Human-computer interface progress vital to success of EHRs
Steve Jobs' passing last week has triggered myriad reflections on his immense contribution to the modern world. While much emphasis has been placed on Apple's recent trendsetting products--the iPod, iPhone and iPad--the signal contribution of Apple under the leadership of Jobs and Steve Wozniak was to make the personal computer practical and useful. Beyond the Apple operating system itself, the invention of the Macintosh computer--which used a mouse-based graphical user interface derived from an experimental Xerox product--radically redefined the relationship between humans and computers.
Despite all the technological advances since then, however, physicians continue to struggle with that relationship. I recall that back in the 1990s, when I began covering this space, there was considerable disagreement among experts about whether electronic medical records were ready for prime time. Clement McDonald, MD, who helped pioneer health IT at Indianapolis' Regenstrief Institute, once told me that doctors would never accept EMRs (now known as EHRs) until they could dictate their notes and have them transcribed automatically into discrete data.
In retrospect, that prediction probably was a shade too pessimistic. According to the Centers for Disease Control, 10 percent of office-based physicians were using a fully functional EHR system in 2009. Preliminary results of the CDC's 2010 survey show that the percentage ranged from 10 to 27 percent in various states. But the vast majority of doctors still were not using an EHR for most of their clinical work.
While the cost of EHRs has long been a problem for physicians, the federal incentives for Meaningful Use have lowered that barrier. A more significant problem today is the inability of many physicians to use EHRs without losing productivity. Steven Waldren, MD, director of the Center for Health Information Technology of the American Academy of Family Physicians, recently told me, "EHRs are more usable [than they used to be], but they still have a long way to go."
Can EHRs, if properly used, increase efficiency and improve the quality of care? There are plenty of arguments on both sides. Moreover, some EHRs work much better than others, and EHRs are easier to adapt in some care settings and specialties than others. But with those caveats, EHRs could stand improvement in a few areas, including:
- Initial data entry: One of the biggest challenges to adopting EHRs is transferring data from patient charts into the electronic record. Scanning and manual chart abstraction are both expensive and difficult. Mark Anderson, a health IT consultant in Montgomery, Texas, suggests that perhaps practices should get lab interfaces first and automatically bring in their reference labs' historical results. That would allow them to build problem and encounter lists without data entry, he notes. But EHR vendors would have to get on board to make this possible.
- Documentation: Many physicians hate the point-and-click templates that are standard in most EHRs. They're hard to customize and slow for documenting physical exam findings. With the work currently underway by Nuance and the University of Pittsburgh Medical Center, using IBM's Watson technology, perhaps natural language understanding will get a jump start, and McDonald's prediction will come true, after all.
- Clinical decision support: The leading EHRs have alerts about adverse drug interactions, prompts to provide preventive and chronic care, and other forms of clinical decision support. But "alert fatigue" leads many physicians to ignore alerts, especially when they're frequently wrong. More research needs to be done on how to ensure that EHRs trigger only essential alerts.
- Interoperability: The lack of interoperability among disparate EHRs has long been a sore point for physicians. One drawback, Waldren pointed out, is that when clinical summaries are sent from one EHR to another, they may go into the receiving EHR as documents, but the structured data in them does not go into the appropriate fields. That's "real frustrating" for doctors, he noted. This doesn't seem like an insoluble problem in an age when computer whizzes are constantly inventing new products and trying to become the next Steve Jobs.