The Big Data, No Data Paradox

The current premise is that the more data an organization can access and analyze, the greater the insights that can be acted upon – so focus on data, data and more data. This approach loses sight of “NO” Data, though, which if detected can often represent even more value. But there’s a paradox: “NO” Data can become more visible in a world of Big Data, while blindly focusing on Big Data can diminish the visibility and potential value of “NO” Data.

The push for digital medical records and the ability to exchange that information across the healthcare delivery system is foundational to the industry and long overdue. Combine this with the accelerated amount of medical discovery such as the human genome, increasing number of therapeutics, smarter web-enabled medical devices, and sensors, and the result is a tsunami of data that organizations are struggling to manage, rationalize, expense, monetize and use to deliver meaningful results.

Hidden away in this tsunami are many instances of “NO” Data that Big Data may now make more visible, and if detected, much more valuable. Here are some “NO” Data scenarios in healthcare, and how they may generate quality and cost of care value.

 

NOT Data: The Search for Missing Data

NOT Data is probably the most perplexing of the “NO” Data healthcare scenarios. One example is pre-authorizations and the claim that never arrives.  Although approved as medically necessary, the patient did not obtain the required diagnostic tests, medical procedures or services, durable medical equipment, or prescription drugs. The ability to detect this missing data in a timely fashion could trigger appropriate actions that could prevent medical issues or even save lives.

Another instance is a provider’s inability to easily monitor and follow up with patients that did not receive a flu vaccine even though the physician’s EHR system informed patients to get the vaccine, particularly high risk patients such as those with congestive heart failure or COPD.

Other examples of NOT Data include prescriptions that are not renewed or refilled, diagnostic test results not received, dental cleanings not performed, and office visits missed to name just a few. Although NOT Data has always existed, in many instances it is now easier to detect its absence because of Big Data.

 

NULL Data: Finding the Value

NULL Data is an instance where there is a physical place for the data to reside, but the result in that space either purposely or by default is a null value. Here are two examples at the far opposite ends of the NULL Data spectrum.

First there’s the case of a high risk patient that has co-morbidities. The individual is under a highly customized, case management regimen provided by the primary care physician. Electronic Medical Record systems can record events such as appointment schedules, test results, and prescription orders but until the time of an occurrence  contains null values. The null value is not an error per se, but represents the absence of an event or result. However, detecting the null value can at play a critical role, particularly with someone that is high risk, such as the canceled appointment that is not rescheduled, test results not reported, and authorization for prescription refills not requested.

At the other end of the NULL Data spectrum are the periods of silence in a conversation, real-time or recorded. There are technologies that can now analyze these white spaces along with cadence and other verbal characteristics and determine potential medical symptoms or behavioral intentions. For example, if a health plan is offering a smoking cessation program to admitted smokers, it would be most efficient to make a second call to those that are truly considering, but unsure about attending.

 

NAUGHTY Data: Detecting Error or Fraud

NAUGHTY Data is valid data in and of itself, but when put in the context of other data its veracity comes into question. The end result is either error or fraud.

Take, for instance, a health plan member who, at the request of their surgeon, calls the member service center to see if bariatric surgery is covered for reimbursement, and if so, wants to find out the qualifying conditions. In this case, natural language processing technology could detect “bariatric surgery,” setting a trigger to watch this member’s future claims for an appendectomy submission -- potentially a fraudulent coding for bariatric surgery. 

Or, how about adding a family member to a plan of benefits that they are not eligible for? For instance, someone has a child over the age of 26 that becomes unemployed and loses their employer’s health insurance coverage. The parent enrolls the adult child in their plan of benefits with a contrived birth date. The child provides the health plan with their Twitter account or access to Facebook as part of a promotional offering.  That social media information might reveal that a great time was had by all at the child’s 30^th birthday celebration.

Unfortunately there are many instances of NAUGHTY Data which go undetected. The FBI estimates that healthcare fraud costs American taxpayers $80 billion a year. That is more than naughty, that is out and out big, bad and ugly. Big Data approaches can provide more context and reduce the time to detect NAUGHTY Data, resulting in faster answers and actions to reduce the cost of fraud.   

 

Carl Ascenzo heads the healthcare practice for NewVantage Partners, a consulting firm that  provides expertise and guidance to Fortune 1000 business and technology executives who are seeking to leverage data and analytics to gain business insights and  derive business value. He was formerly CIO of Blue Cross Blue Shield of Massachusetts where he led all aspects of information technology. Reach him at cascenzo@newvantage.com.