Bob Plumridge

The Obstacle Course of Big Data Healthcare

Blog Post created by Bob Plumridge on Jul 14, 2016

Digital technologies are often hailed as a solution for exploding healthcare costs, and Big Data analytics is arguably leading this digital transformation of the healthcare sector. By helping us to understand correlations between lifestyle, medical history and health provision, data can turn healthcare on its head: instead of treating people only once they are ill, we may be able to target at-risk individuals for prevention, so that they are less likely to become patients.

 

It’s a big promise, but in real life the implementation of Big Data healthcare faces an obstacle course. Healthcare professionals find themselves battling dozens of incompatible data formats, struggle to make huge volumes of data both portable and secure, and find it challenging to meet strict guidelines for data privacy.

 

Big Data 1 new.png

 

Your Big Data Doctor Will See You Now

 

Who keeps you healthy? Your family doctor? A nurse practitioner? The consultant in your local hospital? Your vital health tips might actually come from a data base, as digital technologies are finally transforming healthcare.

 

We know that Big Data helps patients make more informed choices, allows clinicians to understand medical conditions better, and makes it easier for administrators to deploy resources when and where they are needed.  But to get there, caregivers first need to make the right technology choices.

 

Connecting the dots

 

Many healthcare providers are beginning to realise that data they need is either trapped in medical devices or stuck in the silos of medical departments, and further tied down by incompatible data formats and standards. Setting these data free is only the first step, though. To turn data into insight, it needs to be standardised in a central data repository. Only then can caregivers connect the medical dots and have a unified patient view, instead of poring over paper records.

 

The digital transformation is also powered by a new set of data, generated by wearable devices like smartwatches and fitness trackers. By gathering patient data in real time, not every check-up requires a visit to the hospital; clinicians can now monitor patients in their day-to-day environment.

 

Big Data analytics is also revolutionising medical research. Carefully designed studies – with randomized control groups – will continue to be relevant, but they can now be complemented by analysing data sets that are based on sample sizes close to 100% of the patient population. This makes it possible to identify health trends early, and helps to spot correlations between people’s health, their lifestyles and personal circumstances.

Using data insights, patient care can be tailored for the individual like never before; health risks can be flagged and tackled before they affect a person’s wellbeing. Medical teams, meanwhile, can monitor their own performance to spot problems and deal with inefficiencies.

 

What we are seeing is the emergence of an Internet of Health. It is the medical incarnation of the Internet of Things, helping patients to reap the benefits of the combined power of sensors, connectivity and data.

The Internet of Health, however, will work only if care providers manage to overcome the Big Data obstacle course.

 

The Power of Data

 

Take type 2 diabetes, a rapidly growing, but also largely preventable disease. Across France, Germany, Italy, Spain, and the UK the total direct cost burden of people with diabetes is €90 billion.

 

In Salford, a city in the North of England, around one in 10 men over the age of 60 is suffering from diabetes, with obesity a key factor in triggering the disease. Even relatively small lifestyle changes can make a difference, though. Every kilogram of weight loss, for example, reduces the risk of diabetes by up to 15%. However, instead of broadcasting this message to everybody, a local partnership of three healthcare providers is using Big Data insights to identify patients who are at risk. Local doctors then approach them to join a project to stop the illness in its tracks. With some mentoring and the help of wearable devices, the participants are encouraged to change their lifestyle – like doing more exercise, drink less alcohol and eat healthier. It’s possible because Salford has an integrated electronic records system, which is now delivering significant savings and giving the at-risk group a better quality of life.

 

In Poland, data helped a hospital to tackle another serious problem: Too many patients who had a hip operation were readmitted to hospital with an infection. There is a simple remedy: patients need to push themselves and walk more after an operation. So the hospital gave its patients wearable fitness trackers and began to monitor their mobility; those who did not meet their daily step targets were contacted and encouraged to get out and about. The outcome: readmission rates dropped by more than 40%.

 

In the Netherlands, the St Antonius group of clinical training hospitals wanted to use data to improve patient care and reduce operating costs. Patient and research data, however, were trapped in the silos of numerous hospital departments. By creating a central data warehouse and using the Pentaho business intelligence platform, the hospitals could rapidly identify problems and develop solutions to improve patient care. As a result, the hospitals saw a 20% reduction in emergency room turnaround times, an improved use of surgery rooms and personnel, better access to research data, and superior tools in the hands of doctors that allowed them to interpret patient data.

 

The Obstacle Course

 

Yielding the power of the Internet of Health does not come easy, though.

 

Data Fragmentation

 

Most healthcare systems have badly fragmented IT set-ups, with every doctor, every surgery, every department in every hospital making its own choices of system providers and data formats. To gain a unified patient view and introduce Big Data analytics, healthcare providers need to create integrated data repositories that can take in and standardise data from 50 to 100 medical systems, made by 30 different vendors, using 20 data formats.

 

Data Volumes

 

They also have to cope with the sheer volume of data generated; a single MRI scan can easily take up several gigabytes of storage. Adding to the load are external data sources, whether it is wearable devices or social media analytics, which is a new way for hospitals to spot early if patients are unhappy with the service they provide.

 

Data management

 

Next, all the data must be integrated, to ensure the data specific to a patient is available at the point of care. At the same time, the data has to be fed into anonymised data sets, so that medical researchers have information that is both accurate, but also ensures the privacy of each individual patient.

 

Data portability

 

Patients and clinicians also need to be able to access the information – across multiple locations, and with mobile devices on the go. It’s easily said, but to do so reliably continues to be a challenge for organisations that have built not a data repository, but a data archipelago.

 

Data security

 

Then there is the security issue, because not all devices used in today’s hospitals are built for the connected world. As healthcare providers connect them to the internet, they need to ensure that data is encrypted the moment it leaves the machine to go to a central data base, using at least SHA-256 hash encryption.

 

We at Hitachi know how to tackle all these challenges at its root: the data storage level. Big Data healthcare and the Internet of Health need a clinical data repository that supports integrated data analytics and business intelligence solutions.

 

The results are not only dramatically reduced costs, but – more importantly - instant clinical benefits and better patient outcomes.

 

If you want to find out more about the power of digital transformation, then a good starting point will be my Webinar, available online from 19 July.

Outcomes