Academic journal article Frontiers of Health Services Management

Priming the Pump for Big Data at Sentara Healthcare

Academic journal article Frontiers of Health Services Management

Priming the Pump for Big Data at Sentara Healthcare

Article excerpt

DEFINING BIG DATA

Like many technology terms used today, big data is not easy to define (Press 2014). Some consider it to mean a data set that is so large it presents significant logistical challenges regarding manipulation and management. This is a subjective definition, identified as such in a 2011 study by a multinational management consulting firm that defined big data relative to an organization's size and capabilities (Manyika et al. 2011). For small, rural hospitals, for example, managing an EHR can (and in many cases does) pose a significant challenge. These hospitals may consider the information contained in an EHR to be big data. Sentara's EHR database includes about 7 terabytes of data. We store approximately 50 terabytes of data (including the 7-terabyte EHR database) for analytics. When we include other types of data, such as voice (i.e., physician dictation systems), unstructured data (e.g., physician documentation, scanned documents), laboratory test results, and imaging, the total data storage load increases to about 100 terabytes. Although it may seem like a lot, we do not consider this to be big data. To put size into perspective, one large health system with a database of 9 million patients reported that it stores more than 30 petabytes (30,000 terabytes) of data-more than three times the digitized storage of the Library of Congress-and its database grows by approximately 2 terabytes per day (Jaret 2013).

Sentara considers health data collected in an EHR, data shared via a health information exchange, and financial data to be systems of record (Bersin 2012). We rely on these systems to document and support care delivery and to run our day-today business. Although systems of record store a lot of information, Sentara does not consider this information to be big data.

Systems of engagement in healthcare include those for collaboration, social networking, and learning (Bersin 2012; Huston 2015). These systems can be used to reengineer the patient experience, increase patient satisfaction, create a more productive workforce, deliver more effective healthcare, and provide targeted, personalized education for healthcare providers and patients. For example, the MyHealth MyChart patient portal at Sentara provides patients with secure access to their health information via Epic's MyChart. The explosion in the use of systems of engagement for consumers, both in healthcare and in other consumer-driven sectors, mandates the integration of these systems with traditional systems of record. Ideally, the systems of engagement will be layered on top of the systems of record, with a welldesigned interface between the two. Even technologically advanced industries are continuing to evolve (e.g., the Internet of Things), and healthcare will follow suit.

At Sentara, big data refers to the unique ways internal and external data are combined and leveraged to improve patient care, streamline operations, and decrease costs.

TURNING DATA INTO ACTION

In healthcare, sources of big data include EHRs, wearable sensors (e.g., pedometers, blood pressure monitors), insurance plans, consumer purchasing behaviors (credit card use, credit scores), environmental reports (weather, pollen), educational records, genetics, dietary habits, employment histories, hobbies, transportation patterns, socioeconomic data, social media (tweets, texts, social media posts), and information about emotional status. Creating big data sets is not a trivial undertaking, and validating and mapping the data can be more of an art than a science, requiring specialized expertise not typically found in health systems.

Furthermore, the data are useless unless the organization can create actionable outcomes (e.g., prevent admissions via patient self-management, increase smoking cessation by means of incentives) through analysis. Thus, along with systems of record and systems of engagement, systems of insight are needed to merge business discipline with technology to harness insights and consistently turn data into effective decisions (Flopkins 2015). …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.