A Business Case for Clinical Use of Big Data Analytics

The hype around the use of Big Data analytics would lead many to believe that we should be able to understand all things about anything, within any business, as deep or as shallow as we want to go. The reality is that most health care organizations are far away from the use of any high-end analytics systems to provide the insights that clinicians, physicians, and administrators should have available to them.


The hype around the use of Big Data analytics would lead many to believe that we should be able to understand all things about anything, within any business, as deep or as shallow as we want to go. The reality is that most health care organizations are far away from the use of any high-end analytics systems to provide the insights that clinicians, physicians, and administrators should have available to them. 

The core issue is cost. Despite the fact that much of the technology that organizations can leverage for Big Data and Big Data analytics is open source, the amount of time and money it will take implement these systems is far beyond the budgets of most providers. 

That said, we could be quickly reaching a point where the ROI for this technology is just too compelling. Indeed, it could lead to a drastic reduction of risk pertaining to clinical errors, as well as the ability to provide better care by having better data points in combination with deep and proactive data analysis.

For example, a recent study conducted by Kaiser Permanente examined the frequency of blood clots among women who were prescribed oral contraceptive formulas. Kaiser’s analysis found that one formula containing drospirenone increased the likelihood of blood clots by 77 percent compared with women taking other oral contraceptive formulas.

The Big Data analytics success stories continue to flow in. New York-Presbyterian has reduced the rate of potentially fatal blood clots by about a third with the use of the technology. Seton Healthcare Family found out through their Big Data system that a bulging jugular vein is a predictor that a patient admitted for congestive heart failure is likely to make a return trip to the hospital at some point. 

The core idea is that by leveraging Big Data analytics, typically linked into the clinical systems and clinical data, we’re able to consider almost all of the data for analysis. This includes patient history, medications, treatments, and outcome data, together with external data, which can quickly find these emerging patterns of risks, and take corrective action. The result is better care, with obvious moral and financial benefits.

Big Data analytics is rapidly growing. According to research firm IDC, the growth of Big Data was $34 billion in 2012, and will certainly increase this year. This is in part because of increased use in the health-care industry, which is finding more of a direct benefit than other verticals, considering the relative investment in I.T.

Moreover, Big Data has a clear business case for hospitals. By making “meaningful use” of computer systems, hospitals are eligible for millions of dollars in government funding from the Obama administration’s $14.6 billion program launched in 2009 that encourages the adoption of electronic medical records.

But, the path to Big Data analytics for most health care providers will be difficult. Most providers have a hodgepodge of siloed systems and databases that don’t have good points of integration. Moreover, typically there is no common understanding of the data under management and thus it’s difficult to determine the analytical models, even in the abstract.

The best approach is to obtain a data-level understanding of the problem domain you’re looking to address, including the location, structure, and meaning of the data housed within the silos. Then, you need to determine if the data can be leveraged in place, as a node on a Big Data system, or if the data should be migrated and altered in terms of semantics and structure as it moves from the source database or system to the target Big Data system. 

In many instances, the use of public cloud-based Big Data systems should be considered, including the potential reduction in the cost of operations and the speed in deployment. While many providers are nervous about leveraging infrastructure that they don’t own or control, if you’re willing to put some planning into place, as well as pay attention to both compliance and security, public clouds make great platforms for massive amounts of anonymous patient data. The number of server instances can be elastically scaled up, or down, to support the analytical processing required.

The business case for Big Data analytics is rather easy to make for most health care providers, and you can do so on many levels. However, the path to deployment of these systems is costly and complex. Many providers just don’t have the budget or the time to deploy these systems. 

But as time passes, that business case will be too compelling to ignore. Patients will demand it, and I suspect the government will as well. It’s time to get to work. 

David Linthicum is an SVP with Cloud Technology Partners, a cloud computing consulting and advisory firm. David’s latest book is "Cloud Computing and SOA Convergence in Your Enterprise, a Step-by-Step Approach." His Web site is www.davidlinthicum.com/ 

 

More for you

Loading data for hdm_tax_topic #better-outcomes...