3 Must-Read Resources to Understand Big Data in Healthcare

As the New Year begins, we can reflect on the past year’s technology investments’ maturation. Fittingly, the December Harvard Business Review closed 2013 with a spotlight on “Making Your Company Data-Friendly.” Three of the articles provide some edgy considerations for healthcare IT in 2014.

Analytics 3.0

Thomas H. Davenport in Analytics 3.0 says to envision analytics beyond just big data, which he describes as Analytics 2.0. Davenport describes analytics sampling data vast in sources and numbers, creating “customer-facing products, services and features.”

Healthcare delivery’s complexity and depth produces opportunities beyond the manufacture of inkjet cartridges. In healthcare, the “customers” are many and interconnected. Obviously, the first customer that comes to mind is the patient, but so are the medical staff, community, private insurers and CMS.

Big data should allow decision-makers to focus on the customer and value their own technology investments.  If big data is used correctly, every aspect of healthcare service can improve.

Acute care facilities must appoint great physicians to their medical staff. Excellent physicians raise the level of care provided to patients, enhance other physicians’ and providers’ performances and optimize the enterprise’s entire operation. Truly, we have only begun to appreciate the metrics that describe our own and caregivers’ performances.

Data’s Credibility Problem

This article by Thomas C. Redman reminded me of the challenge dirty data poses for value-added healthcare delivery. Redman reminds us that “the quality of data is fixed at the moment of creation.” He also says that the data creator is not necessarily the data user, which may result in poor data capture. In healthcare, the data creator is frequently the end data user, but this still does not protect care providers from dirty data and its consequences.

Anyone who was ever spent time in a Health Information Center, the department formally known as Medical Records, knows the horror of duplicate records and can recite tales of inaccurate face sheets. Despite the best training, system design, personal selection and intentions, patient registration is still a tight rope walk across an erupting volcano.

Registering a patient in the busy ER violates every recommendation for optimal man-machine interface. Add to this environment the problem of the unique name identifier.

Consider Charles Lutwidge Dodgson, better known as his penname, Lewis Carroll. If he lived in modern times, his potential aliases seem endless. Fortunately, most patient access/registration modules are programmed to facilitate correct, unique patient identification. Yet we still create duplicate records. Could big data improve registration?

In the early 19th century, virtually every Briton knew Arthur Wellesley was the Duke of Wellington, or ”The Iron Duke.” But I’m not sure modern health information technology would. This is the realm of unstructured data search. Is it worth the effort to improve it? Absolutely! It’s more harmful to lose a critical ER patient than to show a hospital’s board a bad data point.

You May Not Need Big Data After All

The last article to launch us into 2014 is “You May Not Need Big Data After All.” The authors observe that “ … big data has been hyped so heavily the companies are expected to deliver more value than it actually can.” Expecting big data to add profit when many companies fail to use their existing data to optimize management is a bad career decision.

Providing decision-makers with effective information leads to better decisions. Simply put, evidence-based management often beats seat-of-the-pants management, which is a good thing for those of us in healthcare information technology to keep sight of as we power into 2014.