“The widespread use of electronic health records (EHRs) in the United States is inevitable. EHRs will improve caregivers' decisions and patients' outcomes … But inevitability does not mean easy transition.”
-- David Blumenthal, M.D., and Marilyn Tavenner, R.N., writing in the New England Journal of Medicine in 2010
Peter Greene, M.D., chief medical information officer for Johns Hopkins Medicine in Baltimore, often uses the above quote in presentations to remind listeners what a complex undertaking an electronic health record system rollout is. He has enterprise-wide leadership roles at Johns Hopkins in the deployment of Epic and in establishing a data governance program.
He used the quote again in a PowerPoint slide during a keynote speech at last week’s 2014 Federal Healthcare Forum in Washington, D.C.
Recognizing the level of complexity of the transition to electronic records, Greene said the adoption rates achieved to date are a reason to celebrate. “What an extraordinary contribution the HITECH Act has made to healthcare,” he said. “I have been nose-down getting Epic deployed at Hopkins over the last few years, and what a difference it has made to have that infrastructure being built out across the system.”
He noted that close to 90 percent of hospitals are participating in the meaningful use program and even on the ambulatory side, more than 70 percent have some form of EHR. But Greene said recent surveys suggest that many physician offices aren’t ready for Stage 2 of meaningful use.
“As you plot all the capabilities that are important for meaningful use Stage 2, and you ask how many are ready to participate in Stage 2 and have a system with all the capabilities, that number was 13 percent of physicians,” he said. “So my concern is that we have made wonderful progress on the acute side, and with large medical practices that are affiliated, but we may grow a bit of a digital divide in that outpatient provider community. Those folks won’t have systems that provide good care and they won’t be reachable with our interoperability efforts as we try to send summaries of care and other information. The curves are going in the right direction, but I think we need to be responsive to figure out how to bring the whole industry along.”
Greene also used Hopkins as an example of how difficult the transition has been, even for a sophisticated academic medical center. For years Hopkins had been focused on quality in the acute-care setting, but its challenge was how to deliver on continuity of care as it expanded to 30 outpatient clinics, ambulatory surgery centers and five hospitals. “We began to realize we wanted to merge these worlds, but they were on a patchwork quilt of information systems, all of which had some level of integration but not enough,” he said. “Sadly, the patients thought it was better integrated than it was. Many didn’t realize it was our clinicians who were the interface, looking in multiple systems and gluing them together. They would much rather be spending that time with their patients.”
In order for Hopkins to be centered on the patient, it was clear they had to have a single record across the enterprise, he added. “It was not just the medical record; you want the protocols, the orders, the decision support, the reminders, everything the same across the enterprise.”
It is crucial not to underestimate the complexity of healthcare, Greene said. “I have seen wonderfully successful people and technology strategies fail miserably when they come over to healthcare.”
The Epic deployment at Hopkins involves 20,000 users, 400 types of users roles. 93 clinical specialties, 460 clinical departments, 200 message folder types, 561 insurance plans and 11 different lab interfaces. “We have to redesign every microsystem of care,” Greene said. Care takes place at the level of clinical microsystems with shared objectives and shared information systems. And in order for Hopkins to become a learning health system — for its use of the system and data to continually improve — it needs feedback streams.
Greene spoke about what it would take to improve measurement of the EHR’s impact on healthcare. The head of computer science in the school of engineering at Johns Hopkins told him, ”you’re asking all these questions about your system, but you are not instrumented to answer them.” It is like building a robot without any sensors, he said. “Meaningful use is building that infrastructure,” Greene said, “and we are now getting instrumented.”
Johns Hopkins is moving quickly into EHRs on mobile devices, pulled along by clinical users. Hopkins initially designed everything to be done on workstations, and it is struggling to adjust to users who want everything they need to do to be mobile. The pull is coming from clinicians who don’t want to be at a workstation, he said. They want to show their patients something vivid and interact with them.
Greene believes that once EHRs have been in place for a while, there will be an emphasis on improving their usability. “Our medical students have extraordinary expectations,” he said. “They want an app to use and if it’s no good, they want another app to try. That's the bar we have now. The systems still are very hard to use. The vendor community is exhausted with trying to keep up with some of the requirements that they have,” he said. “But we really need to focus on this, because it has become a distraction from good care, and better is certainly possible.”