Healthcare organizations are becoming incredibly data-rich but are in fact embarrassingly insight-poor, because only a fraction of the available data is being used to its fullest possible extent. The increasing volume, variety and velocity of healthcare data, plus the potential ways in which it can be used, has highlighted an analytics gap within organizations.
Advances in healthcare information technology are resulting in massive increases in both the volume and complexity of healthcare data. According to the Institute for Health Technology Transformation report on big data in healthcare, healthcare data in the U.S. alone has well exceeded 150 exabytes and is on track to reach the zettabyte and even yottabyte scale.
This growth in data is not alone attributable to electronic health records or other clinical systems. In fact, big data in healthcare is said to comprise data from multiple sources, including the following:
- Clickstream and interaction data from social media and the Web;
- Machine-to-machine data, such as readings from sensors and other medical equipment;
- Transaction data, such as health insurance claims and billing;
- Biometric data, ranging from fingerprints, genetics, vital signs and even diagnostic imaging data; and
- Data generated by humans that is recorded as structured and unstructured data in electronic medical records and other source systems.
The improvements in healthcare information technology offer exciting potential to improve the quality, efficiency and safety of healthcare, and to aid in the discovery of new methods and tools to detect and treat disease. But the growth also poses a challenge in that healthcare organizations must be ready and able to analyze and utilize the data to achieve the many potential benefits of healthcare analytics.
It is clear that a significant portion of data being collected by or available to healthcare organizations ranges far beyond the financial, clinical and operational data familiar to most healthcare analysts and managers. The process and clinical data that typically has been the focus of analysis at most healthcare organizations (HCOs) can now be augmented by streams of diagnostic, vital signs, location tracking and genomic data, to name but a few examples. Even trending data from sources such as Google and Twitter can be used to draw attention to emerging diseases, outbreaks and population health management, and even patient satisfaction.
The availability of such data does not necessarily mean it is immediately capitalized upon by HCOs. New data sources are coming online more quickly than can be integrated into existing data warehouse structures and put to use using only traditional integration and data analysis methods. Any potential data source that is left untapped is a lost opportunity for more deeply understanding and improving the clinical and operational performance of a healthcare organization.
Signs of an analytics gap
There are many signs of an analytics gap within healthcare organizations, although they can be cleverly masked as "the way decisions have always been made." An organization may be awash in dashboards, reports and other information, yet may still not be making significant progress toward achieving its quality, performance, safety or clinical research goals. But data and reporting is really only one part of the healthcare analytics equation. How well data is analyzed and the resultant insight applied to decision making are key indicators of whether a gap exists.
An analytics gap might exist within your HCO if, despite continued adoption of health IT and increasing amounts of data being available, important operational and clinical decisions throughout the HCO are being made without the benefit of accurate, timely and readily available insight and evidence. Although not an exhaustive list by any means, below are several common signs of an analytics gap within an HCO:
- Lack of a data-centered culture. Perhaps the most worrying sign of an analytics gap in healthcare is not technical at all. In some organizations, a necessary culture of "evidence-based decision making" has not taken hold. This may be because decision makers do not trust available data -- stating things like "we know the data is garbage, so we don't use it" -- or because of the fear that big data analytics is more akin to Big Brother," wedging in to remove room for human judgment and dictate every action taken by clinicians, and even administrators. Healthcare organizations must embrace the information and insight available to them to enhance and optimize decision making by administrators and clinicians, not replace them. No level of analytic sophistication can help an organization that does not fully embrace analytics in the first place.
- Poor decision-making tools. Healthcare organizations that have become data-centric (and by extension, more patient-centric) and desire to make timely decisions based on the most recent and best evidence available may in fact be stymied by poor decision-making tools available. These suboptimal decision-making tools, such as reports, dashboards and other "analytic" applications, tend to be static, retrospective and often out-of-date.
- Analytics are not pervasive. Although many HCOs have pockets of analytic excellence, most are often contained within silos, such as a program or department, and are typically operated from a single-source system. The value of analytics within an HCO can only be realized, however, when timely and relevant information becomes available to everyone throughout the organization in a format that is usable and insightful to the user.
HCOs that fail to address this gap in analytics capability will be unable to gain the deeper insight into their clinical and business operations required to transform healthcare. They may also be missing out on significant clinical breakthroughs. In an environment where HCOs face increasing regulatory, financial and quality pressures, failure to leverage data analytics to their fullest extent is doing a disservice and, quite possibly, doing harm to the most important stakeholders in healthcare -- the patients themselves.
Part two of this article will uncover several of the major causes of the analytics gap in healthcare and will discuss key mitigation strategies.
About the author:
Trevor Strome, M.S., PMP, leads the development of informatics and analytics tools that enable evidence-informed decision making by clinicians and healthcare leaders, and his experience spans public, private and startup-phase organizations. A popular speaker, author and blogger, Strome is the founder of HealthcareAnalytics.info, and his book, Healthcare Analytics for Quality and Performance Improvement, was recently published by John Wiley & Sons Inc.