Two HHMI physician postdoctoral fellows have developed Medsite Navigator to assist patients, physicians and researchers in locating high-quality medical and scientific information.

Want to be a Web Head without wasting time? Trying to find useful biomedical information on the dross-ridden electronic highway? Robert Sikorski and Richard Peters can help.

The two HHMI physician postdoctoral fellows have developed Medsite Navigator to help patients, physicians and researchers locate high-quality medical and scientific information. Since April, they have also been writing a monthly column for Nature Biotechnology, in which they highlight useful Web sites for scientists.

Their Medsite Navigator, which can be found at , offers links to Web sites dealing with topics as varied as tissue culture, antibiotics, gene sequences and disease incidence. There's even a link to a gestational age calculator for pregnant women, something handy "for the clinician who has forgotten the formula," said Peters, a Belgian who is pursuing his fellowship with structural biologist Gerhard Wagner at Harvard Medical School.

The two physicians say there's more to come. Speaking at a conference of HHMI predoctoral and postdoctoral fellows at the Institute in June, Sikorski compared the mushrooming of scientific material available on the Internet with advances in DNA sequencing. "What other fields have become so data intensive so quickly?" he asked.

Sikorski is studying cancer genetics in the laboratory of Harold Varmus, director of the National Institutes of Health. Like Peters, he holds both an M.D. and a Ph.D. degree, and worked as a clinician before receiving the three-year fellowship from HHMI.

Last year the pair attended a conference on information systems for health care professionals, where they were struck by the emphasis on "back office issues: billing, claims and banking stuff," Peters said. "We were amazed that no one seemed interested in using this wonderful technology to manage knowledge for the health care provider or the patient. We decided to work on a system of software and human editors that could deliver high-quality biomedical information when you need it, fast."

"In the summer of 1995," Sikorski said, "it became obvious that the Internet was the gold mine of information we had in mind."

Since then, both physicians have worked steadily to develop new software and database tools to help scientists cut through the growing Internet data jungle. To find ten splendid sites for their navigator, they say, they must check 500 or more. "What's worrying are the problem sites that promote bad medicine," Peters said. He cited one Web site that touts an antibiotic as a miracle drug for AIDS because it has been used to treat a different illness.

Web sites located at academic institutions and government facilities such as the NIH and the Food and Drug Administration tend to have reliable information, Peters said. "But remember," he adds, "any individual affiliated with a university can have an address with an '.edu' suffix. This does not mean the information is endorsed by the institution."

Peters says he used a Web site at the Centers for Disease Control and Prevention to determine which malaria prophylaxis was appropriate for a patient traveling in the Caribbean. Sikorski described using the Web on a gene mapping problem. He visited a computer in England for the gene sequence, then used a computer in France to determine which primer was appropriate for the experiment. Finally, he turned to a computer in Alabama to order the primer. "It's hard to imagine getting information any quicker," he said.

Both researchers are excited about the advent of Java and other new programming languages that enable Internet users to run software directly instead of just downloading information or programs. Peters said he and others who study the structures of proteins might be able to twist and turn images on their screens while connected to a site halfway around the world. Scientific journals such as Nature Structural Biology already provide readers with protein structures on-line, Sikorski noted, and in a format that can be manipulated while on the Web with the help of a small program called a browser plug-in.

The technology is moving so rapidly that it is difficult to predict how the Web and other advances will affect biomedical researchers, the two fellows said at the conference and during a meeting the following day with HHMI staff. Commercial companies already are offering to store all of a lab's data in cyberspace, and researchers are using the Web to identify patients around the world who might participate in clinical studies of rare diseases. Similar examples abound.

"Maybe someday physicians will be able to tap into the Web to find out everything from a patient's medical records to their living will, and serve them more effectively," Peters said. "For better or worse," Sikorski added, "the Internet will change how we do science and practice medicine. We are now only limited by what we can imagine."

The two physicians met as interns at Massachusetts General Hospital, where they developed an effective strategy for working together. "MGH is very progressive and let us try a tag-team approach," Peters recalled. "One of us, 'the mule,' would do the grunt work: starting IVs, drawing blood, sending specimens and so forth. The other, 'the fox,' would focus on patient care. Patients loved it because they received the attention of two physicians. We were more efficient and it was fun. To this day, we have kept the 'fox-mule' relationship, although, of course, we alternate roles."