In 2004 Baltimore's Sheppard Pratt Health System rolled out a
state-of-the-art electronic medical record (EMR) system.
The system has yielded many of the benefits that EMRs have promised to
deliver—rapid access to patient records, an end to illegible charts, and
enhanced continuity of care by allowing doctors and nurses in disparate
departments instant access to the same records.
But suppose, for instance, that a doctor wants to pull out a discrete piece
of data about a patient's recent hospitalization—say, blood
pressure—from another institution's EMRs across town.
"It is easy to find any unique clinical data element from within your
own familiar clinician-user interface," said psychiatrist John Boronow,
M.D., of Sheppard Pratt. "But to find that same element in another
system with a totally different interface and data dictionary and database
structure is still impossible. Even within our own system, to actually
identify that same unique clinical data and export it in such a way that
another EMR system could use the data, requires extremely time-consuming
He is medical director for adult services and physician liaison for
information services at Sheppard Pratt. He is also a member of the APA
Corresponding Committee on Electronic Health Records.
(EMRs are variously and more broadly called electronic health records
[EHRs] within the health information technology [HIT] lexicon.)
Boronow cited as an example that the hospital has been unable to coordinate
its EMR efforts with neighboring Johns Hopkins Medical Center, though both
institutions use the same vendor and have a similar vision of the psychiatric
data that are clinically important.
"The actual details of each institution's installation are just too
different and incompatible," he told Psychiatric News."
The only alternative at the present would be to share data in a third
database, called a data repository, which is really a research tool and not
something that would facilitate real-time exchange of clinical
The problems encountered at Sheppard Pratt are examples of why even true
believers in the promise of EMRs like Boronow have come to realize that the
grand goal of "interoperability"—the ability to share data
across the country and across systems and institutions—is proving more
elusive than previously imagined.
He noted that paper-based discharge summaries were invented a hundred years
ago to as a way of summarizing an enormous amount of data, using human
clinical judgment to select what was important with a goal toward synthesizing
and communicating a coherent message. "But with an electronic record, if
you are talking about drilling down to share all the data seamlessly between
EMRs in real time without exercising that human judgment [about a patient's
treatment]—that is a nontrivial exercise," Boronow said.
"Even if we assume that everyone has the technology, no one can agree
on the definition of the data elements that should be shared," Boronow
said. "And these data sets are huge. For a six-day medical-surgical
hospitalization of any sort, there are countless data elements. How to make it
all interoperable is proving much harder than anyone was able to
These bumps and potholes in the road to rapid electronic exchange and
efficient use of health records are among the reasons why the original goal of
achieving a national health information network by 2014—as envisioned by
President George W. Bush in 2004—is not likely to be met.
Interviews with leaders in HIT and the few formal studies of EMR adoption
that exist suggest that there has been only slow progress in the last two
years, and many of the problems that have doctors reluctant to adopt an EMR
system—cost and lack of standards for knowing what the necessary
components are before investing—have yet to be resolved.
Moreover, privacy concerns (see Privacy Must Take Precedence) continue to
worry physicians, especially psychiatrists.
"It's generally agreed that it's going to take longer than
anticipated to achieve widespread use of EHRs," said Robert Plovnick,
M.D., director of APA's Department of Quality Improvement and Psychiatric
Plovnick also said that APA's Council on Psychiatry and law has asked the
Corresponding Committee on Electronic Health Records and the Corresponding
Committee on Confidentiality to review and revise the 1997 position statement"
Confidentiality of Computerized Medical Records."
That statement reads as follows:
"Computerization of medical records exacerbates the threats to
patients' confidentiality. This is a particularly important issue in
psychiatric treatment, where sensitive information is frequently involved. If,
after discussion between the psychiatrist and the patient, the security of the
computerized data is unacceptable to the patient, psychiatrists should have
the option of entering the information into a noncomputerized
A clear picture of the rate of adoption of EMRs by physicians is hard to
discern. After all, what exactly does it mean to have an electronic medical
"A big challenge for adoption surveys is that even the term
'electronic health record' is so loose," Plovnick said. "It
depends on what you are defining as adoption. If you define it as anyone who
has a computer, you can come up with a high number."
But between owning a computer and having a fully operational EMR system
lies a lot of variation. "One of the goals is to at least come up with a
standard for what constitutes an operational EMR system," Plovnick said."
Until someone does that, it's hard to get a hold on the
An October 2007 study in the Journal of Medical Quality surveyed
whether physicians in Florida who had adopted EMRs had systems that included
23 desirable functions. The survey was sent to all Florida primary care
physicians and a random sample of specialists, including psychiatrists.
Of 4,302 who returned surveys, 995 indicated they were using some kind of
EMR. Most adopters were missing key features. While 99.3 percent of those with
an EMR system indicated they could store and retrieve clinical notes, only 60
percent said they had electronic prescribing, and just 34 percent said that
their systems provided preventive-service reminders.
The list of functions (see EMR Function Checklist) was derived from a 2003
report by the Institute of Medicine titled "key Capabilities of an
Electronic Health Record System."
Data from the Journal of Medical Quality study were not broken out
by specialty, but Plovnick believes that generally psychiatrists have been
slower adopters than other specialists.
As reported in the September 15, 2005, Psychiatric News,
physicians say that costs are prohibitive, especially for small-group and solo
practitioners, and because standards are still evolving for what kinds of
functions are necessary, doctors are reluctant to invest in an expensive
system they may have to update or replace in a few years.
"There is no real push from below from users," Boronow said."
My physician colleagues don't want this stuff because all they see in
the short term is expense—both direct in terms of purchasing and
indirect in terms of interference with their immediate productivity during the
learning curve period, which in the era of managed care is a big
So what needs to be done to speed up adoption of EMRs?
Boronow looks for answers in the success of the Department of Veterans
Affairs (VA) and the British national Health Service—both"
top-down" systems in which the government has had an active role
in promoting EMRs and determining standards.
While that cannot be duplicated in a privatized health system, he believes
the government has to take a more active role in defining standards. And he
recalls how the government defined mandatory billing requirements for Medicare
(with the ubiquitous HCFA 1500 form) as a precedent for introducing
Still, despite the slower than anticipated adoption of EMRs, there have
been positive steps taken by the government to provide incentives to
physicians (see HHS Seed Money Helps Build Health-Information). And a number
of legislative bills regarding EMR systems continue to percolate in Congress
(see box on facing
Though widespread adoption may take longer than anticipated, leaders in
medical information technology believe the benefits of EMRs so outweigh the
disadvantages that there continues to be an aura of inevitability about the
movement; and psychiatrists familiar with EMRs continue to tout its advantages
to the specialty.
Perhaps the most successful and comprehensive implementation of an EMR
system has been at the VA.
"Fifteen years ago we were the laughingstock of the nation's health
care system," psychiatrist Peter Fore, M.D., told Psychiatric
News. Fore is chief of the Outpatient Psychiatry Section at the VA in
Chicago. "Now we have six years in a row of beating the private sector
in patient satisfaction, largely through electronic medical
The VA's system has virtually all of the patient-safety features delineated
by the Institute of Medicine. "Once you have an electronic system, the
chart is no longer a dead entity," Fore said. "Rather, it becomes
involved in clinical decision making. We have clinical reminders that alert
the physician to certain things. When you have that, you can make a big step
toward implementing practice guidelines.
"We have an abnormal involuntary movement scale, and we have
reminders to do that test as well as to check for smoking and hepatitis
C," he said. "The VA has become one of the best in terms of
prevention and primary care because of all the prevention reminders built into
our electronic system. We have pop-ups to remind physicians to screen for
depression, PTSD, and problem drinking. And our physicians are held
accountable—that's not always the most popular feature, but it has made
a tremendous difference in quality of care."
The study "Incomplete Adoption of EHR: Late Uptake of Patient
Safety and Cost Control Functions" is posted at<http://ajm.sagepub.com/cgi/content/abstract/22/5/319>;
the report "Key Capabilities of an Electronic Health Record
System" is available for purchase; a description is posted at<www.iom.edu/CMS/3809/4629/14391.aspx>.▪