidence-based professional, scientific expertise and support.

We exist to protect and improve the nation’s health and wellbeing, and reduce health inequalities.

PHE was established on 1 April 2013 to bring together public health specialists from more than 70 organisations into a single public health service

Fingertips is our main data page, this is where we display most of our indicators. These profiles allow users to: * Browse indicators at different geographical levels * Benchmark against the regional or England average * Export data to use locally

Data Science at PHE

The Public Health Data Science (PHDS) team is located within the national Knowledge and Intelligence Service and is part of the Health Improvement Directorate within PHE.

PHDS team was formed in 2015 following reorganisation.

Data science aims to generate insight and knowledge from data and is a developing field within broad public health. PHE is interested in how data science can better support decision making across it’s work in prevention, improving health and wellbeing and healthcare variation and inequality.

PHE’s Public Health Data Science team consists of public health specialists, statisticians and analytical staff. It works closely with colleagues across PHE, supporting the knowledge and Intelligence needs of the local and national public health system and developing close working relationships with external stakeholders to promote the development of data science across the public health system.

5 main areas of work:

• Data
• Methods and indicators
• GBD and analysis
• Developing and applying data science to PHI problems
• Developing capacity and capability

Our vision

Our vision has been to “consolidate, automate, innovate”. Each step makes the subsequent step easier.

The PHDS vision (acknowledgement to Michael Heasman)

Consolidate

The consolidate step is essentially part of the overall data strategy and is a fundamental building block.

We have filled a “Data Lake” built on tidy data principles.

Datasets included:

• Hospital Admission Episodes
• ONS births data
• ONS deaths data
• Global burden of disease data
• Health Survey for England
• Lookups (e.g. geography lookup codes to name)
• Populations

Automate

Having tidy data in a central database allows us to create efficiencies.

There are a number of projects working on automation within Knowledge and Intelligence involving multiple teams across the organisation.

Two broad themes:

1. Data management and automated indicator production
2. Reporting and presenting data

The fingertips pipeline

Data management and automated indicator production

With around 1,700 indicators in our fingertips platform, work has begun to automate many using SQL and R. Such as:

• Mortality Indicators

  

  Automating the production of directly standardised mortality rates directly from the datalake and outputs that can be loaded directly into our fingertips platform. The first release using the R scripts happen this month.

• HES Indicators
  Similarly to the mortality indicator production creating scripts to enable automation.

• Non-datalake indicators
  A tool has been developed to get data from ‘non-data lake’ sources and convert into a suitable format. This tool has been developed using python.

Reporting and presenting data

Application Programming Interface (API)

Fingertips has opened an API to its data, making the data more accessible to users. This includes data from our local health tool too. In recent months we have started to see stakeholders accessing data in this way

How to access it

• Via Fingertips website, go to the download tab and select the API link
• Direct link API