American hospitals fall into three groups in their adoption of electronic health records
, according to a study in Health Affairs embargoed for a Thursday morning release. A small set of high achievers represents 5.8% of hospitals. These can meet the second-stage standards set forth by the federal government. A larger second group is in the middle, representing 53.1% of hospitals. These have adopted either a basic or comprehensive record. Last, a significant minority has neither a basic nor comprehensive record, making up the remaining 41.1%.
The survey “really shows where the pain points are for hospitals,” University of Michigan professor and co-author Julia Adler-Milstein said in an interview. The study examined 2,764 hospitals' EHR use in 2013.
The vast majority of hospitals, the study found, can meet many second-stage objectives. Adler-Milstein said the main problems for hospitals are “around data exchange and data-sharing among care-delivery organizations, and among patients.” Particularly problematic for hospitals are the requirements that hospitals allow patients to view, download and transmit their data in machine-readable format.
“What can be controlled within the hospitals' four walls, they seem like they're doing well on. But when you need to coordinate with an external party, that's where they're behind,” she said.
Despite those problems, Ashish Jha, a professor at the Harvard School of Public Health and another co-author, said he thinks the survey shows things are “on track.”
“The reason I think (these results are) pretty good is that if you think about other countries, it's hard to point to any country that has made this much progress this fast,” he said. “It's also hard to point to any country with our size and complexity that has made this much progress this fast.”
The survey shows just-under 10% adoption of comprehensive or basic EHRs in 2008; five years later, the number has increased more than sixfold.
Jha, however, agrees that information exchange is the key problem for hospitals, and a key challenge to solve going forward. “We've always thought that a lot of the value in these things comes when information begins to flow outside of an organization and people start sharing data.”
Both also agree that the large number of hospitals without basic or comprehensive EHRs is a problem worth pondering. Adler-Milstein cautions that there's a bit of nuance lost in the definition. A hospital can have some basic information technology
, but not necessarily have enough technology to meet the basic EHR standard.
“It's not that 60% have an EHR and 40% have paper, it's just that they haven't adopted all the key features,” she said. Nevertheless, she was still concerned.
“I think we need to think seriously about how we get the remaining 40% up and running, and whether the incentives may not be enough to bring them along,” she said, noting that the hospitals that haven't yet met the basic standard tend to be rural, small and critical-access hospitals.
That opens the question of whether these hospitals “have the capability to do an EHR adoption process,” she said.
The other possibility, she speculated, was that some hospitals were gambling that penalties wouldn't materialize. “It would be really interesting to understand those who have not pursued meaningful use, what combination of those reasons are in play.”
Jha, looking at data for hospitals and physicians alike, speculated that perhaps the problem is with the meaningful-use
requirements, which specify a floor for quality. Software innovation is often disruptive, that is, it often offers a low-cost, low-quality product that everyone can use, as opposed to a higher-cost, higher-quality product whose cost is prohibitive for mass adoption.
“They might need” a disruptive system, he mused. “Will we tolerate that? Will we say, that's OK for some number of providers to have a cheap and not-very-good EHR?” Follow Darius Tahir on Twitter: @dariustahir