Publishing a Clinical Trial


Fig. 9.1
CONSORT 2010 checklist of information to include when reporting a randomized trial

In general, the manuscript should focus on what is novel about the study and what it adds to the body of medical literature. Write clearly and succinctly, with minimal jargon, and keep in mind the journal’s target audience. For this reason, many investigators choose the journal to which they will be submitting their manuscript prior to beginning the writing. When choosing a journal, it may be helpful to review back issues to get a sense of whether or not the trial is consistent with the topics covered by the journal. Asking more senior colleagues for recommendations or contacting journal editors directly can be helpful in gauging a journal’s interest in and likelihood of publishing the trial results. Every major journal has their author instructions freely available on their prospective website. Follow those instructions regarding length and format exactly as they are stated. Failure to do so may result in rejection of the manuscript without further review.

For those whose primary language is not English, it may be necessary to utilize a professional editing service or to enlist the help of coauthors or colleagues with a strong command of the English language. Be sure that there is no patient-identifying information in the manuscript. Many journals offer online supplements/appendices to the printed article. Take advantage of this chance to ensure transparency and reproducibility of the research.

The title should be descriptive of the study design and the scientific question being asked. If the trial is a randomized controlled trial, then that should be stated in the title. Avoid catchy titles and puns—these are best reserved for other settings, such as editorials. Typically, the rest of the manuscript will contain the abstract, introduction, methods, results, discussion, and funding/disclosures section.

Often it is best to write the abstract after completing the rest of the manuscript. It should begin with one or two sentences describing the context and importance of the trial, followed by a general overview of the methods of the trial. The results component of the abstract should specifically state the findings of the prespecified primary outcome. The conclusion should offer interpretation of the findings and their significance. The overall direction of the manuscript, from introduction to the discussion sections, should parallel what is written in the abstract.

The introduction of the manuscript should not be a thorough literature review, nor should it describe what is already well known by the readership. Rather, it should adequately provide the context and frame the motivation for undertaking the trial. The hypothesis should be stated clearly and unambiguously.

The methods section should be explicit about the trial design, the primary and secondary outcomes, and the arms of the study. Clearly defining the inclusion and exclusion criteria allows the reader to draw appropriate conclusions regarding the study’s external validity. The randomization process, sample size calculations, and statistical analysis methods should be delineated in the methods section. To describe the study simply as “randomized” without further information is inadequate reporting. The details of the randomization process are important for the reviewers and readers to determine the possibility of bias, particularly selection bias. Whether patients were placed in a study arm by random assignment, through randomized blocks, or an adaptive allocation processes must be described, as should any stratification strategies that were employed. Allocation concealment, if any, from patients and or study staff should be described as well. Measures taken to reduce bias in the randomization and allocation concealment process strengthen a study’s internal validity, and surprisingly, many authors fail to report them even when they were performed [32].

The results should begin by accounting for all patients in the study, most often in the context of a flow diagram. Baseline demographics of the subjects in the various study arms should be described. The results of the analyses that were prespecified, including the primary and secondary outcomes, are to be presented in this section. Any other analyses should be presented here as well, but their post hoc status should be clearly stated. Any deviations from the trial protocol and/or usual practice should also be explained.

The discussion section should begin with a brief summary of the findings of the study, followed by a comparison of the findings with those of other previously published studies. Conclusions drawn about the study must be consistent with the results of the trial. Avoid the tendency to over-extrapolate the study’s implications. An honest appraisal of the trial’s limitations and strengths should be delineated in this section as well.

Images, figures, and tables represent an opportunity to express the components and findings of the study in a visual manner. Data are often more easily presented visually than in text and are remembered better by the reader. In fact, many readers (as well as reviewers and editors) focus on these and the abstract before or in place of reading the text, underscoring the importance of making them as visually compelling as possible. Numerous texts and articles have been published on the effective presentation of data, and the reader is referred to them for in-depth details [3336]. Figure 9.2 demonstrates examples of good and bad visual displays of data. In general, the figures should be clear, as simple as possible to convey effectively the intended message, be of high resolution, and be submitted in the target journal’s preferred format. Some journals will allow color printing, but may charge the authors for that amenity. Figures ought to be stand-alone, meaning that the reader should be able to understand the investigator’s message without having to read the rest of the article.


Fig. 9.2
Poor and good examples of visual displays of fictitious data. Panel a displays survival data, but the figure is mostly empty space. The title is not descriptive, and the legend and the axes lack captions. Panel b displays the same data; however, instead of displaying survival, death data are shown, which, in this case, allows for better use of space. In addition, the legend and title are more descriptive. The axes have captions, and the P value for the difference between curves is displayed. In pane c, there are abbreviations used that are not described in the figure. The figure lacks a title, there is unnecessary use of three-dimensional effects, and the data are displayed in a way that does not allow the reader to determine the incidences of the complications. The use of color may not be helpful when printed in black and white. Panel d has a title, avoids unexplained abbreviations, will be helpful for the reader when printed in black and white, and allows the reader to determine the individual complication rates and still compare between repair type

References are most easily handled with the reference management software of the author’s choice. There are many options on the market of varying capabilities and price, and most allow the user to choose the reference style to be consistent with the target journal’s requirements, avoiding the hassle of manually reformatting them.

During the writing process, the help of a trusted colleague who is not intimately involved in the trial can be invaluable. For those immersed in the details of a trial, some aspects may seem clear or obvious, but that may not be true for those who have not read the protocol and been involved in the trial. A colleague’s feedback about the level of detail provided in the manuscript and the clarity of the writing can help point out areas worthy of a more careful explanation. It goes without saying that a thorough review for grammar and spelling, ideally by more than one person, is mandatory prior to submitting the manuscript. While the authors will have a chance to review proofs prior to publication, inattention to spelling and other grammatical details may raise questions in the reviewer’s minds as to the investigators’ attention to details in the planning, conduct, and analysis of the trial as well.

While most people recognize that copying text verbatim from someone else’s work without utilizing quotation marks constitutes plagiarism, fewer recognize that plagiarism is the use of another’s words or ideas without appropriate attribution or permission. There are various forms of plagiarism, including plagiarism of ideas; plagiarism of text verbatim; mosaic plagiarism, which involves weaving one’s own words and ideas into those of another without proper citation; and self-plagiarism [37]. Perhaps under-recognized by researchers, utilizing one’s own previous work without appropriate citation is considered plagiarism. The availability of software today allows journals to easily and accurately identify plagiarism. At best, it can result in a delay in publication or manuscript rejection, though it constitutes academic misconduct and can have even more dramatic ramifications for the offender. A thorough understanding of plagiarism is key to preventing it.

9.6 Responding to Reviewers

Very rarely is a manuscript accepted as is without revisions. Successful publications often will have required at least one round of revisions. Receiving criticism from the reviewers can be frustrating; however, the feedback should be viewed as an opportunity to improve the manuscript. Keep in mind that reviewers take the time out of their busy schedules to offer this feedback, and they are not compensated to do so. Be appreciative of their time and efforts. When responding, be sure to respond to every single point thoroughly and diplomatically.

9.7 Commit to Publishing and Do So in a Timely Manner

Regardless of the results of the trial, commit to publishing them. For previously enumerated reasons, there are ethical grounds for publishing the results of human subjects research, particularly clinical trials. Several studies have confirmed that only approximately half of all clinical trials are eventually published several years after completing the trial [38, 39]. Trials with null results are even less likely to be published, and for those that are, the delay is even longer than for trials with positive findings. The onus is on investigators to publish the results in a timely manner, irrespective of what they are.

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

Sep 26, 2017 | Posted by in GENERAL & FAMILY MEDICINE | Comments Off on Publishing a Clinical Trial

Full access? Get Clinical Tree

Get Clinical Tree app for offline access