Doctors and scientists at the National Institutes of Health in Bethesda are used to change. Administrators change places. Researchers come and go.

But the last 15 months have seen a remarkable series of changes in NIH leadership, starting with the arrival of Dr. James Wyngaarden, a Duke University research doctor, as director of the NIH.

Resignations and departures have let Wyngaarden name new directors of six of NIH's main research institutes, as well as a new director of the Clinical Center, the agency's research hospital. There will be an opening soon for a director of the NIH's National Library of Medicine.

The new leadership comes at a time of severe financial and political pressure on the practice of pure medical research.

Rep. Henry A. Waxman (D-Calif.), head of a House health subcommittee, wants an arthritis institute and for Congress to have far greater power to dictate which diseases the NIH will investigate. Budget constraints are forcing Wyngaarden to choose between grants for basic research and grants for bedside science, where the basic research findings are applied.

Several bills in Congress seek to regulate, and generally limit, the use of animals in medical research. Other proposals would limit drastically the use of human fetuses or fetal material in research.

NIH scientists are generally puzzled that this barrage of attempts to control them comes when medical research seems to be making its greatests contributions to control of human illness.

QUICK REFERENCE . . . The doctors were puzzled. A blood specialist said the partly paralyzed young man in Summit, N.J., had a blood disease, thalassemia. A nerve specialist said spinal cord compression was causing the paralysis. He advised surgery.

Could the patient have both conditions? And what to do? The hospital asked MEDLINE, the National Library of Medicine's nearly 20-year-old computerized scientific bibliography. With more than 6 million entries, it is the world's largest medical information retrieval system.

The answer came back. According to a scientific article, the patient could have both conditions, and radiation, not surgery, was the best treatment. The patient recovered.

The story is one of many to the credit of MEDLARS (for Medical Literature Analysis and Retrieval System). In 1971 MEDLARS added MEDLINE (MEDLARS on line) to give hospitals immediate access. The network now has 2,000 users: every medical school, research organizations, private companies, other government agencies (including the CIA) and 1,000 hospitals.

Hospitals and research centers in 14 other countries also use the system.

Robert Mehnert, NLM publications chief, says MEDLINE users are expected to conduct about 2.5 million searches this year.

EVEN QUICKER REFERENCE . . . If you go to the National Library of Medicine to look up something, you'll be directed to a computer terminal.

"The system works marvelously well," Mehnert reports. "People find what they want faster and easier . . . . We're no longer putting cards in our catalogue."