Posted by: Jenny Laurello
Big data, EHR application design, EHRs, user interface
I highly recommend reading Dr. Rick Weinhaus’ article on HIStalk about user-centered design of electronic health records (EHRs). I completely agree with his premise that there is much work to be done in making EHRs more useable, since rows and columns of numbers and complex navigation are not only annoying to end users, but can also lead to data entry and interpretation mistakes.
So how can we begin to understand what works in EHR user interface design, and what doesn’t? I suggest we put “big data” to work and as a first step to understand how end users are using their current applications, and what usage patterns are yielding the best clinical outcomes at the lowest cost.
There is a great deal of excitement about the potential for big data in financial services, communications, retail and other industries. Informatica is a thought leader on big data. However, health care organizations have not necessarily jumped on the big data bandwagon. This is understandable as even on a national scale in health care, we don’t have the same big data challenges that compare with 900 million Facebook users updating their status every hour, or a wireless provider tracking billions of cell phone calls per day. But if we look closer to home, and at far more pedestrian log data, we actually have a treasure-trove of big data that can help us redesign our EHRs to deliver more efficient, better quality care.
The log data I am referring to are the application logs that capture how physicians and other end users interact with an EHR application. This log data can capture the activities of thousands of end users, spending hours a day pointing and clicking, telling us exactly how they navigate and use their EHRs to care for patients. But what’s interesting is that virtually nobody is looking at this valuable cache of data that we can learn from. While a great deal of focus is placed on using EHR data to understand objective findings, effectiveness of interventions and evaluate outcomes, completely absent from the analysis to date is looking at how the design of the application itself influences outcomes.
By comparison, every single major online retailer tracks and profiles each visit to their website to understand end-user behavior and what design elements yield the most revenue. Nowhere in the health care field do I see anything similar being applied to EHR application design, where we profile how providers who get the best outcomes utilize the system, and then purposefully approach the user interface design to passively reinforce the desired behaviors that result in the best outcomes.
The exciting news is that big data techniques from other industries make this problem achievable if someone were to choose the path of investigating log data captured by EHRs. Unfortunately, most application vendors and IT shops look at the huge volume of data as a problem to be managed (e.g., stuffed into log files, written to tape, hope nobody ever asks to use it…) rather than an opportunity to be leveraged.
That said, I don’t have a great answer to the question Dr. Weinhaus poses in his blog of “What exactly is user-centered design?” However, I do think the idea of investigating user interaction data to yield insight into how end users are actually navigating EHR applications today, and what types of behavior is yielding the best clinical outcomes and fewest errors, certainly has merit and hope for the future.
Please visit www.informatica.com for more information.