PROPOSAL: COMPUTERIZED INTERNATIONAL PUBLICATION OF AIDS RESEARCH RESULTS

Despite the annual international conferences on AIDS, serious obstacles impede the rapid dissemination of research information. For example, the abstracts published at each Conference are due in January but not available to the public until five months later. Each Conference casts a shadow ahead during this five-month period; as the June meeting approaches, researchers become less and less willing to talk about their results. (Part of the problem here is the language on the abstract submission forms. Technically it does not require authors to avoid publishing their work during the five-month interval before the Conference, but the language does suggest such a requirement, and most researchers have interpreted it that way.) And aside from the Conference, the mechanics of journal publication usually delay information sharing for months, often for over a year.

Two years ago, organizers of the international AIDS conference in Stockholm had a "computer conference" system running at the meeting. Intended to provide international communication and publication for AIDS researchers, this system ran on a computer in Sweden but could be accessed internationally. However, we found that it was difficult for persons without academic affiliation to reach the Swedish system from a home computer in the United States. It may be less difficult today. (Note the International Global Teleconference now being organized in Sweden; it is listed above at the end of "Satellite Meetings and Conferences.")

How can we combine the need for selecting, editing, and peer review with the desire for an open system which avoids censorship by old-boy networks and narrow viewpoints? Here computerization can provide a unique advantage. Almost any abstract submitted could go onto the system; but also, expert referees, usually known to the system's users, would mark those abstracts which they considered interesting. These reviewers would usually be leading physicians and scientists, but they could also include representatives of grassroots organizations like ACT UP. Referees could make new recommendations, or cancel their previous ones, at any time. System users could choose the referees whose judgment they trusted, and then have the computer ignore all abstracts not recommended by any of them. (Users could also choose to ignore the referees entirely and look at any abstracts.)

How would the referees have time to look at all the abstracts, when thousands would eventually be submitted? They would not need to. Instead, the authors of new abstracts would be responsible for bringing them to the attention of those referees whose approval they wanted. Improvements could be negotiated, as is done with conventional peer-reviewed journals today. Authors could publish a rough draft immediately, then revise it later to obtain referee endorsements. This "instant peer review" provides the benefit of traditional refereed journals, while completely eliminating the publication delay which is their major disadvantage. It provides selectivity without censorship, by allowing readers to choose exactly whose selectivity they want.

Many medical and scientific journals which otherwise demand an exclusive from their authors will now allow publication of abstracts at scientific conferences, without ruling the work ineligible for formal publication. (Otherwise scientists would be reluctant to submit work to the International Conference on AIDS, for example.) This precedent should allow researchers to publish an abstract on a "computer conference," too.

For this system to work, it needs software which is easy to use. We frequently use online research databases, and we have found that two commands -- FIND and PRINT -- would be enough for all but advanced users. For example, a typical search might be

FIND AZT AND DDC

which would immediately print the titles of all abstracts which contained both those words, starting with the most recent. Beside each title would be a sequence number. Then

PRINT (followed by sequence numbers)

would print the full abstracts corresponding to the sequence numbers selected -- or PRINT by itself, with no sequence numbers, would print them all. Anyone could learn this software in minutes. Yet these two commands would be powerful enough to provide convenient access to databases containing tens of thousands of abstracts.

Such a system would provide instant publication and communication, at all times and in all countries. The instant peer review with customized selection of referees by the reader, which we described above, would allow the same system to serve researchers, practicing physicians, AIDS activists, educators, legal experts, and others, giving each group its own view of the data, while allowing anyone to experiment with other views if they want. It would be like a permanent International Conference which anyone could attend at any time, without leaving their home or office.

Many components of such a system already exist; there is no need to re-invent them. The purpose of this article is to give one picture of how a useful computer publication system could work. If you can help with computer publication and communication, write to us at AIDS TREATMENT NEWS, attn: computer conference.