|
Abstract
|
Key Points
- Big data is integral to driving and evaluating system level change.
- Governance structures enabling access to large-linked data sets needs to be timely and cost efficient.
- Turning information in to knowledge requires clinicians to be engaged in the interpretation of data.
Wikipedia refers to big data as 'data sets that are too large or complex to be dealt with by traditional data-processing application software' [1]. The concept is not new but it is only in the last two decades that IT solutions for capturing, analysing and storing data have been realised. Major tech companies have invested in systems of data capture and analytics, which turn exa-, zetta- and yottabytes of data into information and knowledge that ultimately serve to enhance revenue streams. Using predictive analytics Netflix can offer us a range of tailored viewing options whilst facebook delivers feeds and personalised ads to our handheld devices based on browsing activities.
So has the behemoth affectionately known as 'health' embraced big data? The short answer is a qualified yes but not at the pace seen in organisations where data is power and power is money. The REDUCE (REducing unwarranted variation in the Delivery of high-qUality hip fraCture services in England and Wales) study reported by Patel et al. [2] in this edition of Age and Ageing is a great example of how we can harness the power of big data to better understand how our systems and processes of care impact outcomes. In this case the authors have targeted a high volume, high cost, high impact condition--hip fracture--and, using linked datasets, have explored how organisational factors affect outcomes for this population. They specifically sought to look at variation in care across hospitals and more importantly where variation in care led to different outcomes--unwarranted clinical variation. The learning derived from this paper extends way beyond a single clinical condition and serves as an example of what is possible using big data.
So how does big data help shape our thinking and actions. It is often divided into a number of characteristics--volume, variety, velocity, veracity, value and variability. Patel et al. [2] interrogate multiple linked datasets (variety) to provide information on almost 179,000 people (volume) with a hip fracture. Big data has its critics who bemoan the lack of granularity that a clinician derived trial dataset contains but what it lacks in granularity it makes up for in volume and in most cases power.
Language: en |