Event summary: Data and healthcare innovation

Event summary: Data and healthcare innovation

Written by Lauren Howells on Tue 11 Jul 2017

Big data can revolutionise the diagnosis and treatment for millions globally. But the power of data is only harnessed when we analyse, interpret and derive action from it.

Edo and data partner White Swan, are piloting programmes that use big data, and complex data modelling for early health condition diagnosis. We recently held a roundtable for healthcare and charity professionals, to understand how organisations are applying data and technology to achieve their mandates. Attendees included British Heart Foundation, AgeUK, Royal College of Physicians, Sue Ryder among others. 

Introducing White Swan

Steve King is the founder of Black Swan, a technology and data science company. In 2015, his sister, Julie King, was wheelchair bound and in danger of losing her life.  For 8 years her health had been deteriorating, yet despite countless tests doctors were at a loss to explain her condition. Sadly, her family were told to prepare for the worst. Steve enlisted the help of his Black Swan data scientists, and by analysing Julie’s symptoms they were able to diagnose a her with a rare form of Parkinson’s. 

White Swan was born. Using Black Swan’s systems and know-how, White Swan help experts and ‘not-for-profit’ organisations to help people get, and stay, healthy. White Swan’s mission is to improve the health of society by accelerating the path to diagnosis, improving the effectiveness of treatment and ultimately helping prevent illness through the use of technology and analytics.

White Swan work with both public and private data sources, fusing them together to generate new powerful insights.  They also visualise data for clients in order to give them the power to create actionable insight.  

What do we mean by data?

Any symbol, character, word or number that can be accessed for analysis such as tweets, emojis, step count, air temperature, time, locations… the list goes on and on! 

We can analyse sentiment, can identify common words and phrases and can even train the tools to pick up on sarcasm. 

It is critical that the whole picture is assessed - not just one element; the data must all work together to map insight. For example, analysis of air temperature coupled with sentiment and status on feelings, can give insight into the effect weather has on mood and health conditions.

What can we do with data? 

The power isn’t the data, but what is done with the insight. Data can be used to map patient journeys to diagnosis, their experience of treatment, give insight to reduce suffering and so on. Some examples of how White Swan and Edo have worked to use data effectively are below: 

Case study: Understanding Dementia with Alzheimer’s Society (White Swan and Edo)

White Swan and Edo have been working together to help Alzheimer’s Society gain an understanding of the patient and carer experience for people living with a dementia diagnosis -  spanning from how a dementia diagnosis affects a person, to end of life care. This work will ensure the charity can most effectively support its service users.

Over 1.2m conversations across news, social media and the Talking Point Forum (run by Alzheimer’s Society) were analysed, across the diagnosis and treatment journey. Utilising Black Swan’s proprietary technology an in-depth understanding of the dementia patient and carer experience was mapped out.  

With this, a clear set of findings and recommendations to shape Alzheimer’s Society future services was mapped across the patient journey, including: 

  • Where fears and challenges exist, and the importance of them.
  • The tipping points for when carers seek help and the support they need.
  • The challenges of reaching a diagnosis and the paths patients go through.

Case study: Identifying triggers for Meniere’s (White Swan)

Meniere’s is a disorder of the inner ear which impacts balance and creates ringing in the ears. There isn’t a cure and therefore suffers have to learn how to live with (and minimise) the symptoms. 

White Swan analysed data collected from the Meniere’s Monitor app. They collated symptom patterns, and mapped to sensing data (such as the weather) to identify patterns and distinguish whether they were able to predict attacks brought on by the disorder. 

While the work is still ongoing, links have been found between air pressure changes and Meniere’s attacks. Further analysis is being done to investigate sufferers’ potential behavioural or lifestyle factors, such as alcohol and salt intake. 

Case study: Predicting A&E attendance (White Swan)

It is often difficult to staff Accident and Emergency (A&E) departments effectively as there is uncertainty around busy and quiet periods. To predict A&E attendance and admission rates with greater accuracy would enable better planning of resources across hospitals.

White Swan conducted analysis of three years of historical aggregated and anonymised A&E admittance and attendance data for a major NHS Trust coupled with localised social, weather and flu season proxy data created a predictive model with 86% accuracy 48hrs in advance for attendance. The data was also able to accurately predict spikes in flu admissions (at 72% accuracy) and trauma admissions (at 85% accuracy).

The analysis also identified the most important drivers for A&E admissions, the majority of which were available days or weeks in advance, highlighting potential for longer term forecasts.

Data-rich user experience

Any successful customer experience design will derive from the right mix of qualitative and quantitative data that forms the evidence-base by which it is informed. In addition to Edo’s tried and tested research methodology, White Swan added a complementary component centred around social listening and predictive analytics. Put simply, they have the ability to examine the conversations about issues such as dementia, arthritis, mental health and Parkinson’s across a wide-range of online spaces. This allows for very high volumes of data to be collected, making trend analysis as robust as possible. Couple this with the invaluable qualitative, more targeted data (captured in person and remotely) that is more traditionally associated with a user-centred design process, and you have a powerful and compelling foundation upon which to create the most socially impactful outcomes.