There was no Epic or Allscripts. It was long before the days of meaningful use. In fact, there was no full electronic medical record system in the nation to speak of. Clement McDonald, MD, recalls those days with ease.
It all took place beginning in 1972 at the Regenstrief Institute when McDonald and a few colleagues took it upon themselves to develop what was widely considered to be the first EMR system in the country.
What's more, no one had asked him to build it. "We were sort of mavericks all along," he says, chuckling. McDonald had been planning on building a system since the mid 1960s, a plan stemming from what he observed while interning at Boston City Hospital, where hospital admissions were an ordeal in their own right.
"It took five or six hours to admit one patient before we had all the data," he told Healthcare IT News.
"I just knew that it was stupid that we were spending so much time finding information to make good decisions about patients," he says. Oftentimes, they couldn't even locate the right information. "It seemed like the effort should be to do the best job for the patient not to spend the time scrounging."
He knew computers could transform the way medicine was practiced. He also realized he was capable of pioneering the project, as between his internship and residency, McDonald earned his master's degree in biomedical engineering and subsequently spent two years at the National Institutes of Health where he built an automated laboratory system. "That convinced me it was really doable," he recalls. So he went to work.
McDonald started small. His and his team's first goal was to do the data capture display and reminders – what's now decision support – for the Wishard Diabetes Clinic, which had 35 patients at the time.
After each patient visit, McDonald explains, they'd reprint the whole medical record. It would contain the flow sheet detailing the time course of each observation – blood pressures, drugs, test measurements – together with another piece of paper containing reminders like, "it's time for your mammogram."
Then, there was a third paper form, the "encounter form," which included pre-printed data the clinic already knew about the patient. The patient would fill in items for that particular visit. Then these forms would be optically scanned into the computer.
And the computer?
After a brief stint with a PDP-11/44, put out by the Digital Equipment Corporation, Regenstrief soon purchased an upgrade, the PDP-11/45.
As McDonald recalls, the computer had a 10-megabyte disc, 64 kilobytes of ram and cost around $170,000.
Within a year and a half, they had things up and running for the Wishard Diabetes Clinic – then they completed the entire medicine clinic by 1974. The project expanded to the hospitals in the '90s, and now carries data on more than 1.4 million patients.
McDonald went further. Together with the medical record project, he and his team also built a lab system, a pharmacy system and a scheduling system, using their own interfacing and codes.
Interfacing was no easy task, though, especially when hospitals started a buying frenzy with new EKG systems and cardio echo systems. "We would do interfaces with them in really awkward and not successful ways, like you take the printer output and you run a wire to your computer, and then you capture the report and then you parse out the information," he recalls. "If the printer line wasn't plugged in, it just sent (data) out into the ether. It didn't warn you like it does today, like your printer is offline."
Another issue with interfacing arose in the early nineties when Regenstrief was connecting to other hospital systems. Every hospital had their own code system, and the names they used weren't necessarily understandable, which led McDonald to embark on yet another project – establishing the Logical Observation Identifiers Names and Codes, or LOINC, for standardizing purposes. He was also the co-founder of the HL7 message standards.
His work, he says, is still not over. McDonald continues to dedicate his time to health information technology, now serving as the director of the Lister Hill National Center for Biomedical Communications at the National Institutes of Health, where he leads research efforts on clinical informatics and EMRs.
When asked to describe the environment of when he first started his career developing the nation's first EMR, he recounts fondly, "There was a time when the prospective interns coming to interview would ask if we had computer – in the singular, like it's a substance, like sugar. 'Do you have computer?'" Those, he adds, were the days.