Data Analytics is (Still) the Next Big Thing02.06.15
My not-so-bold prediction for 2015 is that it’ll be a busy year for the healthcare data analytics business. I admit that making 2015 predictions now is a little like placing a bet after the horses have run the first quarter mile, but the ONC’s recently published draft 10-year interoperability roadmap shows that there’s a lot of galloping left to do.
The ONC’s roadmap advocates a “nationwide learning health system” that “requires a high degree of information sharing between individuals, providers and organizations.” It espouses an “ecosystem” of information sharing that includes governmental agencies (e.g., state and local health departments, emergency responders and public safety), hospitals, healthcare professionals, diagnostic laboratories, researchers, and non-governmental social services, advocacy and community based organizations.
That vision of a data utopia leaves me wondering who’s going to read all that stuff. I don’t expect that my primary care physician will have the time to read the novel of my life in data every time I show up for treatment of a sinus infection. And I seriously doubt that he or my health insurance will be willing to pay to have someone with sufficient medical knowledge regularly review my personal stream of health data to identify issues that should be flagged for review by my physician.
That’s why I think 2015 will be good for healthcare data analytics companies.
Sure, healthcare data analytics has been trumpeted repeatedly over the last few years as the next big thing. (See, e.g., Data analytics poised for big growth (2012), Clinical analytics ‘next big thing’ (2013), Big Love for Big Data? The Remedy for Healthcare Quality Improvements (2014)). But I’m not talking about pie-in-the-sky data analytics that offer the “tantalizing promise of big data” that will enable healthcare providers “to plan, achieve, and measure…quantifiable program improvements.” (I borrowed those quotes from the “Big Love” article.) I’m talking about automated systems that will review all of the data that the ONC’s interoperability roadmap says should flow into my electronic health record (EHR) maintained by my physician’s practice and determine the relevance of that data to my physician’s treatment plan for me.
In its December 2013 final report, a technical expert panel convened by the National eHealth Collaborative to consider best practices for integrating patient-generated health data (e.g., by smartphone health apps) into clinical workflows said that “decision support algorithms to reliably identify situations requiring human notification and review of PGHI [patient-generated health data], will be increasingly needed to improve workload and reduce liability concerns. Voluminous (or even modest amount of) PGHI may exceed the human resources available, making technologies that help improve workload essential.” From both a patient-care perspective and a medical malpractice perspective, a healthcare provider is responsible for knowing all data that is incorporated into a patient’s EHR--regardless of its source--and the relevance of that data to patient care.
To borrow a concept from the ONC, sharing data is merely an abstract goal unless the recipients of the data can meaningfully use it. We must first have data analytics systems that can create actionable information from all of the data created by patients, healthcare providers, government agencies, and everyone else mentioned in the ONC’s roadmap.