Over the past 5 to 10 years, hospitals and physician offices have been in a mad dash to implement electronic health records (EHRs) in order to meet governmental regulatory requirements.  Now that most projects are either complete or well on their way, what are we doing with all of the data that EHRs promised to generate?

From my experience as a physician at a large academic medical center with one of the top EHRs in the industry, I have to say the data wasn’t readily available in any sort of usable format.  If I wanted performance data, cost data, outcomes data, etc., I would have to either track it myself or become good friends with someone in IT.  Both of these options required something I didn’t have a lot of: time.

But I thought the new EHR was going to make this easy for me?  It didn’t.

It turns out that EHRs are mainly designed to be transactional.  This means they perform operational tasks like ordering tests, reporting results, documenting procedure notes.  They exist primarily to facilitate data entry, gathering, and retrieval, but most lack any robust analytic capabilities.  This leaves the responsibility for data analysis to the end user, especially in organizations that lack significant internal IT infrastructure for clinical analytics.

So while many EHR vendors state that they have data analytics as part of their platform, most systems are fairly rudimentary and require a lot of on-site set up and maintenance.  When you look at the massive investment most EHRs require, the majority of institutions (even large academic centers) forgo budgeting significant resources toward analytics and focus their goals mainly on the transactional side of health IT.

This leads to inadequate resources for clinicians interested in performance improvement projects that go beyond the usual parameters that are measured for existing reporting requirements such as sentinel events, patient safety measures, etc.  It also leads to internal strain on IT departments and CMIOs that are tasked with developing data analytics platforms with widespread utility to the health system that they are clearly not capable of producing.

There are many vendors who assert that they are clinical data analytics providers, but there are really just a handful of companies that are competitive in that space today.  Isn’t it a better use of resources for health care systems to implement a proven analytics platform and have their internal IT infrastructure co-implement and support it rather than waste money and personnel resources developing an inferior home-baked version?  This would free up CMIOs and IT specialists to take a more active role in identifying strategies to use data analytics to improve patient care, rather than sitting behind a computer developing software.

Alexandra S. Brown is associate director, Healthcare Delivery Institute, HORNE LLP.