Care.data and big data will fill 'dangerous gaps' in NHS and futureproof it with genomics, argues Tim Kelsey

National director for patients and information Tim Kelsey believes 2015 will be the year big data goes big in the NHS

Care.data, the NHS's controversial data-harvesting programme, is "starting again" and will enable the health service to close "dangerous" gaps in patient care through the proper harvesting of big data and analytics.

That's according to Tim Kelsey, NHS England's national director for patients and information, who was speaking as part of WANdisco's Big Data Breakfast event at City Hall.

Kelsey told the audience that through Care.data and other big data and analytics initiatives, the NHS will be able to greatly improve patient care during 2015.

That's despite MPs suggesting that the Care.data programme should be put on hold due to continuing concerns over patient privacy.

Kelsey believes using big data tools to analyse previously unstructured - or some instances, uncollected - information will help to create a better National Health Service (NHS).

"Care.data is starting again now, there are some pathfinders which are beginning to work on extracting data, linking GP data and hospital data in new ways which mean we can analyse the pathway of care that patients travel on," said Kelsey.

He said this will also help the NHS to properly monitor key information, such as how many patients a doctor has seen.

"It's just an amazing thing - and I sure most people don't realise - that the NHS isn't capable, currently, of telling you how many patients are undergoing chemotherapy, for example," Kelsey said. "There are gaps so big, so dangerous that they just have to be filled from a moral, as well as a political, stance."

"That's what we're going to be doing this year," he added.

However, there's still a lot of controversy surrounding the Care.data project, with one-third of GPs saying they would opt their patients out of the scheme, even if that meant flouting the law. There are also questions over private firms having access to Care.data information, something that many individuals would object to.

Insurance is just one area of business that could benefit from mining patient information - and possibly to the detriment of the person attempting to get insurance. After all, why would a company agree to hand out a policy to a person whose data suggests has a high risk of a heart attack?

Kelsey admitted that the idea of using big data in healthcare is "controversial" but pointed to instances where patients have been asked to provide feedback on their hospital experience, something which, with the help of analytics, he claimed has already helped to improve patient care.

"It's been incredibly interesting to watch the friends and family tests, the crowdsourcing of opinion about services has already improved services, there are millions of examples of how hospitals have acted in real time to make a difference," he said, before adding that the amount of information available is only going to get larger.

"This is a huge dataset which is unparalleled in the world and is about to grow exponentially larger, and just needs some big data analytics to derive some insight from it."

Kelsey said part of his role as national director for patients and information is "to ensure the NHS is very futureproofed and takes proper advantage of the genomics revolution".

Genomics is set to become a hugely important area for the NHS, he told the audience. "This is incredibly exciting and very complex, but at its heart is a big data problem on a scale we've never had before."

Kelsey argued that the analysis of the genome represents the true definition of big data at work.

"We've wrongly used the words ‘big data' to describe things like an administrative dataset which runs through a hospital. They're actually structured datasets, they don't conform to the normal definition of big data, by which we mean random, unstructured data," he said.

Kelsey said high-powered data analysis could open up the human genome and aid in the personalisation of patient care.

"There is one dataset which does do that in healthcare, which we haven't yet understood, and that is in our own molecules. [We could] develop diagnostics in healthcare which not only allows us to understand our gene sequences, but also to relate those to molecules in our bodies, in real time," he explained.

However, this raises big technical challenges, he added.

"The dataset will challenge our current concepts of how we might store data and the volume of that data," he said, before going on to describe how much information can be unlocked by examining unstructured data.

Kelsey used the example of two Imperial College London machines that were used to examine athletes' urine at the London 2012 Olympic Games - but now act as molecular diagnostic machines - to make this point.

"A year ago we thought there were 2,000 molecules in urine, now we know there are 10,000. Those additional 8,000 are new to science, we've never seen them before and this machine is beginning to understand them," he said

"But that machine is producing more data per day than Hammersmith Hospital is producing: these are vast, vast databases," said Kelsey.

"We have to develop collaboration so the NHS can have big data at bedside," he concluded.