Harnessing and capitalizing upon the monstrous amounts of available healthcare information
By Peter Edelstein, MD, Elsevier
Big Data. Population Health Management. Patient Engagement.
Healthcare reform churns out buzzwords at an alarming rate. But at least big data has a more defined meaning, having come to linguistic life long before the Affordable Care Act was a gleam in President Obama’s eye presumably.
Today’s world runs on big data.” It’s big data that allows millions of us to almost instantaneously receive insurance quotes online; creates your credit score; select a mortgage; and pushes pop-up advertising that just happens to be exactly what you were looking for yesterday.
As is our healthcare industry’s history, big data is yet another capability that has entered the medical arena long after becoming an integral part of non-medical sectors. That said, big data is (finally) here to stay, in our hospitals, our pharmacies, in our insurance systems (where it has been the longest), and in our ambulatory care centers.
Big Data Goals
And like population health management, patient education, and other buzzwords, understanding our specific big data goals and how to achieve them is critical if we are to maximize the success of healthcare reform. So the first question is, What Are Our Healthcare Goals for Big Data?
If an underlying goal of healthcare reform itself is to improve the quality and cost efficiency of care for populations and for individual patients, then we must turn away from reactive care provided in the acute, inpatient facility and strive for proactive, preventative, and maintenance care provided in the ambulatory world (both the outpatient physician office and in the place where patients spend virtually 100% of their time: their homes and workplaces).
Linking to this goal, Big Data can drive the identification of individuals and populations at risk of suboptimal quality and/or cost of care and then to guide intervention to reduce or prevent the realization of the identified risks.
Already, Big Data is playing a foundational role in the first part of this goal. Monstrous amounts of claims data serve to feed clinical analytics models, including predictive models. Such powerful tools allow us to predict which patient populations and individuals are at risk of specific forms of clinical deterioration, high cost care, and/or unanticipated hospitalization and Emergency Department visits.
And recently, the incorporation of public records Big Data (including moving, home ownership, eviction, lien, and property value history; estimated annual income, wealth index and financial stress; and accident, fraud, burglary, and criminal history) along with health claims data has allowed for the development of even more powerful predictive analytics models. (For example, inclusion of such non-medical data may more accurately predict risk of early post-discharge hospital readmission and/or risk of failure to pay).
Big Data Expansion
Today, non-clinical Big Data is expanding from the clinical analytics world into site of health care delivery. The Institute of Medicine is recommending the inclusion of social and behavioral data within the EHR, where (as with analytics) this expansion of Big Data is projected to more clearly and accurately guide patient care.
Whether empowering clinical analytics models or more clearly defining individual patients and populations from within the EHR, the ultimate impact of our evolving Big Data is in guiding evidence-based content and clinical decision support tools which are targeted to meet the specific needs of an identified patient population, subpopulation, or individual, content which can be pushed to every point of care (hospital, ambulatory setting, patient home, etc.) and delivered in a format appropriate for the specific provider (doctor, nurse, patient, etc.).
As we widen the net of data sources and types included within our analytics models and electronic information systems, we are increasing our ability to hone down, to fine-tune our understanding of specific and populations’ and patients’ risks, needs, and opportunities to improve both the quality and cost efficiency of their care. To provide the most appropriate evidence-based content wherever it needed, whenever it is needed, for whoever needs it.
Peter Edelstein is chief medical officer, Elsevier Clinical Solutions