.excerpt-thumb {display: block !important;}

Survey by KPMG on Reliability of Healthcare Analytics Mirrors Findings by Reltio in Life Sciences

Click here for the full article at HealthITAnalytics.com

Why Can’t Providers Trust Their Healthcare Big Data Analytics?

By Jennifer Bresnick on July 29, 2015

A recent survey of life science organizations, conducted by Reltio, found similar results: three-quarters of organizations stated they have deep concerns about the integrity of the big data they use for research and development. EHR data integrity is such a pervasive concern that it has also featured highly on a number of ECRI patient safety and health IT hazard lists over the past few years.

Healthcare organizations have not yet overcome the enormous difficulties of developing a big data analytics ecosystem in the healthcare industry, according to new research from KPMG, and continue to suffer from a lack of trust in the integrity and accuracy of their data. 

While healthcare organizations are eager to achieve higher levels of productivity, market share growth, and financial savings by adopting clinical analytics, decision support, and data-driven population health management technologies, fundamental problems with interoperability and data integrity have stymied their progress in comparison to other sectors of the economy.

The cross-industry survey found that entities of all kinds are integrating big data into their strategic visions.  Unsurprisingly, industries such as retail, banking, and telecommunications are far ahead of their healthcare counterparts when it comes to leveraging big data for actionable insights – 92 percent of all organizations participating in the survey said they use big data for marketing purposes, while 41 percent have used their data assets to target specific types of customers.  But healthcare has not been able to achieve the same high levels of adoption and the resulting financial benefits just yet.

“Clearly certain sectors are ahead of others in terms of where they’re embracing data analytics and their level of maturity. Any sector dealing with customers is probably on the forefront because it’s pretty intuitive that the more you know about your customer, the better you can tailor what you sell and, in doing so, the more you will sell,” said Brad Fisher, a partner at KPMG. “Then there are others that have made massive strides over the past few years but are still probably behind.”

“Healthcare is a perfect example of an industry that is only now moving towards full digitization, and healthcare providers, payers and even governments are soaking up all they can in data and analytics to help them transform,” he continued. “But while the sector is certainly moving quickly to catch up, the fact remains that their level of maturity – and the sophistication of their systems – may be inferior to those data-driven sectors, such as retail, that have been using data and analytics for years.”

Healthcare isn’t just coming a little late to the party.  The industry has its own unique obstacles to overcome.  Trust in the validity, accuracy, and completeness of big data is at the top of that list, the survey found.  Seventy percent of healthcare organizations expressed very low confidence in their data integrity, compared to just 45 percent of the banking industry.

This is not a new issue for healthcare providers, most of whom continue to struggle with EHR documentation, a lack of health data interoperability, and a fragmented delivery system that hasn’t quite figured out how to bring health information exchange to scale. 

recent survey of life science organizations, conducted by Reltio, found similar results: three-quarters of organizations stated they have deep concerns about the integrity of the big data they use for research and development.  EHR data integrity is such a pervasive concern that it has also featured highly on a number of ECRI patient safety and health IT hazard lists over the past few years.   

Why can’t healthcare’s big data be trusted?  It all starts with the EHR.  Designed as transactional systems to record specific information that is used to inform providers in a specific, time-limited circumstance, big data is basically a byproduct of the healthcare system.  

For a long time, EHRs have been modeled on paper charts that sat dormant in file cabinets when not actively in use.  EHR data has largely done the same.  Big data analytics and dynamic, real-time population health management were simply not a concern when most providers were exploring their first health IT purchases, and organizations that implemented more-or-less static, episode-based documentation infrastructures have had a difficult time retooling their systems to meet more modern demands for data liquidity, interoperability, analytics, and health information exchange.

Poorly optimized EHRs have let messy workaround creep into the documentation process, fragmenting data while frustrating human operators.  Data siloes remain a critical concern as providers strain to generate complete and longitudinal portraits of their patient populations.  New attitudes towards the creation, collection, and transmission of data are not easy to cultivate, and the vendor community is just starting to offer affordable, interoperable products that meet advanced big data analytics needs.     

While the healthcare industry has made huge strides towards remedying these problems in a very short time frame, they are still at the beginning of the process.  The challenges are many, and their impact runs deeply through an industry that has more than a few competing health IT struggles on its plate.  Healthcare organizations must embrace a stronger sense of information governance if they are to make headway against the tide of untrustworthy data, and many are doing so.

Healthcare providers are willing to learn from the trials and successes of other industries, and would do well to remember that the learning curve for big data analytics is a steep one.  Healthcare organizations that make the commitment to data-driven population health management, patient coordination, and accountable care can see rapid and meaningful results, but they must focus on building trust and confidence in complete, accurate, easily accessible data if they are to cash in on the promises of big data.