Commentary (Glass): NCI's Cancer Information Systems-Bringing Medical Knowledge to Clinicians

In 1995, everyone is on the "Information Superhighway." The National Cancer Institute, not to be upstaged, has developed its Cancer Information Systems into a powerful and easy-to-use resource for all to use. The article by Hubbard, Martin, and Thurn describes some of these systems, stressing the techniques employed to develop the databases, the expert panels used, and the processes used in fashioning recommendations. PDQ is designed for use both by physicians and the general public, a unique configuration as far as I am aware. Two other programs, CANCERLIT and the Information Associates, are directed at oncologists. Hubbard et al only mention these briefly, and nothing more will be said about them in this commentary.

PDQ is the popular and heavily used resource for oncologists. In an age of information overload, when every physician complains of his or her inability to keep up with the medical literature, PDQ is a ready expert consultant and counselor. It is encyclopedic, full of consensus recommendations from recognized experts, easy to use, and free. Judging by the figures quoted by Hubbard et al and the frequent mention of PDQ in conversations among oncologists, this is one resource that is often used. Almost every physician is computer literate, and has a PC in his or her office and usually one at home as well. The mechanics of connecting to the database seem to be the least of problems, but several other questions do arise.

For What Purposes Do Physicians Use PDQ?

Do physicians use PDQ to confirm their management of "routine" cases, or do they reserve it for particularly rare cancers for which management decisions may need to be made in the absence of any personal experience? One suspects that the latter behavior is more common. Faced with an unusual cancer that he or she has not seen before or in years, it is comforting to know that useful management recommendations are readily available at any time of the day or night.

On the other hand, any surveyor of practice patterns in American oncology must be impressed with the degree to which the most modern techniques are today being used in every medical setting, no matter how physically removed from large cities and universities. PDQ, in this light, may be one of the many medical educators in the land, serving to raise the overall quality of care for both common and rare cancers.

Other Questions

Do physicians use the material merely to confirm their strongly held prior opinions, or do they seek new information for patient care? This is a more difficult question to answer. Physicians are clearly under great scrutiny to practice modern, error-free medicine. Nowhere is this more evident than in oncology, as newspapers, magazines, and television feature stories about "breakthroughs" and "cures" on an almost daily basis. PDQ may serve as a helpful confirmation of previously held opinions about management of a difficult case, and may further serve as an "authority" that can be quoted for the benefit of patients, family members, and colleagues. Finally, as PDQ is a public database with a significant "readership" among the lay public, the practicing physician may do well to know what is in this database, as his or her patients may quote it and ask about their own care in comparison to what they have read.

Is PDQ Independent?

Along similar lines, what is to be made of the expert, consensus opinions expressed in PDQ? In this time of questioning the central government, has PDQ retained its independence of the opinions of the National Cancer Institute bureaucracy? I certainly find the material free of such political taint, but does everyone feel likewise? Are the views expressed about the management of a particular cancer too slanted toward surgery or radiotherapy or combined modality treatment? Are their views concerning the quality of evidence necessary to justify a recommendation too slanted toward prospectively randomized trials, especially now that such trials are under such hot attack? What is the influence of the National Cancer Institute staff concerning their criteria for acceptance of evidence concerning screening and prevention strategies? These and other questions remain unanswered.

A major thrust of PDQ since its inception has been the furthering of the clinical trial effort. It is well-known that only a tiny fraction of patients enter clinical trials. The numbers vary from cancer to cancer, but hover at about 2% of newly diagnosed cases. This figure has not changed in years, despite the major Community Clinical Oncology Program (CCOP) effort, increased funding for clinical research, intergroup studies, etc.

Many entries in PDQ clearly state that there is no single recommended treatment for a particular situation, and recommend that patients be entered on this or that clinical trial. Further, there is a large section in the database that offers clinicians full information about entering their patients on these trials, including whom to call to enroll participants. Does this work? Has anyone counted how many patients have indeed been enrolled in such trials as a result of a PDQ search and recommendation? Anecdotal experience tells me that most of the searches of PDQ that wind up at a protocol schema result in the clinician using one of the arms of the protocol that seems more attractive. The rationale for a clinical trial may, in this setting, be ignored for the sake of a convenient recipe for the waiting patient. If this is the more common scenario, and I would love to be wrong, much of the promise of PDQ is still unrealized, and considerably more work needs to be done.