Seeing into the future used to be the stuff of fantasy and science fiction, but as we become more adept at collecting and analyzing data, it’s starting to look a lot less fantastical and a lot more scientific. For years now, marketers have been honing their skills at forecasting what products and services an individual is likely to buy based on income, age, gender and cultural background, as well as personal interests and past purchases — all of it gleaned from data already out there in the digital universe. You’ve no doubt seen more posts appearing on your social media accounts that are directed at you personally. Who knew that you were into cooking or favored a particular genre of TV show or were a big fan of country music? Companies of all kinds are tracking that information and then betting that by crafting messages based on it, you will open your wallets to whatever it is they’re promoting.
Of course, predictive analytics holds much promise beyond just the purely commercial. For instance, some child welfare agencies are using data to identify kids at high risk of abuse so that they can keep a close eye on them and intervene if necessary.
And, as we reported last April, combining predictive and prescriptive analytics holds enormous potential for health care providers to improve efficiency and quality of care, and even save lives.
By tracking historical patterns and emerging trends, hospitals can predict disease outbreaks in their communities and brace for an influx of patients. Doing so before or in the earliest stages of the outbreak allows hospitals to be fully prepared by ensuring there are enough medications and staff on hand.
Predictive analytics also can enhance population health management by painting a detailed picture of health and wellness issues in a whole community or cohort of the community. What disease patterns can you spot across generations, what lifestyle traits are prevalent and what interventions are most likely to work positively for a specific group?
While some hospital leaders may find the very term “predictive analytics” intimidating, many are testing ways to put the concept into practical use. They are starting out by focusing on narrow but critical challenges, such as patients who are at higher risk for falls or are most likely to require rehospitalization.
In this month’s cover story, Rebecca Vesely takes a close look at how Indiana University Health tackled one of its most vexing problems — a high rate of central-line infections among its patients. By embracing predictive analytics, IU Health now knows which patients are most at risk for CLABSI, and already it has significantly cut the rate of infection.
As providers struggle to figure out their place within the transforming health care system, predictive analytics also provides useful insights into how their markets will change over time and which service lines will be most in demand in the future. Will the population of a community become disproportionately older? What will the ethnic or socioeconomic makeup of the community be? Will patterns emerge from historical data about future wellness and medical needs and how a provider can meet those needs either independently or by partnering with others?
We’ve been hearing a lot these past several years about the promise of Big Data and about machine learning and data mining. Thanks to organizations willing to pave the way, that promise is now being realized.