Health IT In 2018: Crystal Ball Predictions

10 Mobile Health Apps From Uncle Sam

In a new paper in the Journal of the American Medical Informatics Association (JAMIA), three health IT experts summarize the progress made in the field to date, list the barriers that remain, and make predictions about what will happen in health IT over the next five years.

Donald W. Simborg, Don Eugene Detmer and Eta S. Berner, who forecasted the current explosion of health IT in a 2005 article now predict that by 2018 we can expect:

-- Near-universal adoption of EHRs.

-- Greater emphasis and progress on standards and interoperability.

-- New breakthroughs in user interfaces.

-- An emphasis in clinical decision support systems on genetics and personalized medicine.

-- A resurgence of computer-assisted diagnosis.

-- A better understanding of the strengths and weaknesses of big data.

-- A blurring of the distinction between telemedicine and EHRs.

[ Health IT gets a boost from a popular former president. Read Bill Clinton Stumps For Health IT At HIMSS. ]

In an interview with InformationWeek Healthcare, lead author Don Simborg, a health IT veteran who helped found HL7, said that in 2005, he and his colleagues expected that EHRs would spread rapidly, but only if there was a change in the financial environment. What they didn't expect was that that change would come from the federal government in the form of $19 billion in EHR incentives.

The JAMIA paper cites the serious deficiencies in EHR usability that physicians have encountered. Simborg attributed those problems partly to the fee-for-service payment system, which rewards providers for the volume of documentation that EHRs make possible.

"That's driven a lot of things that make usability ugly: for example, the requirement to do a complete review of systems, along with family history, past history and so forth. As a result, the record is really unreadable and, in fact, unreliable. Physicians don't trust these records that are computer generated by these large templates. So what will change that is a change in the payment system and a greater emphasis on some kind of pay for performance -- the kind of shift we're starting to see, where you don't get paid by documenting eight pages of patient history, you get paid by improving the healthcare of the patient. That will improve documentation as well."

Another usability problem is related to the difficulty of documenting patient encounters with point-and-click templates, he noted. "Doctors are storytellers. They like free text, they don't like dropdown menus, because they don't express enough. So I think we're going to see breakthroughs in natural language processing, which will make a big difference." The area where this is most needed, he added, is in the history of the present illness, which doesn't lend itself to templated documentation. Interoperability between disparate EHRs is still largely lacking. Many observers believe that the problem is lack of standards or poor implementation of those standards, but Simborg disagrees.

"The real problem with interoperability is not standards," he said. "We've got more standards than we can deal with. The problem is the buyers of healthcare -- the hospitals, the big healthcare organizations, the integrated delivery networks. Once they see it's in their economic interest to demand that the vendors be interoperable, that will change things. But so far, they haven't done that."

Simborg also believes that clinical decision support needs to be rethought to get widespread physician adoption. "Physicians resist decision support. They hate getting alerts and reminders; in fact, they get so many of them they ignore them. So this is a usability problem. How do you deliver this knowledge to physicians at the point of care?"

The right way to do it, he said, is to provide the decision support in ways that reduce the amount of work for physicians. For example, an "intelligent system" could use clinical protocols to enter default orders that the doctor could approve, which would save him the time required to input those orders himself.

Regarding the prediction in his article that CDSS will start to emphasize genetic data, Simborg agreed with other experts who recently observed that EHRs aren't designed to use genomic information and don't have sufficient storage capacity, in any case. He believes that EHRs will have to change to accommodate this data and that they will in the next five years.

The distinction between telemedicine and EHRs will blur, he said, because "more and more medicine will happen through the Internet. Lots of healthcare will happen without there being a physical encounter, and the EHR will have to cope with patient inputs a lot more. So we won't call it telemedicine anymore."

Regulatory requirements dominate, our research shows. The challenge is to innovate with technology, not just dot the i's and cross the t's. Also in the new, all-digital The Right Health IT Priorities? issue of InformationWeek Healthcare: Real change takes much more than technology. (Free registration required.)