“That was the infrastructure for what became ARAMIS,” he explains. “And it was the concept with which I was recruited back to Stanford by Hal Holman. [He had originally graduated with a B.A. from Stanford in 1960.] At that time, I think you could say I was the only IT expert in rheumatology.”
In the late 1960s, Halsted R. Holman, MD, was chair of the department of medicine at Stanford University’s School of Medicine. Now the Guggenhime professor of medicine, emeritus, Dr. Holman recalls that “the striking thing that [Dr. Fries] wanted to do—and that I thoroughly agreed with—was to begin to do good measurements of healthcare outcomes. Now, that was not just setting up an instrument of some form to which you could assign numbers and do number crunching. It was much more substantive, in that he wanted to put together whatever data there were about patients’ experiences from illness and from our efforts at treatment.
“Our treatments for rheumatic diseases were rather meager at that time,” continues Dr. Holman. “What he tried to do was break down the relationship between different physiological abnormalities, what we could measure in the laboratory, and clinical states that the patient had. Jim wanted to put together this information in a composite of what—on average—happened to patients, and to watch this evolve out over time in both treated and untreated patients.
“So, that was the beginning and it was something that wasn’t being done widely at all,” says Dr. Holman. “We thought it was important to do, and encouraged him to do it—and you know the results.”
Following his fellowship at Stanford, Dr. Fries became a junior faculty member in Dr. Holman’s department. Work on the concept of the computerized rheumatic disease unit began slowly. Dr. Fries had the “usual junior faculty problems” of raising money to build the unit. However, he enjoyed fruitful collaborations with Stanford’s programming community. At that point, the computer was mainly seen as a computational tool, and storage of data slow and prohibitively expensive.
“In retrospect,” he says, “it was awfully primitive. The computer was essentially a big hand calculator, with a minimum of data storage possibility. We had to develop what’s been termed the first relational database because there weren’t any anywhere at that time.”
These efforts yielded a number of papers describing the IT aspects of the team’s work, including the chronic disease databank model, whose cornerstone was establishing a rolling cohort of patients, with continuous open entry.4,5
ARAMIS Launched
The Arthritis Act of 1974 facilitated the formal launch of ARAMIS. Dr. Fries credits Evelyn Hess, MD, an adult rheumatologist at the University of Cincinnati, for her efforts to include in the act provisions and funding to establish a nationalized patient database. Dr. Fries’ Stanford group applied to the National Institutes of Health (NIH) to become the national arthritis data resource. Larry Shulman, MD, PhD, director of the National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS), also helped lead that effort. “You could say that the program actually formally began with an act of Congress,” says Dr. Fries.
