Big data is the practice of taking masses of information, and mining it to cull out specific trends and characteristics, using sophisticated software and powerful hardware. In healthcare as well as in many other areas, data sets are growing in size because from information sensing mobile devices to remote sensing technologies to wireless networks, data is constantly collected and stored.
The healthcare industry has finally recognized that the days of keeping health data sequestered in proprietary silos is unacceptable. The adoption and near ubiquitous acceptance by clinicians of electronic health record systems requires that there are innovative analytics solutions to peel back the new layers of information and insights contained in these records. A similar growth in electronically available hospital data, ambulatory data, pharmacy and payer claims information and data from personal health devices such as home monitors and smartphones continues to grow.
Consumers are eager to gain access to their health histories and to share them with their care providers. Payers, employers, providers and government agencies have come to realize that reforming healthcare is, in large measure, dependent on sharing important clinical data. Furthermore, the shift towards paying for health services based on outcomes rather than on procedures has focused attention on analytic techniques that can accurately distinguish between good and poor results.
The technology behind these gargantuan data sets now enables researchers to begin to understand underlying cause-and-effect relationships that previously remained hidden in too much information. For example, one important outcome of big data analytics is the ability to predict cases of hospital readmission. Armed with this information, providers are able to implement better interventions among those patients whose conditions and behaviors are associated with a high degree of probability that they will return to the hospital within 30 days of an in-patient stay.
Patients on average spend an hour a year with a clinician, and in that time the clinician and the patient are expected to make effective treatment decisions. With big data they will have the information to understand outcomes of large patient populations which provides the clinical context required for effective decision-making, including ways to reduce medical errors, and more effectively deal with chronic disease management.
The big data analytics approach, includes identifying hidden drug interactions as well as the care processes, conditions and patient characteristics that affect safety and efficacy. This deep analysis enables organizations to determine which approaches are most effective, where gaps exist, and how to chart the structure of complex treatments. As we work with big data, the models will become increasingly predictive and, hopefully, will lead to cures for diseases that today seem intractable
The advent of personalized medicine which includes whole genome sequencing further changes what can be understood and applied in healthcare treatment and outcomes. With deep data analytics, drug developers can match patient types to therapies most likely to work for them, as well as identify new therapeutic approaches, while providers can tailor those remedies to the specific needs of each individual patient. Additionally, health plans can use the information to develop comparative effectiveness models for specific scenarios and develop more efficient and cost-effective approaches to healthcare delivery.
The goal of medicine is to provide the right care, at the right time, every time, for every patient. From a quality of care perspective, it has always been important for healthcare organizations to have a full picture of what is happening in their patients’ health. Predictive analytics with big data is truly the pathway to achieving this type of value based healthcare. It represents a sea-change for the future of medicine for everyone.
