Health system leaders are between a rock and a hard place.
They have invested millions of dollars in an electronic medical record (EMR) system that their physicians likely hate because it slows them down, adds work to their day and doesn’t help them deliver better care.
Ripping and replacing the EMR is not an option for most hospitals and health systems due to cost and disruption. That means enterprises have turned to third-party apps to bolster their EMR functionality that hasn’t changed the core application, simply stacking technology on technology on technology. Other health systems are using scribes, but that, too, doesn’t solve the essential problem that physicians simply do not have reliable, structured and actionable information easily accessible at the point of care.
Now health system leaders are hearing about artificial intelligence (AI) and machine learning technology to solve their EMR usability challenges. Yet neither of those solutions solve the inherent problems with data quality that most physicians face daily.
Supplying physicians with more data will not measurably improve patient care. Instead, what providers need is contextually relevant and trustworthy medical information that offers insight into a patient’s medical history and current problems. Data is just data – and not actionable information – unless it is logically compiled and transformed into information.
Lack of structure leads to lack of coherence
EMRs do have very tangible benefits, most notably, all a patient’s data is stored in a single record. But unless it is well-managed, clinicians end up drowning in data and are forced to waste time sifting through irrelevant information to find what is usable.
The reason for this endless searching is as much as 80 percent of the patient information within EMRs is stored as free-text and not mapped to data standards. Such a problem isn’t efficiently solved by adding scribes either. As Atul Gawande, MD, pointed out in his New Yorker article, physicians still need to review and correct the final chart after the scribe has entered it. At the same time, physicians may miss critical alerts at the point of care if the scribe has all the EMR screen time in the exam room.
Even with scribes, the majority of enterprises are not able to clean and standardize the huge amounts of clinical data created over the last couple of decades as EHR adoption has grown. Data capture can be time-consuming, but more importantly, providers need proven methods to help them easily interpret relevant data at the point of care.
Turn data into actionable information
That is why implementing AI or machine learning technology in most EMR systems is akin to buying a sports car for a child who just learned to ride a bicycle. Many buzzed-about technologies are far from maturity and have yet to deliver any tangible results at hospitals or health systems. What’s worse, they may compromise clinician productivity, or even more troublingly, compromise patient care.
For example, providers would likely love solutions that leverage natural language process (NLP) and be able to convert dictated chart notes to free text, and then free text to data that’s stored in an actionable format. Unfortunately, the error rates for converting speech to text to data are, at best, between 8 and 10 percent – which is unacceptably high for safe and effective clinical decision-making.
Instead, CIO and other health-system executives’ priority should be to seek established, widely implemented technology that intelligently identifies and interprets the disorganized and complex arrays of medical information from multiple sources to support providers in their goal to deliver better care. More specifically, the technology should convert data into structured, actionable formats, such as ICD-10, SNOMED, RxNorm, and LOINC, and present the relevant elements for each patient encounter.
Finally, the technology needs to seamlessly deliver this filtered, high-value information in the 10 minutes or less that physicians typically have with patients at the point of care. That means leaders need to witness such functionality in the real-world at a peer organization, not just in a product demo.
Listen and collaborate
Perhaps most importantly, executive leadership needs to listen and collaborate with their physicians on these issues to find third-party applications that can rapidly deliver clinically relevant and useful data at the point of care. For years, healthcare senior executives have taken the view that if physicians do not like the enterprise’s technology or processes that they can find someplace else to practice. That isn’t a sustainable model.
Clinicians must be given a larger voice in decisions that impact their workflows. If not, organizations risk losing more physicians to burnout or to competitors who understand that the importance of supplying clinicians with the tools they need to maximize the value of their education, skills and experience and deliver optimal outcomes.
Through interdepartmental collaboration and optimized clinical systems, physicians will be able to easily identify and interpret all the disorganized data within a patient’s chart and make safe and effective decisions in less time and with less frustration. Together, executive leaders and physicians will finally have the key to translate disparate data into relevant information that delivers improved outcomes and better financial performance.
About the author:
Jay Anders, MD, is Chief Medical Officer of Medicomp Systems