In an article in this issue, Wenger and Lieberman provide "an assessment of orthopaedic surgeons' knowledge of medical ethics." The authors note: "The respondents appropriately handled questions involving economic aspects, truth-telling, confidentiality, and an incompetent colleague. However, there was poorer understanding of proper ethical conduct with regard to informed consent." Having served as Editor of The Journal for twelve and a half years, and having been confronted with the difficulty that authors have understanding proper ethical conduct with regard to informed consent, the findings of Wenger and Lieberman come as no surprise to me.
Informed consent is not the only area that requires an understanding of proper ethical conduct by medical authors. Thus, the article by Wenger and Lieberman caused me to reflect on the ethics of authorship as they apply before a study is begun, while the study is performed, and during the period in which the manuscript is written and prepared for publication. My reflection included the logical steps that an author should take to perform and report a study in an ethical fashion as well as the ethics of multiple authorship.
The first ethical issue that must be considered is the reason for performing a study and for publishing its results. Basically, the major ethical reason for publishing material is to provide information that will help improve the care of patients. The second reason is to record the data from a well done study in an accurate fashion so that present and future generations of readers will be able to build upon solid research. Regrettably, many articles are written with the writer (and his or her curriculum vitae), rather than the reader and the patient, in mind. Many authors write to provide yet another version of previously published material rather than to advance scientific knowledge and improve patient care. When such writing is done for a textbook or for a specific review article, the ethics of authorship are not violated; on the other hand, when the material is repackaged for yet another journal article, the author's ethics must be called into question. Such dual publication takes up space in the medical literature that could have been devoted to primary publication. A more drastic problem is that dual publication wastes the reader's time. Often, the two publications have minor differences and the reader is forced to read both in order to obtain a full comprehension of the subject; sometimes, there are major differences and the reader must decide which article is correct. An equally disturbing variation of dual publication is so-called salami publication, in which the author releases information to the reader in a serial fashion by segregation of data. Although authors sometimes argue that they have too much data to include in one publication, readers deserve to have all material relating to a particular series of patients published in one place so that they do not have to search through a number of publications for information regarding a single group of patients. Frequently, salami publication, rather than providing additional thin slices of information, provides bologna.
The ethics of medical writing must be considered during each part of the process, from the planning of the study to the completion of the project. Each study begins with an idea, but before a study is undertaken the author should review the literature to be certain that the idea is a new one. There is an ethical and an unethical way to review the literature.
The ethical way demands that the author read the entire article, not just the abstract. It is appropriate to keep a written record of the full citation and a summary of the basic ideas contained in the article. Ethical behavior implies responsibility, and it is the responsibility of the author to be certain that a published article is cited correctly and that the information drawn from it is correct as well. Frequently, when manuscripts submitted to The Journal are reviewed, it is apparent that the author has not read the full article and does not understand its content. Often, I realize that the author has not read the full article because the reference given to The Journal is cited incorrectly, just as it was in a previously published bibliography. Citing a referenced work without reading it is unethical.
The next ethical step, before a study is begun, is to formulate a question to be answered. The cost of performing the study should be considered, as should the benefit to be derived from the conclusion that is to be obtained. It is unethical to begin a study when the author has no idea what benefits might result from it. Also, once the question has been formulated, the study must be designed to answer that question. This requires consideration of the proper demographics of the population to be studied; the best way to select patients to be evaluated; the appropriate clinical, laboratory, and radiographic studies; and the most suitable methods of statistical evaluation. It is unethical to start a study before all of these parameters are in place. For example, one should not perform a study and then contact a statistician afterward; the statistician should be involved from the start.
A concern of paramount importance that must be addressed prior to any study is the rights of the patients involved. Patients who are to participate in a study must be fully informed so that they can provide informed consent, and the study must be evaluated and approved by an Institutional Review Board before it is begun. While many authors understand the need for informed consent when two treatment programs are being offered in a randomized fashion, fewer authors understand the more subtle implications when such consent is obtained in other types of investigations. As I have stated previously3, I believe that all patients should be adequately informed about all aspects of a study even when they are involved in that study only to the extent that they are asked to return for a follow-up evaluation. Such follow-up evaluations may inconvenience patients or their families. The Institutional Review Board must ensure that patients are fully informed of their rights regarding their participation or non-participation in any study.
While some might question whether ethics are involved in the writing of the manuscript per se, I believe that there is an ethical way to prepare a manuscript for publication. First, the author must strive for accuracy when reporting the information obtained from the study; it must always be kept in mind that this information will be used by the reader to manage patients. It is imperative that the author take pride in what he or she does. Second best is not good enough when managing patients, and second best is not good enough when preparing material for publication. Thus, the author must write and rewrite the manuscript to ensure clarity. Coauthors have a similar obligation to aid in the writing, to make suggestions, to read the final manuscript, and to approve that manuscript before it is submitted. This is the ethical responsibility of each individual who is listed as an author of a manuscript.
Ethical behavior also comes into play when an author submits a manuscript. Ethical submission mainly involves telling the truth. Sadly, during the last year, there were a number of instances in which authors did not tell the truth during the process of submitting an article to The Journal.The Journal requests that each author specifically sign the following statement: "The author(s) further represent(s) that the article is original, that it is not under consideration by another journal, and that this material has not been previously published." During the past year, one group of authors signed such a letter and sent the same letter to two different journals. In addition, several authors specifically stated that the material that they submitted for publication had not been published previously, but it was later discovered that illustrations said to be original had been published in other journals.
The author must also exhibit ethical behavior during the process of revising a manuscript for publication. The revision process gives the author the opportunity to take advantage of the evaluation of peer reviewers who are experts in their field and of the advice of the Editor of The Journal. The reviewers and the Editor make suggestions so that the author can provide information to the reader in the most accurate and clearest way possible; these suggestions should not be taken lightly. Authors often assume that their major task is to satisfy the Editor, but this could not be further from the truth. Again, the purpose of the exchange during the revision process is to provide the best possible information to The Journal's readers. To be sure, the reviewers and the Editor may have made suggestions that are difficult, or sometimes even impossible, to address, but the author should make an honest effort to reply to questions that can be answered and to explain the reasons why others cannot be answered. The author should strive for perfection during the revision of the manuscript, as he or she should at every other stage of the preparation of the material for publication. Finally, the corresponding author should not return the manuscript to The Journal office until all authors are certain that all aspects of the article are correct. Final proofreading of the manuscript by all of the authors represents ethical behavior. Returning the manuscript to the Editor with persistent typographical errors, with the data in the abstract in disagreement with the data in the body of the paper, and with the data in the tables disparate from both suggests a lack of pride in the submitted work. A lack of pride calls into question the reason for doing the study and preparing the material for publication in the first place.
Authors must remember that The Journal's Instructions to Authors state: "It is to be clearly understood that each author has participated in the design of the study, has contributed to the collection of the data, has participated in the writing of the manuscript, and assumes full responsibility for the content of the manuscript." This ethical behavior is required of each and every author listed on a manuscript2. It is unethical for an author to allow his or her name to appear on a manuscript unless these criteria have been satisfied, yet this happens regularly. Also, many coauthors play no part in the revision process; such behavior is unethical as well. Any author of a manuscript should be able to defend the manuscript in a public forum.
Regrettably, fraud must be considered in any discussion of the ethics of medical writing. Most readers would like to believe that there is little fraud in the medical literature, but one only has to read Betrayers of the Truth. Fraud and Deceit in the Halls of Science, by Broad and Wade, to realize that fraud is rampant. The use of the word fraud is upsetting, but fraud comes in both large and small packages. Altering data by dropping out recorded points is one example of fraud.
Fisher, writing in 1936 about Mendel's work, stated that: "A serious and almost inexplicable discrepancy has, however, appeared, in that in one series of results the numbers observed agree excellently with the two to one ratio, which Mendel himself expected, but differ significantly from what should have been expected had his theory been corrected to allow for the small size of his test progenies." The fact that Mendel's data may have been altered is hardly comforting and is not a reason to suggest the alteration of data by others, yet authors often wish to drop outlying data from a study.
Finally, authors must continually be aware of the responsibility that they incur by submitting material for publication. They must remember that, because their work will be used by readers to manage patients, they must write each word, each sentence, and each paragraph with the same care and concern that they show when they manage patients. Thus, the authors' ethical responsibility requires clarity of thought, accuracy of the written word, and proper execution of the deeds of authorship in preparing a manuscript for publication.
Henry R. Cowell, M.D., Ph.D.