What if we could know who will get sick? Or figure out who is at risk of getting an infection before it happens? What if we could prevent an influenza outbreak?
Providing insight into challenges like these is the promise of big data analytics (BDA). When applied to healthcare, analysis of big data can help detect diseases at earlier stages in their progression, manage patient and population health, and assist insurance companies in uncovering cases of healthcare fraud. BDA uses data collected in the provision of care for secondary purposes; analysis that allows hidden patterns to be uncovered and emerging trends to be revealed in real-time, or near real-time.
It is this type of analysis that created FluView, a website and mobile application developed by the Centers for Disease Control and Prevention (CDC). By tracking influenza-like illness activity across the United States, FluView helps health professionals and the general public monitor the prevalence of reported flu symptoms in their state. Using a ‘heat map’ it provides a visual representation of where a flu outbreak may be occurring. A study published in the journal Vaccine estimated that influenza epidemics in the U.S. result in medical costs of $10 billion annually. This is just one example of how the secondary use of health data can improve efficiency in the delivery of care and provide insights to healthcare providers.
The digitization of healthcare is generating data on a massive scale. Individual hospitals are storing petabytes worth (the equivalent of one million GB) of patient data. This creates rich sources of information for data analytics but introduces issues of how to manage and protect that data. Greater amounts of healthcare data being captured mean greater amounts of patients’ protected health information (PHI) being stored. This can put patient privacy at risk. Not properly managing these resources to mitigate the chances that a patient could be re-identified from their data, could put an organization in the crosshairs of government regulators.
PHI can exist in structured data formats in health databases, but more and more often it is unstructured data that is being added to the digital record. The conversion of traditional written forms of health information has led one top-tier consulting company to estimate that 80% of all health data stored is unstructured data. This data includes physician’s notes, prescriptions and pathology reports into digital form, the widespread use of MRI’s and CT scans, raw data from medical research, genotyping, and data from social media sites. Adequately protecting this patient data requires the ability to deal with a variety of data formats.
We are just starting to see the promise that big data analytics holds for the future of healthcare. As the market matures, opportunities to share and aggregate data will proliferate as more organizations aim to take advantage of powerful analytic tools and techniques. Data sharing will also be driven by bold initiatives like the ‘cancer moonshot’ announced by President Obama in the State of the Union address. Ensuring compliance with HIPAA and other privacy laws by knowing how to de-identify medical data, particularly unstructured medical data, will be an essential step to moving forward.
The Promise of Big Data Analytics is the first in the Big Data Analytics Series by Privacy Analytics. Next week: The Rise of Big Data in Healthcare.
The post The Promise of Big Data Analytics appeared first on Privacy Analytics.