Computer Consult: EMR? Not exactly. Good enough? For now

August 9, 2002

These Kentucky FPs discover that a crude EMR is better than none at all.


Computer Consult

EMR? Not exactly. Good enough? For now

Jump to:Choose article section...EMR? Not exactly. Good enough? For now "Our doctors finally got comfortable with a mouse" The return on investment has been considerable

These Kentucky FPs discovered that a crude EMR is better than none at all.

By Robert Lowes

Intimidated by the thought of converting to electronic medical records? If so, you're in good company—plenty of it.

But don't let the fear of sudden, drastic change paralyze you. You can gradually automate your records while reaping immediate benefits. That's what Family Practice Associates of Lexington (KY) did three years ago, when it began to scan paper charts and convert them into static electronic images.

The Lexington FPs arguably have the most primitive form of EMR. They use computers primarily for viewing patient information. Unfortunately, scanned charts don't give the physicians the kind of searchable database that's at the core of a full-fledged EMR. Consequently, the doctors can't electronically query their records to produce, for example, a list of all diabetic patients whose most recent HbA1c level is 8 or higher.

But there's the rub: To have an ideal EMR with a searchable database, somebody must do the tedious work of filling in all those data fields using a keyboard, mouse clicks, stylus taps, or voice-recognition software. FP Jeff Foxx, the group practice's CEO, knew his computer-challenged colleagues weren't ready for that. "It would have been too much computer at once," he says.

As it is, his primitive EMR has increased office efficiency—goodbye chart hunting—and reduced costs enough that the system has almost paid for itself. What's more, initial successes emboldened the seven FPs and two physician assistants to graduate to electronic prescribing. They're also beginning to use electronic forms that they complete themselves. You might call it the incremental EMR.

Convincing the average doctor to automate clinical documentation sometimes requires that sort of go-slow approach, says Mark Anderson, a medical informatics consultant in suburban Houston. Right now, only 5 percent of physicians rely on an EMR system, according to the Medical Records Institute in Newton, MA, a group that promotes computerization in health care. Anderson says his own polling reveals that most of these doctors work in academic medical centers and megagroups like the Mayo Clinic. In other words, the EMR still hasn't made a substantial beachhead in Main Street medicine.

For doctors loath to leap into a sophisticated EMR program, says Anderson, scanning technology is a sensible first step.

"Our doctors finally got comfortable with a mouse"

Family Practice Associates chose a scanning and document management program called IMPACT.MD, from Louisville-based Advanced Imaging Concepts. The program allowed the practice to eliminate almost 20,000 paper charts.

That feat took about two years, however. At first, when an established patient was scheduled for an appointment, a staffer scanned his chart on a machine that could process about 25 pages per minute. Then the staffer would assign individual pages to customized online folders with titles such as "office visit notes," "medications," "in-house labs," and "immunizations."

When the patient arrived for the appointment, seven or eight key pages of the chart—the note from the previous office visit, for instance—were printed for the doctor. At day's end, these "minicharts" were put in a box to be shredded.

Before the practice started using IMPACT.MD, the doctors dictated the office notes for each visit. A hard copy went into the paper chart, and an electronic version was stored in the practice management system. That way, if the paper chart was misplaced, a doctor could at least find the last office visit in the computer.

Once IMPACT.MD was in place, however, doctors began documenting office visits on a one-page form that incorporated checklists and handwritten notes. This form was then scanned. "If we'd stuck with paper records, we wouldn't have given up dictation, because we wanted those electronic transcription files for backup," says practice administrator Susan Miller. "But when we switched to IMPACT.MD, we didn't need a backup system. So we dropped dictation, except for complicated visits."

It didn't take long for the FPs to computerize further. By the summer of 2000, they had weaned themselves from printed minicharts and were viewing patient data on desktop computers in each exam room. In March 2001, they began prescribing electronically with a software program from Allscripts Healthcare Solutions. "Our doctors were finally comfortable fiddling with a mouse and entering data," says Miller. Then, last January, the practice started to scan in lab and diagnostic imaging reports that the doctors could electronically annotate.

The return on investment has been considerable

Over two and half years, the Lexington FPs invested about $165,000 to launch and expand their EMR system, says Miller—80 percent of which went toward hardware, including two servers, desktop computers, laptops, a $5,300 high-speed double-sided scanner, and three slower, one-sided scanners that cost $900 each. The practice received a discount on the IMPACT.MD software because it served as a beta (testing) site.

The practice saves around $16,000 a year on the cost of paper charts for new patients, and $24,000 more on transcription. Within eight months, the practice also eliminated two file-clerk positions, which cut payroll costs by $30,000 a year. All told, the EMR system has added roughly $160,000 to the bottom line, says Miller. "Within a few months, we'll recover what we spent," she says.

Junking paper records freed up enough space for an extra exam room and a patient waiting area at the practice's main office. This building reconfiguration, in turn, helped justify hiring an additional FP in 2001, says Foxx.

Another benefit is that the EMR puts information at the physicians' fingertips. Foxx and his colleagues now access patient charts on their home computers while they're on call. "I don't have to rack my brain anymore trying to remember what I prescribed for someone," he says.

To be sure, an EMR consisting of scanned patient records has major drawbacks. With a full EMR, you could electronically retrieve a patient's blood pressure readings over two years and automatically graph them, for example. But the shallow database you'll get from scanning won't allow that. Moreover, it doesn't necessarily improve the quality of a physician's clinical documentation. If you scan in illegible handwritten notes, that's just what you'll see on the computer screen. In contrast, full-fledged EMR systems yield legible, organized, and thorough documentation.

The best EMR systems also incorporate clinical evidence and guidelines that assist doctors in treating patients at the point of care. They suggest how to code an encounter and interact with practice management software to produce a superbill. Such functions surpass the powers of scanned patient records.

Should the deficiencies of scanned records deter doctors from this option? Consultant Anderson thinks not. With most practice management programs, he says, doctors can drill down for data such as patient demographics, diagnostic codes, treatment codes, and medications. And while a full-fledged EMR cubbyholes much more data than that, the average doctor doesn't need one to practice medicine.

For Family Practice Associates, the scanned patient chart isn't the end of the EMR road. The FPs plan to introduce computerized forms for patient telephone messages, referral authorizations, and precertifications. Jeff Foxx dreams of a quick-and-easy form for progress notes. "Eventually we'll get to a true EMR," says Foxx. "I just can't change everything I do all at once."

The author, who is based in St. Louis, is a Senior Editor of Medical Economics.


Robert Lowes. Computer Consult: EMR? Not exactly. Good enough? For now. Medical Economics 2002;15:32.