Using big data to modernise care delivery

Harnessing the potential of big data is one of the biggest opportunities facing the NHS. It will be at the heart of how we make care better, safer and more affordable in the future. The question, therefore, is not whether big data will transform care, but how to maximise the benefits. The danger for the NHS is that it focuses on data at a macro level, building pictures of health trends across populations. Useful though that is, it merely classifies problems. The real breakthrough comes from using data to identify solutions, helping doctors and nurses to make real-time decisions. The NHS does not need to choose between using data to plot problems or identify solutions; it needs to be ambitious in embracing both, and using big data to transform the quality of care.

Overcoming barriers
Before the NHS conquers data, however, it first needs to overcome its cultural aversion to sharing it. Some industries, such as retail and banking, have been quick to see monetary value in high quality customer data and have developed advanced systems to capture, trade and use it for commercial advantage. The NHS is different. The lack of commercial drivers, combined with greater sensitivity and legal constraints around medical data has created a nervousness about sharing it. The consequences for care are stark; clinicians are left facing decisions without vital pieces of the jigsaw that did not follow the patient through the system.
    
Fortunately for big data advocates, the tide is turning on the reluctance of the NHS to share information. In May last year, the Department of Health’s information strategy ‘The Power of Information’ explicitly stated that “not sharing information has the potential to do more harm than sharing it.” This was followed in March 2013 by the ‘Caldicott2’ report, a Government-commissioned review tasked with identifying how NHS information sharing could be improved without compromising patient confidentiality. The Health Secretary Jeremy Hunt has spoken passionately about the need to improve the quality of data as part of his plan to make Britain a “global hub” for health technology. It would be naïve to think political will can reverse decades of cultural conservatism overnight, but the direction of travel is the right one.

Tracking illness in the population
In seeking to harness the power of big data, the easiest approach for the NHS would be to focus on its secondary benefits. This involves joining up data currently locked in silos to identify health opportunities across the population. Some work has already started on this, with national disease registries and more recently with NHS England’s care.data project  already extracting data from General Practice and seeking to extract and link data from hospitals, and down the line from community and mental health providers across the NHS.
    
The US has arguably gone further. Originally designed in 2003 as an early detection mechanism for bioterrorism, the BioSense project now uses pooled data to track emerging public health problems in real time. The cloud-hosted system is a model for how the NHS could integrate data to proactively identify and respond to health challenges.

Other secondary benefits of big data for the NHS would include capturing performance data that could benchmark the performance of hospital departments and highlight pockets of poor care. Such a benchmarking system was recommended by the Francis Report into failings at Mid Staffordshire NHS Trust, and would allow for interventions before – rather than after – problems compromised patient safety.

Primary benefits
The more ambitious target, however, is for the NHS to realise the direct or primary benefits of improved data that can enhance quality and safety at the level of an individual patient. This means improving the flow of information to clinicians to enable them to make quicker, more informed decisions. One example is sepsis, which kills close to a third of affected patients. Survival hinges on detecting sepsis in the initial 6 hours of onset, meaning every hour counts. Applied intelligently and in real time, data can directly increase the chance of survival. Cerner’s St John’s Sepsis Agent technology continuously monitors key clinical indicators and where sepsis patterns are detected an automatic alert is sent to care teams who can intervene. By using big data to structure and process data directly in clinical practice, the NHS can directly improve care and save lives.

Cloud and real time data analytics will enable the proliferation of algorithms and decision support to direct patient care and the care of patient cohorts. This will expand with the availability of genomics data mapped to patients’ electronic medical records. Decision support using this linked information enables personalised care. Also, the progressing genetic re-classification of disease will change existing handfuls of histopathological cancer types into hundreds of types per cancer. We will require huge patient databases to understand how to treat smaller numbers of each genetic type, if not down to the individual patient level. By merging the scale of big data with the detail of ‘small’ data, information can be applied to individual patients at the point of care. This is no longer science fiction but an achievable technological reality.

Using data to make savings
In advocating the transformation of care, we have to be mindful of the financial realities facing the NHS. In the short-term, the NHS needs to realise £20 billion of efficiency savings by 2015. Longer-term, a financial time‑bomb caused by the rise of chronic diseases in an ageing population makes the current model of care delivery unsustainable. Data offers a solution and should underpin the effort to make care affordable. By mapping health trends across a population, at risks groups can be identified and resources targeted more effectively. At an individual level, enabling patients to manage their own health and avoid hospital admissions dramatically reduces the cost of care. In the US, a pilot to join up patient data between the Department of Veterans Affairs and Kaiser Permanente, a leading Californian care consortium, reduced patient visits by over 26 per cent. If the NHS is to remain affordable, it has to transform how it delivers care. Harnessing big data to better target resources and prevent hospital admissions should be at the heart of this. Far from undermining the case for investment in big data technology, budget constraint reinforces it.

Embracing change
The NHS has much to be proud of as it celebrates its sixty fifth birthday. It remains a model of what an ambitious, progressive society can achieve, even in times of austerity. But if the NHS is to celebrate another sixty five years, it needs to radically change how it cares for Britain. By embracing data, the NHS has the possibility to improve the quality, safety and affordability of care. The question, however, is how ambitious the NHS is prepared to be. If it restricts itself to secondary benefits of smarter data, the NHS will be a more informed purchaser, with accurate pictures of health trends, hospital performance and at risk patients. To truly transform care, however, the NHS needs to go further, directly applying data into clinical practice to improve outcomes, strengthen patient safety and prevent hospital admissions.
    
The challenge is not unique; it is the same as that faced by companies such as Google, Facebook and Amazon. In harvesting aggregated data, online companies build a valuable business asset, but their success hinges on analysing and applying that data at the level of an individual consumer. That allows for tailored interventions, such as targeted advertising on Google or suggested purchases on Amazon that change how consumers behave.
    
The context for the NHS is different, but the objective is the same; using data to better understand a patient’s health and target specific interventions to improve it. It will require ambition to make it a reality, but the prize of a better NHS makes it worth striving for.

Further information
www.cerner.com