"The combination of a library degree with a science degree is a powerful one," says Kate Oliver, "and I have always found it easy to find jobs that I like."

Oliver is a reference librarian at the NIH. She helps scientists and doctors and other researchers find the published information that they need about diseases, treatments, experiments, basic research, and so on. The tools of her trade include a variety of high-speed high-tech electronic resources, some sixty thousand books, and thousands of medical and scientific journals.

"Searching for biomedical information is," says Oliver, "a bit like the scientific process: you try and retry different approaches. The most important thing -- more important than knowing specific information about a subject -- is to understand how the literature is indexed in a database, that is, the 'structure' of the record. When you understand how data are put into a database, you then can visualize how best to find and pull out the kinds of information that you need."

Every time a researcher comes to Oliver's office looking for information about a topic, a "new puzzle has to be solved." Oliver says it is a lot like playing a game and definitely a lot of fun. "It is not always obvious at first what the best search strategy will be," she commented, and, for that reason, she prefers to do electronic literature searches with the requester sitting at her side. The requester can see the types of articles that are being retrieved and decide whether they are on target. Often, both on-target and off-target references will jog the requester's mind about related subjects and words; these help Oliver to fine-tune the search so that, in the end, just the right articles can be located.

The challenge for a reference librarian, notes Oliver, is to be imaginative. She says that it is crucial not to get discouraged when a search is not immediately yielding sought-after information. Although sometimes Oliver has been stumped, no search ends at a complete impasse: with 14 reference librarians at NIH all having different science backgrounds, one always is able to devise a strategy for ferreting out the literature that is relevant.

There are seven or eight key databases that cover the biomedical literature. Medline, a huge index of some seven million articles that have been published since 1966, is pretty much the "standard" for searching the medical literature.

Oliver says she is especially keen on using Dialog, a database "vendor" that allows her to search through numerous databases, such as Biosis (abstracts of scientific articles that were formerly collected in "Biological Abstracts") and Science Citation Index (abstracts of articles in diverse magazines and journals). Dialog covers many subjects in addition to biomedicine. With Dialog, Oliver simply strings key words together and then asks the computer to find meeting abstracts or titles of articles in journals and newspapers in which the words are juxtaposed.

Oliver spends a lot more time now than she did in the past training researchers to do their own searches. That's because, today, computers are everywhere, software is relatively cheap, and searching is easier than ever before.

Oliver says that she learned years ago when she was in library school that the typical scientist is an "independent library user -- someone who wants to search on his/her own." One of her professors even suggested that, for the scientist, a librarian is "an artificial intermediate." So, now, a significant chunk of her time is spent helping people grow comfortable with search software, search language, search strategies, and search technologies. She calls it "stress management on the receiving end to keep researchers from being overwhelmed or mystified."

Oliver and the other librarians teach search strategies to researchers both in the library's training rooms and in the researchers' offices and laboratories. At the NIH library, drop-in training sessions are offered all the time.

But, because biomedical researchers are mostly interested in doing research rather than finding papers, the need for professional searchers who know the ins and outs of each of the systems and how to find specific and often obscure references is still great. The librarians do, on average, about 600 computer searches each month. Oliver says that she and her coworkers believe that their "professional skills contribute to the scientific effort . . . to discover new treatments for disease, and we are happy to be part of that effort."

The reference librarians at NIH are also responsible for the book and journal collections of the library, which continually grow and change. Each librarian oversees certain biomedical "territories." Oliver tracks pediatrics, obstetrics and gynecology, statistics, epidemiology, and geriatrics. At one time, she also watched the mental health literature. She looks at, reviews, and orders new books in all of her areas and monitors new journals in those fields as well. Although the journal collection at NIH is pretty well established, a new journal will be ordered if sufficient interest in it is expressed by NIH researchers.

Pressure for the reference librarian comes mostly "from what you impose on yourself." Oliver loves what she does and enjoys the interactions with her colleagues, who, she says, are well-informed, interesting people with varied, well-developed interests. A reference librarian is exposed to a variety of research topics and is forced to explore and try different things, exactly what Oliver likes doing best. She thinks that the type of person who chooses this work is someone extremely bright with many interests and unwilling to settle on just one field to pursue.

With modernizations, many libraries have turned into technologic showpieces. But Oliver notes that that is simply not descriptive of the NIH library. It is a true research library and a critical part of the basic biomedical research that takes place at NIH.