" 'Change is scientific; 'progress' is ethical; change is indubitable, whereas progress is a matter of controversy."
Happily ensconced daily in academia, I see my patients, try to be on time, check my e-mails, and read my medical journals. From these activities, I have become increasingly aware that, over the past 10 years, the way in which I am being educated is changing. These changes are remarkably subtle, almost subliminal, and are becoming a regular event.
Recently, I was lamenting over the sad collection of incomplete, out-of-date textbooks and journals in the psychiatric emergency room of a major academic institution. Several staff members seemed perplexed by my grief. "After all," they said, "we have the Internet. Everything's on there." Shortly afterward, I asked a resident to look up the DSM-IV criteria for schizoaffective disorder. To my surprise, he did not reach for the Manual (which was on the table in front of him); instead he took the time to walk across the clinic to a computer, where he searched for the information on the Internet. Later, I, too, flicked on the computer to check my e-mail and, from somewhere on the Web, I received not one, but two unsolicited "informative" pieces of literature. One was a biased, incomplete, unreviewed "study" promoting a selective serotonin reuptake inhibitor (I had to wonder who granted its funding); the other was a page of "e-junk" from a health insurance company; it was titled "Short Office Visits Prove Beneficial in Treatment of Anxiety Disorders"—author, anonymous. (I had to ask the question: "beneficial to whom?")
These "informative" reports—along with other, innumerable pages of unrequested, unclear—and sometimes frankly untrue—pseudoscientific transmissions flash by psychiatrists' retinas to a greater extent every day. What makes this particularly distressing is that I sometimes find myself inadvertently recalling their content. (I see a patient in the afternoon with anxiety and wonder about the benefits of a shorter session—and—oh—what was the name of that SSRI I just read about?) To make matters worse, I, and many other psychiatrists, increasingly rely on the Worldwide Web as a source of accurate information. When a question arises, I just click on a search engine and wait for an answer; thus leaving my cherished medical journals in their respective piles, aging beside my computer, collecting dust. The truth is, at present, searching the Internet for medical information is often like searching for facts in a local public library where every book claims to be nonfiction.
Survey of Psychiatrist and Psychiatric Residents' Sources for Information
To better characterize if and how the manner in which psychiatrists gather information is actually changing, we conducted a self-report survey of 30 academic psychiatrists' and 25 psychiatric residents' information sources from the present and from 5 years ago (mean age of psychiatrists=43.8±6.7 years; mean age of residents=30.9±3.7 years). Subjects were obtained from 200 mailings to four Northern California academic medical centers. Survey response rate was 27.5% (see t1 for subject compositions). The survey required participants to give the best estimation (percentage) of present and 5-years-ago method of information-gathering. The survey asked "When gathering medical/psychiatric information, I use the following [what percent of the time]?" Medical/psychiatric information was defined in the survey as research articles, educational material, drug information, outcome data, case presentations, as well as investigations of theory and practice. The three categories subjects were asked to estimate percentage utilized were 1) print material (textbooks, books, periodicals, journals); 2) live or videotaped conferences, lectures, classes; and 3) the Internet (e-journals, classes, sites, links such as MEDLINE). Subjects were instructed that the total of the three categories had to equal 100%. The survey excluded information gathered by personal experiences, personal communications, personal research, and other less-formalized means of education.
It is hard to appreciate how paperbound our mind-set is until traditional methods of printed communication are challenged. We regard "scientific reports" as something inherently trustworthy, a paper rolling off the presses at regular intervals in immutable ink. However, given the above data, this sacred mountain of print journals appears to be tumbling. Psychiatry, too, has been swept up in this Internet age, the "e-avalanche" roaring down at practitioners to such a volume and degree that only those who are blind and deaf can refuse to step out of the way.
A major boulder accelerating—if not detonating—the avalanche-discussion regarding medicine on the Worldwide Web was the recent propositions for two online entities, dubbed e-Biomed and PubMed Central. Before January 1, 2000, when Dr. Harold Varmus left his position as Director of the National Institutes of Health (NIH) to become the Chief Executive Officer of the Memorial Sloan-Kettering Cancer Center, he proposed creation of an entirely electronic repository of all biomedical publications (1,2). This massive "tome," originally named e-Biomed, would be an NIH-operated site, an expansion of MEDLINE (the bibliographic index collated by the U.S. National Library of Medicine) made available for the electronic distribution of all life-science research reports. Once a manuscript has been accepted by a peer-reviewed journal, e-Biomed would include it in its online library; e-Biomed could broadcast the manuscript over the Internet before it becomes available in print, although there would be some notation as to which journal had accepted and is processing the paper. Several months later, e-Biomed was revised, and PubMed Central was launched. PubMed Central is the initial site of what will be a "barrier-free" and "international" repository for articles. It is similar to e-Biomed, except that its scope was expanded to include life sciences in general, and plant and agricultural research, as well as biology and medicine. PubMed Central endorsed providing a system of peer review that would be "the responsibility of groups that have no direct relationship to the NIH (2). Articles for PubMed Central could be posted before, during, or after printed publication in other journals, as copyrights would remain with the submitting groups.
Included in both plans is a controversial second category of postings for "preprints," those scientific or pseudoscientific papers that have been neither screened nor peer-reviewed. This mechanism could be utilized by authors who were not successful in placing papers. The non-reviewed literature would, according to Dr. Varmus, be used to stimulate discussion and act as a "democratizing force that makes distance and wealth nearly irrelevant.… Criteria for approval of reports must be sufficiently firm to guard against gross abuse of the [information] repositories, but sufficiently flexible to permit rapid posting of virtually any legitimate work (1)." The idea that authors might be charged for entries into the archive was also considered. This raised some eyebrows and questions regarding how such a fee could affect quality and an author's motivation for publication. Certainly many would pay (and already do in time and labor!) the financial tariff required to promote a scientific finding, or a theory, or to share with the community an instance in practice that highlights or elucidates a subtler aspect of psychiatry—just as many might pay to promote their own academic standing, earnings, or vanity.
e-Biomed and PubMed Central are just two more compelling signs that traditional scholarly journals have had themselves launched into the Internet age. Like most national and international endeavors, there is great pressure on psychiatric journals to enter the Worldwide Web. Readers' expectations for accessibility, competitive pressures, the high cost of generating printed material, the growing demand for speed in dissemination of facts—these factors compel annals once shunning Internet platforms to concede. Most major psychiatric journals now produce and maintain websites containing full or abridged versions of abstracts, articles, issues, and back editions, and provide a span of services, such as links, frequently asked questions (FAQs) with replies, e-mail, and "bookstores"—and the list goes on. Although there are obvious advantages in this information being readily available, some concern stems from the blurring distinction between journals and databases (3). The Internet's immense information bank, embodying billions of bits of advice, news, tips, and "facts" is largely ungoverned. Correlation of such information in order to endorse or refute a hypothesis has classically been the task of printed scientific journals, the editors of which have a more or less standardized means of reviewing validity, merit, and quality. However, substantiation of this kind does not customarily occur on the Internet. This problem, accompanied by the vast bulk of data available, makes searching the Internet a process of myopic sifting through an information haystack for the rare needle of germane insight.
Although the fact is regretted by some traditionalists, the above survey and evidence indicate that psychiatric reporters and readers are in an age of adaptation. As with every evolution in communication, at first the rules and boundaries seem to obfuscate, but, in actuality, they are just in the process of being redefined.
The "Pros" of Electronic Journalism
Regarding the highly argued topic of online medical journals and journalism, there are numerous "swords in the stone" of the Internet—most of them double-edged. For instance, electronic journalism will encourage editors to work harder to better their own printed journals. It will expose research and publication misconduct (e.g., fraudulent data, plagiarism, dual publications) because there is a huge database with ability to cross-reference.
A recent Canadian Medical Association Journal editorial regarding Varmus' first proposal, e-Biomed (4) notes several advantages: the almost instantaneous, cost-free, full-text access to biomedical research papers is by far the most exciting. Authors would still enjoy copyright, on the condition that their work would be freely available for transmission, downloading, and publication. Sites such as e-Biomed and PubMed Central would also ensure that electronic versions of journals will be made a part of permanent record.
Also, internal electronic storage and retrieval of information may save substantial public and private money; this would also free considerable space in libraries for other projects or archives.
Both commercial and association-based publishers have been criticized for making monopoly profits and extorting copyrights from authors. Although this may be true for some publishers—in all likelihood a small percent—most seem to be serving the best interests of the scientific community and the public. The addition of electronic media would provide additional competition, thus additional places for submission, tipping the balance of power away from journals and into the hands of the author. However, when all is said and done, this may also have the ultimate effect of weakening and perhaps destroying some of the existing tried-and-true print journals.
In his largely negative editorial in the New England Journal of Medicine (5), Relman concedes that websites that collect, organize, and retrieve data from clinical publications are undoubtedly of enormous aid to the population of increasingly computer-savvy practitioners. They can post material with technical details too lengthy for publication, such as whole-genome sequences or multicenter study results, or add multimedia portions to further illustrate findings (how better to study a case presented than to hear and see the history and physical being performed on the patient?). Readers of online journals can download data for analysis or interact directly with the author. Also, there is the common problem of wondering if there was any response to an article. Traditionally, without an index, one has to search through months and months of journals, whereas, electronically, it's just a point-and-click away; as are links to other related topics, resources, and references. In the past, all of this might or might not be available, depending on the whim and capacity of your local medical library.
And what if a reader is curious as to what occurred in the peer-review process? Electronic journals may opt to provide links to the article in its original form; they might supply the comments of the reviewers, the editorial committee, and the correspondences between them. Presently, much of the virtue of print journals surrounds the peer-review process; the claim is that they have properly qualified and unbiased reviewers who do the job fairly and accurately. — Well, here's an opportunity to publicly evaluate their worth. Peer review is a hot topic of debate in the print-vs.-electronic media arena. This may be due to the fact that, traditionally, editorial peer review was secretive, narrow, and somewhat arbitrary; the choice of reviewers is at the discretion of the editors (6). This process is dramatically different with online review. Some theorists speculate that additional peer review may now take place post-haste, within a virtual community of on-line experts, broadening the range of expertise in (and speed of) review (7). This process may continue even after publication of an article, and, unlike printed responses to articles, corrections and amplifications may appear immediately. This may place authors on an even footing with peer reviewers, providing checks on each others' performances. However, these ideas seem more idealistic that realistic, insofar as they leave many questions unanswered. Who would determine the experts from the anonymous users? Who would moderate the opinions, interpret the conclusions, and untangle the vast jumble of misinformation floating through the Web?
And last, but certainly not least, the financial cost of publishing on the Worldwide Web is about 30% lower than in print media. Many journals report paper, printing, binding, and postage accounts for over half their costs (8). Posting manuscripts electronically only incurs the expense of maintaining an electronic archive, which is less than two cents an article.
All this said, one thing seems certain: where clinical medicine, scientific research, and the Internet collide, clarity, credibility, and personal responsibility come into question.
The "Cons" of Electronic Journalism
Bioethically and "Hippocratically," using the Internet as one's primary method of gathering information in clinical practice is unwise. It is hard to deny—immediate practice of what is preached over the Web may be hazardous to one's health. Those familiar with the Internet have heard of cautionary tales of online predators prowling on youngsters and con artists peddling get-rich-quick schemes. The Web itself seems to breed a certain pathology, causing some individuals, believing themselves anonymous and thus untouchable, to fake medical data as well as illnesses. Accessibility of online information also makes it easier for people feigning illnesses: malingerers, those with factitious disorders such as "Munchausen syndrome," or drug abusers vying for pain medications, to get details about their supposed symptoms. Marc D. Feldman, M.D., a psychiatrist and expert on factitious disorders, has documented accounts of "virtual factitious disorders." He believes online support groups provide fertile ground for people who fake having an illness in order to gain attention, sympathy, or nurturing (9). Overall, taking the suggestions from support groups or chat rooms is unwise. Material from the Web may be of dubious value. Like a public bulletin board, its claims are often unreliable. Therefore, caveat emptor; let the buyer beware. An appropriate caution, even when what's for sale is information.
Relman's pessimistic editorial points out the potentially "disastrous" consequences that electronic distribution of articles may have on clinical medicine (5). He believes the "proposal fails to recognize fundamental differences between basic-science and clinical journals," and goes on to state that basic-science literature is used exclusively by working scientists, those familiar with a given field. Such scientists (for example, a physicist) would want rapid, efficient, inexpensive (electronic) posting of such data. It is unusual for the busy clinical psychiatrist to need such minutiae. Classically, what they need is review and keen assessment of what has been tested and applied in practice. Also, basic scientists rarely deal with dilemmas such as patient safety or economic or public policy. Clinical medicine confronts these issues daily. The medical community, especially those in direct contact with patients, form a readership that is not concerned with methodology; nor do they want their e-mails cluttered by journal-like presentations boosting panaceas and charlatanism that could result in potential harm. Like newspapers, electronic media often emphasize the "human interest" angle of scientific findings, an approach researchers in the field may not prefer. What they need is accompanying editorials in clinical journals to help them interpret the data and place the studies in the context of their own practices. As Relman stated, "Mistakes, inaccuracies, and misinterpretations in clinical research pose a far greater risk to health and the public welfare than [they] do in basic-science research."
On the other hand, the typical printed medical journal has a thorough, yet cumbersome, review process. It is standard that each article be read by at least two editors and often by several others, and by two or three peer reviewers in addition to, as warranted, a biostatistician. Then it is copyedited; there are checks of figures and grafts, structure, and internal consistency. Finally, it is published. Although many online journals unabashedly avow to preserve this process, the potential is there for electronic publications to ignore peer review in order to cut costs. Editing is not free. Fifty percent of a typical printed journal's costs are incurred in processing manuscripts. Most journals rely on subscriptions and advertisements to cover costs (biotechnology company marketing and advertising is already prevalent online). Releasing articles directly to online services would cause many libraries to cancel their subscriptions, justifiably, perhaps, if the information is accessible for free on the Internet. Medical journals face the whole issue of free vs. online paid subscriptions, which can cause publishers to perpetually fear that they may compete with themselves and lose paper subscriptions.
And one more aspect that, undoubtedly, biotechnology companies are aware of: science news influences Wall Street. If results of key studies are released to the public before critical reviews, stocks prices could, theoretically, be affected. One example of this phenomenon was when ICN Pharmaceuticals, manufacturers of the antiviral agent Ribavirin, called a press conference to announce that they had found the drug to be effective in slowing progression of infection of human immunodeficiency virus (HIV). Data were said to be forthcoming. HIV patients' hopes were raised, as was the price of ICN Pharmaceutical stock. Subsequently, the Food and Drug Administration found the claim to be unwarranted (10). More recently, on the coattails of the e-Biomed announcement, came the allegations that an associate of the owners of The New England Journal of Medicine, the Massachusetts Medical Society, was heavily invested in an Internet health-oriented website. Not only could this venture be considered a conflict of interest (a promoter of unbiased medical studies investing in a for-profit medical company), but also, the situation exposes the real concern that print media's presumed preeminence can be tarnished by the prospering electronic media. In part, scientific journals have a respectability because of their presumed immunity to financial market influences. Readers trust established publications and often use them to guide their medical decisions; they assume these reports are not influenced by business considerations or the price of a certain stock. Peer-reviewed sites under NIH or other respected leadership may thwart these assumptions. Both the proposed e-Biomed and PubMed Central would be "NIH-sponsored," yet would encourage posting of reports that have had no meaningful evaluation (peer review) or tested conclusion.
Again, peer review rears its problematic, multifaceted head; and dissatisfaction with this process is growing. In a survey of scientific deception, Grayson (11) found that peer review "has been widely criticized for conservatism, professional prejudice, and sloppy standards. Moreover, journal peer review is under enormous stress because of the explosion in scientific publishing." This was supported in an article by the editor of Nature, John Maddox (12), who said, "Misconduct of some overt kind is on the rise. We all know why that is. Reputations rest on publications as never before, as do promotions and research grants."
This, too, may be the primary factor forcing a change in the nature of science. Whereas the grants supporting research were once bestowed by government or academic institutions that were supposedly not financially motivated, a major source of funding is now coming from corporations with commercial interests. Researchers sponsored by these companies are finding that their scientific results are not becoming public information, but are being appropriated by interest groups seeking monetary gain. Only a few researchers seem to be remarking on these changes (13), even though they represent a radical change in the traditional practice and publication of scientific research.
In a recent article (6), the pros and cons of peer review on the Internet were evaluated, and five major issues arose:
How to rectify these issues remains unclear. The fallback position of most authors is to just use this new electronic medium as a paperless journal, using the same time-tested protocols and venerable practices leading to publication. This may not be such a bad idea, given that the virtues of print and electronic journals are largely the same.
In The Lancet's "Guide to the Internet" (14) Horton extolled the four major virtues of traditional, printed medical journals. First, journals report who did what, and when. Second, a journal legitimizes research through the overall quality of its editorial process, especially peer review. Third, the journal acts as an archive, a historical record from which to research and learn. Fourth, a journal distributes research to its readers. In other words, journals serve communities. The success of a journal depends on how well it serves its community.
This fourth virtue is hard to replace with the Internet. The online, virtual community is different from that which would serve, say, The Croatian Medical Journal—which has an essential national role in reconstructing and strengthening Croatia's medical and scientific culture. Thus, as one writes up a case report in Baltimore or Boston, the Internet author may now need to contemplate the implications of treatment of such cases in Bangladesh or Botswana. One must not forget, of course, that the United States is on the crest of the electronic wave and that most areas of the world are still "working without a(n) (Inter)net." In these regions, as within our own, more crucial than the pure speed of information-dissemination is the evaluation of the implications that information will have on health care.
General medical journals play a role in helping their specific readers by interpretation and discussion of the data presented. The weeks between acceptance and print publication in journals allow commentary and interpretation to develop. The result is editorial commentary, letters to the editor, updates, and notices of corrections. Will the Internet provide this? More important, will the Internet be able to provide quality medical information to its community? (See Appendix 1.) The answer is undoubtedly "Yes and No": some sites will, some sites won't.
What Web journalism may do, however, is put some print journals out of business. More formats for publishing means more competition, and online rivals (bolstered by big investors' financial support) are already beginning to pay generous fees for publishing key articles. This is unlike conventional print publications, who generally believe in intellectual or ethical reparation (that distributing helpful or insightful information is payment enough) or retroactive reimbursement (that publication validates one's work, advancing one's faculty position, or prompting further research grants). Advance payment for articles by online journals would certainly cut down on the number of submissions to traditional journals.
The above begs the question: since the Internet is rapidly becoming a consumer-based tool, will the interpretation—and value—of medical research for physicians and their patients merge with business and marketing? A more appropriate question (one that will not be addressed here) may then be, "Hasn't it already?"
Predicting the Future of Writing on the Internet
As I have stated previously, the scientific community has always been paper-bound. Difficulties with production, distribution, archival storage, and duplication in the past were overlooked because, after all, there was no alternative. Now that there is, and the latent expenses of being paper-bound weigh heavy on library budgets, electronic publications are becoming welcome alternatives. So what's the problem? Just convert written to electronic format—writing is writing, right? Unfortunately, the future of scientific writing is not so simple.
"Never make forecasts, especially about the future," warned famed Hollywood producer Sam Goldwyn. Warning heeded, it's still tempting to predict that medical journalism is fast becoming just another tangled knot in the Web. And the effect the Web will have on scientific writing is anyone's guess. However, some speculations seem to be coming true. The addition of images and sound to articles, multimedia-like presentations linked to references and other articles, and links to sponsors—all are already occurring. In my readings of online journals, I have also noticed an increased length of articles. This may be due to lack of limitations, such as costs of paper printing, distribution, etc., or it may be the result of less critical editing in e-journals.
Is it useless to plead for brevity in medical writing on the Internet? Nowhere else do facts get more clothed in elegant pleonasm than in the electronic media. "If we thus relegate ornament and redundancy to a subordinate place in medical writing, what shall we say of obscurity, and [the] wrong use of words?… wherever the writer of a very long and tiresome paper puts his conclusions at the end, in the form of a synopsis or summary, that part is not only read first, but is frequently the only part read by anybody except the writer and the proofreader. However, short papers are always read, and the great improvement in our medical literature in the past few years… is doubtless owing to the tendency of the time toward pithy articles." Such was the cry of an outraged physician at the Journal of the American Medical Association in 1894, at which time medical journals were beginning to experience their first pains of an increase in printed information (15).
Many have argued that medical writing should be clear, simple, and unambiguous (16); that long-winded "more scientific" words should be replaced with more common phrases. All this is true, but will this change as much as it has in the last 100 years?
More forecasting: what is likely on the Internet is that an article's site, once it has appeared on the Web, will consist of two parts: 1) the scientific report (similar to the present); and 2) an attached area, accessible to clinicians and anyone else interested in commenting, critiquing, and asking lots of questions regarding the nature of the study or ramifications of its findings. This may lead to a lot of rumors—again, we stress the importance of editing and peer review.
This should not be a surprise to us, this changing of the guard. The format for scientific writing has always been unstable. The IMRAD format (Introduction, Materials/Methods, Results, And Discussion manuscript structure) that we currently take for granted began in the 1960s; the structuring of research reports (abstracts, etc.) is even newer. The Web is only the latest phase of this process of continually updating.
As with one's child, as with one's computer, the consequences of human creations have never been fully appreciated except in hindsight. The birth of the computer age is behind us, and the technology has matured to a degree that the consequences are now beginning to be grasped. The changes it has brought about in our lifestyle are greater than many predicted. And, again, the analogy is appropriate: as with children, we are having to learn to adapt to the presence of the computer and the Internet, constantly continuing dialogues and working out problems so that we may operate synergistically with these new beings. Let's hope we can be enthusiastic participants in this undoubtedly changing electronic era, rather than become paralyzed by Bertrand Russell's controversy over whether this bit of progress should be considered ethical.