Abstract thoughts

Abstract

See Article p. 275.This issue of Gastrointestinal Endoscopy contains a report that should draw your attention to particular aspects of the development and dissemination of new knowledge in the field of endoscopy. The study by Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar on the acceptance of abstracts for presentation at the ASGE sessions during Digestive Disease Week (DDW) and subsequent publication of these summaries as complete manuscripts is intriguing and, given the number of questions it raises, provocative. For those who participate in this process, myself included, the data these investigators have obtained will in many respects be troublesome.Before considering specific implications of this study it will be useful, perhaps necessary, to describe certain links in the chain of mechanisms that constitute the machinery for dissemination of new information. The data obtained by Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar in fact reflect the actions of the Scientific Program Committee for 1993-1994. Based on experience as a member of this committee in prior and more recent years, it is probable that the program committee for the year examined did not differ in any important aspect from those of other years. As with all ASGE committees, the President-Elect nominates individuals for membership and the actual appointments are confirmed by the Governing Board of the Society. There is, to the best of my knowledge, no established qualifications or credentials for membership on the Scientific Program Committee. The selection of committee members, in my experience, is a difficult task and the President-elect must consider a range of factors, not the least of which is the willingness of nominees to undertake the work. Most incoming presidents strive to find capable individuals who have a background in research, although there is no doubt a desire to field a committee that is representative of the Society at large, which is to say that political considerations cannot be completely ignored.The process whereby a decision is made to accept or reject an abstract for presentation at the scientific sessions of the Society differs markedly from the peer review process that is fundamental to the decision to accept or reject a manuscript for publication in a journal. Two major factors distinguish these widely dissimilar processes: time constraints and space limitations. Thus, on a certain date the program committee member receives a large notebook containing in most cases hundreds of abstracts along with a set of instructions for assigning a number that reflects the opinion of the member as to whether the abstract should be accepted for presentation. Because the number of submitted abstracts is great, committee members are organized into subgroups to review particular subsets of abstracts. For example, it is usual that one group will rate only those abstracts that deal with GI bleeding. The number of members in each subgroup varies, usually between 3 and 4. Because the dates of the DDW meeting and for publication of the abstracts are established in advance, the scientific program committee must meet and complete its work by a certain date.Because the length of each and every abstract is fixed and uniform, it is often necessary to make a judgment based on limited information. Moreover, many are poorly written. After each subgroup member has ranked the abstracts independently, the scores submitted by all of the subgroup members are averaged and the abstracts are listed in rank order. Once the rank order is established, members of each subgroup can discuss the merits and ranking of the abstracts. This takes place formally in a meeting of the subgroups just before the meeting of the full program committee. The relative merits of each group of abstracts are then discussed by the committee at large. Usually, each subgroup will propose one or more abstracts for inclusion in the plenary and other sessions. Often a topic forum will have been planned in advance based on the number of abstracts in a particular area or topics that the chairman and others consider timely. Throughout this process, the members are blinded as to the names of the authors of the abstracts as well as the institution(s) where the work was performed.The data presented by Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar indicate that about half of the abstracts submitted to the ASGE in 1994 were accepted for presentation at one of the DDW sessions of the Society. A small fraction (16%) of these were platform presentations. Almost 90% came from university or university affiliated institutions and more than three-fourths were from the United States (followed in order by France, Germany, and Japan). The investigators concluded that almost half referred to prospective studies although less than 10% were randomized trials and less than one-fourth claimed statistically significant results. Only 10% of the abstracts were classified as pediatric in nature.Depending on the type of statistical analysis, several factors were found by Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar to be associated with acceptance of an abstract for presentation. These include prospective studies, randomized studies, studies from university and university affiliated centers, and pediatric studies. Factors that were not found to be associated with acceptance included origin from the United States, multicenter studies, and studies with statistically positive results.On the whole, it appears that the committee was reasonably discriminating in its choices. That university centers were associated with acceptance seems reasonable, inasmuch as the majority of submissions came from such institutions. It is gratifying that prospective and randomized data were associated with acceptance. Although the fraction of abstracts that claimed statistically significant results was small, it may seem, at first glance, incongruous that there was no association between statistical significance and acceptance for presentation at the meeting. However, it is likely that Eloubeidi et al.1 did not, or could not, assess the veracity of the claim of statistical significance in each instance. It is more likely that they accepted the word of the author that the results were significant. However, it is not so difficult to come up with a p value that indicates statistical significance despite poor study design. Thus, the lack of association between statistically significant results and acceptance might merely indicate that the committee assigned a greater value to well-designed studies.The association between the category pediatric study and acceptance illustrates another major difference between abstracts and manuscripts. The scientific meeting, for which the abstract is intended, invariably has a range of objectives in addition to the presentation of valid information. A manuscript, to be considered for publication in a journal, must stand or fall, ideally, on merit alone. Thus, the seeming preference for pediatric abstracts undoubtedly arises from a desire on the part of the organizers of the scientific program to provide a forum that would attract pediatric gastroenterologists. In fact, putting a meeting together is as much art as science. As Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar point out, some sessions are clearly designed to attract attention to new technology. These sessions are usually well attended because, after all, that which is new and exciting invariably attracts attention. However, the harder issues, clinical benefit for example, are often submerged in the enthusiasm. When these questions ultimately surface during the peer review process, that which once seemed so novel may not in fact stand up to critical analysis.And now, the really intriguing aspect of the study by Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar: only one-fourth of all abstracts submitted from 1993 to 1994 were ultimately transformed into a published article. Moreover, of the abstracts accepted for presentation, only 32% became part of a published article; for abstracts that were deemed suitable for an oral presentation, only 40% ultimately became published articles. Of the abstracts not accepted for presentation, 16% eventually became reports published in full. It is difficult to know what to think about these numbers, except that they are at best bothersome. According to the investigators, this is the lowest rate of publication among medical societies for which data are available.Abstracts that were more likely to become part of a published article included those that were accepted for presentation, those that presented prospective data, studies reporting statistically positive results, and studies from outside the United States. Furthermore, Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar found that abstracts accepted for oral presentations, those presented at plenary sessions, and those that were concerned with multicenter or pediatric studies were not more likely to be published ultimately as an article. The failure of abstracts presented orally at plenary sessions to find their way into publication as an article seems surprising, even illogical, or at least difficult to explain.Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar speculate as to the reasons for these results, referring to data on publication of abstracts from other scientific meetings. There are a number of possible explanations for the low rates of subsequent publication of abstracts accepted for presentation, plenary presentations in particular. It is possible that many were submitted but not accepted for publication. With regard to abstracts submitted to the ASGE portion of DDW, no data are available. However, rejection seems an unlikely explanation. Perhaps many of the studies outlined in abstract form are still in progress after 4 years; also unlikely. Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar quote the results of several studies of abstracts submitted to other meetings which indicate that failure to publish is attributable in general to a lack of time and interest on the part of the author of the abstract and in some cases inadequate resources. For about 40% to 50% of the abstracts, the studies quoted by Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar suggest that a corresponding manuscript was never prepared. It is probably safe to assume that some of these studies were in fact never completed. It is of interest that some available data appear to indicate a correlation between the academic rank of the senior author and subsequent publication of the data contained in an abstract; that higher rank increases the likelihood of publication. This seems plausible and perhaps a senior author of greater academic rank brings not only experience but also resources in terms of time and funding as well as guidance, motivation, and expectations.Although the majority of abstracts submitted to the ASGE for the years studied by Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar came from institutions within the United States, abstracts from other countries were more likely to be published. In fact, the percentage of manuscripts submitted to Gastrointestinal Endoscopy from countries other than the United States is higher (presently 63%) than suggested by Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar It is no secret that academic practice in the United States, the origin of most abstracts studied by Eloubeidi et al.,1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar is being subjected to pressures and constraints that are seen by many as threatening to the fundamental academic mission. To what extent these changes account for the lack of time, interest, and resources that are thought to explain the failure to ultimately publish data presented initially in abstract form is worthy of further consideration.What conclusions should be drawn from the data presented by Eloubeidi et al.?1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google ScholarUnfortunately, the attempt to gain insight by comparing the methods of the ASGE to other societies is interesting but not especially helpful. There are doubtless substantial differences among individual societies with regard to the objectives and organization of their scientific meetings including their methods and criteria for selection of abstracts. Because the ASGE devotes considerable resources to the promotion of endoscopic research, especially by fellows and junior faculty members, further study of the process by which endoscopic research is conducted and the results disseminated would be beneficial, particularly the failure to publish studies presented at scientific meetings of the Society.A scientific meeting represents a unique opportunity for interaction. In addition to the dissemination of information, such meetings offer an opportunity for scientific and professional interchange with the implicit goal of advancing the knowledge and treatment of disease. Unfortunately, the results of Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar compel the question of whether this goal is actually being achieved. How often have you heard it said, or said yourself, “DDW is just too big.” Undoubtedly, this statement has a range of possible meanings, but for most it conveys a sense of being overwhelmed, an inability to decipher useful, reliable information from the hubbub and racket. In this respect, the study of Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar is not encouraging. Whether the centrifugal character of the present day DDW can or should be altered is perhaps a question that cannot be addressed. But it is worthy of note that the scientific meeting was invented at a time when mass communication was largely by means of that which is now called the “hard copy.” In the present age of paperless, instantaneous, even intrusive communication it will be prudent to ponder the relevance of the large scientific meeting. If so little value is attached by so many investigators to the work they present at DDW that they fail to publish it, where is the value in the meeting?The most obvious and perhaps most important caveat that can be derived from the data presented by Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar is this: be wary of data that are presented only in the form of an abstract. Whether abstracts that are more than 4 years old should be quoted at all in scientific publications is problematic. For articles published in Gastrointestinal Endoscopy the editors presently request that authors simply indicate whether a cited reference is an abstract. (Perhaps this should be an injunction rather than a guideline.) The meta-analysis, despite its limitations, is a useful method for deriving conclusions from data, especially when available data related to a particular question come from studies of marginal statistical power with respect to numbers of patients. It seems reasonable that abstracts, especially aging abstracts, should be excluded from this type of inquiry. Although the majority of abstracts accepted for platform or poster presentations do not ultimately become published articles, much of the data and information presented at scientific meetings is widely disseminated by way of the medical tabloid. Perhaps the most disconcerting aspect of the data of Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar relates to readership surveys conducted by publishers that show that these types of publications are read by large numbers of physicians. It is perhaps worth noting, however, that these newsletters are journalistic in nature.Although they may accurately report information presented at a meeting, this is in no manner or fashion equivalent to publication of an article after peer review. As pointed out by Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar there are data that show that only about one-third of published articles contain data that are similar to that presented in the original abstract. To what extent the readers of these tabloid publications view them as reliable sources of information for use in clinical practice is unknown. But given the results of Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar this approach to the acquisition of information has serious flaws. Without doubt, for the practicing physician the most important message from the study of Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar is, unfortunately, to be circumspect in using information obtained at scientific meetings in the care of patients.Perhaps it is possible to make improvements in the process whereby plenary and poster sessions are organized. However, the constraints of time and space inherent to the review of abstracts cannot be changed so that the process can never be equivalent to peer review. Perhaps DDW registrants are sophisticated enough to have already recognized the problems delineated by the study of Eloubeidi et al.,1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar in which case change is perhaps unnecessary. What is certain, however, is that an effort should be made to understand the failure to publish data presented at the DDW sessions of the ASGE. Perhaps this could be in the form of a request for proposals with allocation of funds for such a study. With such knowledge, and the will to improve, it may be possible to bring a greater integrity to dissemination of scientific information. See Article p. 275. This issue of Gastrointestinal Endoscopy contains a report that should draw your attention to particular aspects of the development and dissemination of new knowledge in the field of endoscopy. The study by Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar on the acceptance of abstracts for presentation at the ASGE sessions during Digestive Disease Week (DDW) and subsequent publication of these summaries as complete manuscripts is intriguing and, given the number of questions it raises, provocative. For those who participate in this process, myself included, the data these investigators have obtained will in many respects be troublesome. Before considering specific implications of this study it will be useful, perhaps necessary, to describe certain links in the chain of mechanisms that constitute the machinery for dissemination of new information. The data obtained by Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar in fact reflect the actions of the Scientific Program Committee for 1993-1994. Based on experience as a member of this committee in prior and more recent years, it is probable that the program committee for the year examined did not differ in any important aspect from those of other years. As with all ASGE committees, the President-Elect nominates individuals for membership and the actual appointments are confirmed by the Governing Board of the Society. There is, to the best of my knowledge, no established qualifications or credentials for membership on the Scientific Program Committee. The selection of committee members, in my experience, is a difficult task and the President-elect must consider a range of factors, not the least of which is the willingness of nominees to undertake the work. Most incoming presidents strive to find capable individuals who have a background in research, although there is no doubt a desire to field a committee that is representative of the Society at large, which is to say that political considerations cannot be completely ignored. The process whereby a decision is made to accept or reject an abstract for presentation at the scientific sessions of the Society differs markedly from the peer review process that is fundamental to the decision to accept or reject a manuscript for publication in a journal. Two major factors distinguish these widely dissimilar processes: time constraints and space limitations. Thus, on a certain date the program committee member receives a large notebook containing in most cases hundreds of abstracts along with a set of instructions for assigning a number that reflects the opinion of the member as to whether the abstract should be accepted for presentation. Because the number of submitted abstracts is great, committee members are organized into subgroups to review particular subsets of abstracts. For example, it is usual that one group will rate only those abstracts that deal with GI bleeding. The number of members in each subgroup varies, usually between 3 and 4. Because the dates of the DDW meeting and for publication of the abstracts are established in advance, the scientific program committee must meet and complete its work by a certain date. Because the length of each and every abstract is fixed and uniform, it is often necessary to make a judgment based on limited information. Moreover, many are poorly written. After each subgroup member has ranked the abstracts independently, the scores submitted by all of the subgroup members are averaged and the abstracts are listed in rank order. Once the rank order is established, members of each subgroup can discuss the merits and ranking of the abstracts. This takes place formally in a meeting of the subgroups just before the meeting of the full program committee. The relative merits of each group of abstracts are then discussed by the committee at large. Usually, each subgroup will propose one or more abstracts for inclusion in the plenary and other sessions. Often a topic forum will have been planned in advance based on the number of abstracts in a particular area or topics that the chairman and others consider timely. Throughout this process, the members are blinded as to the names of the authors of the abstracts as well as the institution(s) where the work was performed. The data presented by Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar indicate that about half of the abstracts submitted to the ASGE in 1994 were accepted for presentation at one of the DDW sessions of the Society. A small fraction (16%) of these were platform presentations. Almost 90% came from university or university affiliated institutions and more than three-fourths were from the United States (followed in order by France, Germany, and Japan). The investigators concluded that almost half referred to prospective studies although less than 10% were randomized trials and less than one-fourth claimed statistically significant results. Only 10% of the abstracts were classified as pediatric in nature. Depending on the type of statistical analysis, several factors were found by Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar to be associated with acceptance of an abstract for presentation. These include prospective studies, randomized studies, studies from university and university affiliated centers, and pediatric studies. Factors that were not found to be associated with acceptance included origin from the United States, multicenter studies, and studies with statistically positive results. On the whole, it appears that the committee was reasonably discriminating in its choices. That university centers were associated with acceptance seems reasonable, inasmuch as the majority of submissions came from such institutions. It is gratifying that prospective and randomized data were associated with acceptance. Although the fraction of abstracts that claimed statistically significant results was small, it may seem, at first glance, incongruous that there was no association between statistical significance and acceptance for presentation at the meeting. However, it is likely that Eloubeidi et al.1 did not, or could not, assess the veracity of the claim of statistical significance in each instance. It is more likely that they accepted the word of the author that the results were significant. However, it is not so difficult to come up with a p value that indicates statistical significance despite poor study design. Thus, the lack of association between statistically significant results and acceptance might merely indicate that the committee assigned a greater value to well-designed studies. The association between the category pediatric study and acceptance illustrates another major difference between abstracts and manuscripts. The scientific meeting, for which the abstract is intended, invariably has a range of objectives in addition to the presentation of valid information. A manuscript, to be considered for publication in a journal, must stand or fall, ideally, on merit alone. Thus, the seeming preference for pediatric abstracts undoubtedly arises from a desire on the part of the organizers of the scientific program to provide a forum that would attract pediatric gastroenterologists. In fact, putting a meeting together is as much art as science. As Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar point out, some sessions are clearly designed to attract attention to new technology. These sessions are usually well attended because, after all, that which is new and exciting invariably attracts attention. However, the harder issues, clinical benefit for example, are often submerged in the enthusiasm. When these questions ultimately surface during the peer review process, that which once seemed so novel may not in fact stand up to critical analysis. And now, the really intriguing aspect of the study by Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar: only one-fourth of all abstracts submitted from 1993 to 1994 were ultimately transformed into a published article. Moreover, of the abstracts accepted for presentation, only 32% became part of a published article; for abstracts that were deemed suitable for an oral presentation, only 40% ultimately became published articles. Of the abstracts not accepted for presentation, 16% eventually became reports published in full. It is difficult to know what to think about these numbers, except that they are at best bothersome. According to the investigators, this is the lowest rate of publication among medical societies for which data are available. Abstracts that were more likely to become part of a published article included those that were accepted for presentation, those that presented prospective data, studies reporting statistically positive results, and studies from outside the United States. Furthermore, Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar found that abstracts accepted for oral presentations, those presented at plenary sessions, and those that were concerned with multicenter or pediatric studies were not more likely to be published ultimately as an article. The failure of abstracts presented orally at plenary sessions to find their way into publication as an article seems surprising, even illogical, or at least difficult to explain. Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar speculate as to the reasons for these results, referring to data on publication of abstracts from other scientific meetings. There are a number of possible explanations for the low rates of subsequent publication of abstracts accepted for presentation, plenary presentations in particular. It is possible that many were submitted but not accepted for publication. With regard to abstracts submitted to the ASGE portion of DDW, no data are available. However, rejection seems an unlikely explanation. Perhaps many of the studies outlined in abstract form are still in progress after 4 years; also unlikely. Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar quote the results of several studies of abstracts submitted to other meetings which indicate that failure to publish is attributable in general to a lack of time and interest on the part of the author of the abstract and in some cases inadequate resources. For about 40% to 50% of the abstracts, the studies quoted by Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar suggest that a corresponding manuscript was never prepared. It is probably safe to assume that some of these studies were in fact never completed. It is of interest that some available data appear to indicate a correlation between the academic rank of the senior author and subsequent publication of the data contained in an abstract; that higher rank increases the likelihood of publication. This seems plausible and perhaps a senior author of greater academic rank brings not only experience but also resources in terms of time and funding as well as guidance, motivation, and expectations. Although the majority of abstracts submitted to the ASGE for the years studied by Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar came from institutions within the United States, abstracts from other countries were more likely to be published. In fact, the percentage of manuscripts submitted to Gastrointestinal Endoscopy from countries other than the United States is higher (presently 63%) than suggested by Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar It is no secret that academic practice in the United States, the origin of most abstracts studied by Eloubeidi et al.,1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar is being subjected to pressures and constraints that are seen by many as threatening to the fundamental academic mission. To what extent these changes account for the lack of time, interest, and resources that are thought to explain the failure to ultimately publish data presented initially in abstract form is worthy of further consideration. What conclusions should be drawn from the data presented by Eloubeidi et al.?1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar Unfortunately, the attempt to gain insight by comparing the methods of the ASGE to other societies is interesting but not especially helpful. There are doubtless substantial differences among individual societies with regard to the objectives and organization of their scientific meetings including their methods and criteria for selection of abstracts. Because the ASGE devotes considerable resources to the promotion of endoscopic research, especially by fellows and junior faculty members, further study of the process by which endoscopic research is conducted and the results disseminated would be beneficial, particularly the failure to publish studies presented at scientific meetings of the Society. A scientific meeting represents a unique opportunity for interaction. In addition to the dissemination of information, such meetings offer an opportunity for scientific and professional interchange with the implicit goal of advancing the knowledge and treatment of disease. Unfortunately, the results of Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar compel the question of whether this goal is actually being achieved. How often have you heard it said, or said yourself, “DDW is just too big.” Undoubtedly, this statement has a range of possible meanings, but for most it conveys a sense of being overwhelmed, an inability to decipher useful, reliable information from the hubbub and racket. In this respect, the study of Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar is not encouraging. Whether the centrifugal character of the present day DDW can or should be altered is perhaps a question that cannot be addressed. But it is worthy of note that the scientific meeting was invented at a time when mass communication was largely by means of that which is now called the “hard copy.” In the present age of paperless, instantaneous, even intrusive communication it will be prudent to ponder the relevance of the large scientific meeting. If so little value is attached by so many investigators to the work they present at DDW that they fail to publish it, where is the value in the meeting? The most obvious and perhaps most important caveat that can be derived from the data presented by Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar is this: be wary of data that are presented only in the form of an abstract. Whether abstracts that are more than 4 years old should be quoted at all in scientific publications is problematic. For articles published in Gastrointestinal Endoscopy the editors presently request that authors simply indicate whether a cited reference is an abstract. (Perhaps this should be an injunction rather than a guideline.) The meta-analysis, despite its limitations, is a useful method for deriving conclusions from data, especially when available data related to a particular question come from studies of marginal statistical power with respect to numbers of patients. It seems reasonable that abstracts, especially aging abstracts, should be excluded from this type of inquiry. Although the majority of abstracts accepted for platform or poster presentations do not ultimately become published articles, much of the data and information presented at scientific meetings is widely disseminated by way of the medical tabloid. Perhaps the most disconcerting aspect of the data of Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar relates to readership surveys conducted by publishers that show that these types of publications are read by large numbers of physicians. It is perhaps worth noting, however, that these newsletters are journalistic in nature. Although they may accurately report information presented at a meeting, this is in no manner or fashion equivalent to publication of an article after peer review. As pointed out by Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar there are data that show that only about one-third of published articles contain data that are similar to that presented in the original abstract. To what extent the readers of these tabloid publications view them as reliable sources of information for use in clinical practice is unknown. But given the results of Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar this approach to the acquisition of information has serious flaws. Without doubt, for the practicing physician the most important message from the study of Eloubeidi et al.1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar is, unfortunately, to be circumspect in using information obtained at scientific meetings in the care of patients. Perhaps it is possible to make improvements in the process whereby plenary and poster sessions are organized. However, the constraints of time and space inherent to the review of abstracts cannot be changed so that the process can never be equivalent to peer review. Perhaps DDW registrants are sophisticated enough to have already recognized the problems delineated by the study of Eloubeidi et al.,1Eloubeidi MA Wade SB Provenzale D Factors associated with acceptance and full publication of gastrointestinal endoscopic research originally published in abstract form.Gastrointest Endosc. 2001; 53: 275-282PubMed Scopus (85) Google Scholar in which case change is perhaps unnecessary. What is certain, however, is that an effort should be made to understand the failure to publish data presented at the DDW sessions of the ASGE. Perhaps this could be in the form of a request for proposals with allocation of funds for such a study. With such knowledge, and the will to improve, it may be possible to bring a greater integrity to dissemination of scientific information.