5 Guidelines To Avoid Analysis Paralysis With Big Data in Healthcare

5 Guidelines To Avoid Analysis Paralysis With Big Data in Healthcare

Senior Vice President, Cash Forshee at Medalogix outlines 5 guidelines to avoid analysis paralysis with big data in healthcare for healthcare organizations. 

The global healthcare industry generates approximately 30 percent of the world’s data.It is estimated by 2015, the average hospital will generate 665 terabytes of data. Every initial homecare visit creates 1,185 points of data.

With so much data available, healthcare visionaries realize the immense benefits of adopting technologies to analyzing their data:

  1. Care improvement: When experienced care providers are equipped with data-driven insights like the top 20 percent of patients most at risk for hospital readmission, they can provide more effective care and keep patients in their homes. The right analytic software and data plan can be the difference between speedy recoveries or additional hospital stays.
  2. Reform incentives. Provisions like the HITECH and Meaningful Use standards­ both require extensive technological adoption from health providers.

Unfortunately, the healthcare industry is faced with a growing epidemic of technologies that fail to provide actionable, meaningful insight. The consequence: big data analytics’ benefits are overshadowed by heaps of irrelevant metrics, presented in indigestible reports, and are of little use to clinicians or their organizations. This epidemic is known as analysis paralysis.

Here are some guidelines healthcare organizations should follow to generate insight that is digestible, applicable and useful to businesses:

1. Use your data. We’ve discovered that agency specific data is more accurate in predicting patient risk than state, nationwide or proprietary benchmarks. If we’re aiming to discover patients most at risk for hospital readmission, we must take into account a venue’s staff, geography, demography, clinical capability, and other factors beyond the clinical composition of the patient. Those factors would be overlooked using data sources other than your own. If you want to maximize data to improve care outcomes, you need to assess your data.
2. Find predictors by understanding workflow. Understanding under what context, and in what workflow your data is generated is essential to effective analysis. A mapping exercise that takes into account your clinical workflow, where data is captured, and how the latter travels throughout your process is essential in identifying risks and opportunities for improvement. Understanding your workflow in advance, an analytics expert can create tools that better accommodate your needs.
3. Generate actionable insights. In industries like health care—where minutes matter—it’s critical your data analysis can deliver insight that clinicians can immediately understand and use. Endless reports and dashboards frustrate people, as no one has time to analyze them while also caring for patients. Your analytics team must assess every application enhancement with a single question: How will having this data improve my organization’s work, and how can I provide it to clinicians and business decision makers in a meaningful way?

Focus on this area by creating a one-page summary that ranks the relevant results. The summary should be the first page presented after logging in, so clinicians can quickly incorporate the information while reviewing their care plans

4. Foster usability. Just as users don’t have time to delve into pages of analytical reports, they don’t have time to read your technology’s manual. The experience of presenting data must be intuitive if you want to positively affect business and care. Two clicks of the mouse or two seconds—that’s the most amount of effort technology should require of the user. Usability may be the most important feature of any business intelligence application.
5. Monitor and optimize constantly. Data and workflows change—so must your analytical models. Constantly analyze, test and if need be, alter your models so you deliver the best output, and always assess your analytics in the context of the clinical workflow in which it is created.

Data is indeed valuable, but it’s the insights unveiled by data that benefit patients and organizations. If your data assessment tool is not designed with your specific organization in mind, it only delivers a bunch of metrics that stifle meaningful change.

What are you thoughts on analysis paralysis? How do you suggest avoiding it? Comment below or email me at cash@medalogix.com.

About the Author

Cash Forshee is the Vice President of Medalogix, provider of unmatched predictive insight into current patients where he focuses on cultivating new partnerships with leaders in the health care sphere and further developing the Medalogix software to meet the evolving needs of the health care industry.

Prior to joining Medalogix, Forshee gained three years of sales and development experience at HealthStream, a health care technology company, where his culminating role was as director of their sales acceleration office, developing and executing new technologies to support their sales organization.

 

Featured image credit: IBM Events via cc

5 Guidelines To Avoid Analysis Paralysis With Big Data in Healthcare by

Get in-depth healthcare technology analysis and commentary delivered straight to your email weekly