During the 2012 presidential elections, the predictions of one shy outsider with a love for baseball statistics outperformed those of hundreds of campaign veterans and beltway insiders. How did he do it? Instead of gut feelings and anecdotes, Nate Silver and his blog FiveThirtyEight used “big data” to correctly predict the outcome of every single state and the District of Columbia. But Silver was not the only person to use data in that election; the Obama campaign also famously employed big data to help them win votes, using it to do everything from tracking the effectiveness of door-to-door canvasing to determining the contents of ads. Now, more and more the same tools and techniques Silver and the Obama campaign used to find trends in everything from election results to poll numbers are being applied to a very different topic: your health.
Big data refers to the use of specialized tools and techniques to understand and distill meaningful insights of massive amounts of information. Imagine using a spreadsheet to record your friend’s favorite restaurants. By looking at the document you might be able to see that most of your friends love Italian or that nobody’s favorite restaurant is seafood. Big data does the same thing, except instead of having information from a handful of friends; the file contains information from thousands, even millions of people. Crunching this cache of data requires more than your basic spreadsheet program and laptop, but the benefit is being able to find nuanced, hidden patterns in otherwise impenetrable masses of information.
Health is particularly well suited to benefit from big data. For decades hospitals, researchers, and government agencies have diligently collected a huge variety of health data, from the success rates of drug trials, to the cost of an average medical procedure, to the demographic information of patients, to the average wait time in emergency rooms. A recent report by the research and consulting firm McKinsey & Company found that there are four “pools” of data in the health care field: pharmaceutical research data (e.g. clinical trial results), clinical data (e.g. patient records), activity and cost data (e.g. estimated procedure costs), and patient behavior data (e.g. health purchases history). Where big data comes in is gathering all this information together in one place, sometimes from many different data warehouses, and using it to gain insights into how our health care system can be better. Want to know which drugs are least likely to have side effects? Which individual doctors have the best outcomes? Which procedures are most cost-effective? Big data could answer these questions and more.
There are three major areas where big data is revolutionizing health care. First, big data is increasingly being used by health care providers to identify patients at high-risk for certain medical conditions before major problems occur. One health care system in Texas is using data from clinical records and insurance claims to identify which patients will be at high-risk of an ailment in the future and using that information to offer them early preventative services. A different provider has been using data from patient records to build predictive models around intervention successes. The analysis provided by the data has successfully cut down on hospital readmissions by 31 percent. Google is also getting involved in identifying health risks. Using data from user search histories, the technology company is able to track the spread of the flu across the world in near real time.
This is where the power of big data becomes the power of knowing. By working with big data experts like Siemens Healthcare, providers are able to use solutions that automate the collection of healthcare data, aggregating and standardizing the data across an enterprise, thereby using predictive analytics to help identify populations at risk, while monitoring performance at all levels of the organization.
Second, big data is also being used to increase the quality of care received by patients. One way this is being done is using data to build a “clinical decision support system,” that is a tool health care providers can use to evaluate their proposed treatments — for example, identifying medical errors before they happen. This short video explains some of the exciting possibilities that can arise when a clinical decision support solution acts more like a helpful assistant, providing relevant information to physicians when it is needed most (link ow.ly/nkt3Y ). A study in the research journal Pediatrics found evidence that one such system in a major US city reduced adverse drug reactions by 40 percent over the course of only two months. In another case, a health care provider in Southern California used clinical data on the actions of staff doctors to find that one physician was using a certain antibiotic far more often than the rest of the staff — potentially increasing the risk of drug-resistant bacteria.
The final area where big data is changing our health system is in helping reduce the mounting costs of health care. Health care is one of the biggest sectors in the US economy and consumes a significant amount of the country’s gross domestic product. At the same time, there is ample evidence that a significant proportion of every dollar spent on health care is wasted, whether by bloated overhead or unnecessary procedures. Big data is taking a big role in bringing these costs down. In one case, clinical data was used to discover which doctors were costing the most money in procedures and other treatments. By reviewing their actions with them, the health care provider was able to identify and reduce duplicate tests and unnecessary procedures. The move not only lowered costs, it improved patient outcomes as well. In the United Kingdom, the National Health Service is using data on the clinical and cost effectiveness of new drugs to help negotiate drug prices with the pharmaceutical industry. In Texas, one health system is using data from 14,000 patients and 6,000 employees to identify individuals likely to require expensive health care services in the future. Using the data they can identify whom to target with preventative care before the costly health issue arises.
McKinsey & Company estimates that the use of big data in health care could produce a savings of up to almost half a trillion dollars. However, as the firm points out, the benefits of big data will not come automatically. It will take strong partnerships between the technology companies like Siemens building the tools and the health care providers using them to achieve even a fraction of the potential savings.
The private sector is not the only area taking notice of the importance of big data in the future of health. In July, the National Institutes of Health announced $24 million in annual funding to build a handful of “Big Data to Knowledge Centers of Excellence.” The centers will help the health care research and clinic community better understand how it can use big data to improve the nation’s health care system.
Health care is about people, not numbers. Big data’s growing role in health care does not change that. Health data, at the end of the day, is also about people, whether exercise statistics or genetic information. By taking advantage of the wealth of health data available through working with big data experts, providers are able to identify areas where improvement is possible and will begin to realize better outcomes, improved efficiency and a more sustainable healthcare system.