If you were to ask five people their definition of Big Data, you would get five different answers. Tons of information is received and stored in a patient record, and electronic health records (EHRs) are the repositories for all this information. How we apply Big Data in meaningful ways is the next step in healthcare. Here I provide various applications that have promise in the future, but before we get to the application of Big Data, first we need to define what makes Big Data special. We can define Big Data as having at least 3 dimensions:
Volume: The reason it’s not just called regular data—the vast amount of information transmitted is volumetrically incomprehensible to conventional understanding. Just think of the amount of Facebook likes, Tweets, and videos sent every day. We are talking about zettabytes and brontobytes of data, not terabytes. Big Data technology is used to understand the data that traditional databases cannot process.
Velocity: How fast data is generated. Not only are large amounts of data being generated, they are generated at lightning quick speeds. Imagine the constant stream of data medical and fitness wearables transmit. Big Data has application in the analysis of this data in real time to improve upon the health of patients.
Variety: Typical data structured according to its relational relevance, the number of sales in a month for example. But imagine a platform Facebook, it is difficult to put photos, statuses, likes, and shares into one table with the purpose of understanding it and creating actionable tasks. The amount of data we have now is varied to the extent that we need Big Data to help us analyze the different dimensions of data.
In relation to healthcare, a simple and analogous definition we can apply is from an article by Travis Good where he states, “… [Big Data] is data of a subject in multiple dimensions. The more dimensions you have on the subject, the more powerful the Big Data and insights are on that subject.” To put that into perspective, look at the dimensions of a patient record he provides:
- Patient Info (MRN, member number)
- Provider Info
- Care Plan Discharge
- Reason for Referral
- Problem List
- Functional and Cognitive Status Results (Labs)
- Social History
- Vital Signs
- Discharge Instructions
- Health Financial Amounts
- Wellness and Care Management
- Programs Alerts
Population Health Management
When we think about improving healthcare in the U.S, we often resort to the “Triple Aim” of healthcare, that is, improving the patient experience of care (including quality and satisfaction), improving the health of populations, and reducing the per capita cost of health care. Where Big Data intersects with this aim is through the application of population health management (PHM), or the aggregation of data across multiple information points to improve overall wellbeing. Data may come from multiple sources that cannot be measured in one table, for example:
- Patient cost across different fields and specialties
- Electronic Quality Reporting Data (PQRS, Meaningful Use, etc.)
- Pattern analysis between different departments on how to provide cost effective care
- Factors that inhibit or contribute to interoperability
Population heath management is essentially Big Data applied to healthcare. A PHM analysis can take huge sets of data on a patient record and identify care gaps, to provide ways for providers to bridge these gaps. Population health management comes down to data—accurate data that is, from sources that can properly manage and quantify information.
We are in the midst of a transition from traditional fee-for-service to value and outcome-based payment models. Value-based payments are predicated on the idea of quantifying the value of care. For example, examine the criteria of one of MACRA’s payment routes, MIPS. In MIPS, Providers are paid based on how much “value” they bring to patients based on a composite score. The factors involved in a MIPS score do not run completely parallel. With Big Data, we are able to intelligently analyze large sets of data to provide value.
Another way Big Data can guide our industry into a value-based care system is through a facet of healthcare we are sorely underperforming in: preventive care. We are great at reactive care—ordering rounds of chemotherapy for example, but we are terrible at providing services which stymie the symptoms before they happen. Trillions of dollars are spent each year on reactive care services, and what Big Data can do is analyze huge sets of patient records and find care gaps that can prevent illnesses before they happen. This comes in the form of education, prescription analysis, encounter analysis, etc. Big Data will be essential to preventive care in the future.
Big Data, Questions and Concerns
Big Data may be a valuable tool in achieving the Triple Aim in healthcare, but we must be mindful of the overuse or misuse of Big Data. In recent news, some have proposed to add another V to the dimensions of Big Data in the form of vulnerability, and for good reason. The purpose of Big Data is to digest and understand the data of everyone and everything, by the gigabyte, with the goal of providing more insight into human behavior—insight that will not always have altruistic means. Invasion of privacy comes to mind, as consumers will have concern over the use of their everyday life data.
It falls to organizations that handle large quantities of data, EHRs for example, to provide adequate data security and management for their clients where servers and networks are constantly bombarded with attempts from unauthorized information seekers. At Bizmatics, we employ simple but effective data security tasks in our daily processes:
- comprehensive scans for malware and any external vulnerabilities
- upgrading all anti-virus and anti-malware programs
- upgrading server and account passwords
- revising firewall configurations
- implementing strict password policies
We also implement an active traffic-monitoring solution called Cloud Defender by Alert Logic for Bizmatics’ network. System security is not static but a continuous process, and we continue to review network configuration to strengthen defenses against cyberattacks.
The future of Big Data in healthcare is utilizing large sets of data in meaningful ways, with the goal of improving patient outcomes. Big Data has useful application in both population health management and value-based care, and we must be mindful of misusing Big Data. How we will navigate between interoperability, security, and privacy in regards to Big Data in healthcare, is up to us in the industry.