Our approach to data

The types of analytics we use

Descriptive: what has happened?

By describing the current situation, we can focus our effort on groups that are most likely to have poorer health outcomes. For example, with the Social Progress Imperative, we’ve created a set of metrics to assess the social progress of our boroughs of Lambeth and Southwark. The result of this partnership is the Urban Health Index.

Diagnostic: why has it happened?

Sometimes, correlations are not enough to understand how to best impact a complex issue. More comprehensive analyses, such as system mapping are useful to uncover the direct and indirect causes of poor health outcomes. Diagnostic analysis helps us tackle the right part of a complex issue to maximise our impact.

Predictive: what is likely to happen in the future?

Understanding the current mechanics at play can help us model possible future outcomes. Typical questions include modelling the propagation of information or estimating the impact of new regulations on future health outcomes. For example, using the prevalence of known risk factors we can identify areas where people are more at risk of poor health outcomes in the future.

Prescriptive: what should we do about it?

By modelling the likely effects of an intervention in the years to come, we can inform our current thinking. For example, we commissioned Health Lumen to model the health and economic impacts of both NO₂ and PM_2.5 exposure under different scenarios.

Interrogating multiple sources

To build the most exhaustive picture of our place, we need to tap into a broad mix of data. We achieve this both through accessing public data, collecting our own and partnering with organisations.

Open-source

Open-source data is a fantastic resource. We rely on sources that are available nationally and offer the granularity we need to understand our place in detail:

• Office for National Statistics is an important repository for demographic characteristics and social determinants of health
• The London Datastore contains the London Atmospheric Emissions Inventory informs our health effects of air pollution programme
• The National Child Measurement Programme is insightful for our children’s health and food programme

Data collection

Sometimes, the data that is available is not enough for us to validate some of the hypotheses that underpin our programmatic work. We then commission collection to specifically plug our knowledge gap. This could be a survey or collecting data from online sources like Twitter to analyse sentiments about a topic or understand how a topic is being discussed online.

Partnerships

Partnering with organisations that share our vision is a great way to enrich both parties’ knowledge bases. In the past, this has allowed us to access and explore data that we otherwise would not have been able to.

We have partnered with organisations to help them develop their product in a way that is also furthering our mission, funded a Data Scientist to be seconded at the Health Foundation to produce insights that respond to both organisations’ research needs or used a partner’s methodology to apply it to our data.

Our approach to partnership is innovative and flexible to allow for the right insights to be surfaced to shape great programmes and policy influencing work.

Responsible use

It is our responsibility as an organisation to ensure our partnerships and the data we use are not putting us or the people we serve at risk. To do this, we have data analysis principles, so that our analysts, consultants and partners have a shared understanding of our approach. We also have a Data Advisory Group that provides us with an external perspective on our work. Our work on data ethics is not set in stone, it’s an evolving toolkit we build on and share to inspire others to use data responsibly.

Creating insights for everyone to use

Our insights are key to shaping our strategies and operational decisions. We aim to support others to use our products to inform their own organisational decisions.

Our Urban Health Index, developed with the Social Progress Imperative, is a great way to move beyond economic indicators as a proxy for health outcomes. We hope that it can inspire more people to use data to focus their interventions.