Turn a clinical study report into a manuscript
A journal manuscript is often produced from a sponsor's clinical study report rather than from the study itself. The writer has to find the key data and messages in a document that can run to thousands of pages, settle with the named authors how much of it the paper carries, and check what the authors add against it.
Working through the report
Producing a high-quality manuscript from a clinical study report can be even more challenging than writing a manuscript in general, because the writer has to deal with a report that may be of less than ideal quality, tease out the vital messages from what may be a tangle embedded in it, and juggle a team whose members may each be pursuing their own agenda for the manuscript.
Even seasoned medical writers can feel daunted when distilling a clinical study report into a 2,500-word manuscript for a peer-reviewed journal. Clinical study reports summarise all aspects of a clinical trial and can exceed 5,000 pages. In one freelance writer's experience they have become more sophisticated and less manageable over time, and a writer new to them can find them quite overwhelming. The same writer holds that, because a freelancer can charge only so much for their time, the work has to be as efficient as possible.
A report includes hundreds of pages of patient narratives, lists of the laboratories that took part in the trial and other data usually not relevant to the manuscript. Rather than scroll endlessly through these sections, the writer can use the report's clickable table of contents to go to specific sections. The synopsis, which typically spans several pages, highlights the methodology and results the writer should include in the outline and the manuscript. In one writer's view, the synopsis and the table of contents are the most useful parts of the report. The writer can search the report for specific words and combinations of words to locate information and confirm that results are accurate.
Programs such as Adobe Acrobat and doPDF let the writer highlight and add notes to the report, which is invaluable in a document of that length. Annotating the manuscript outline as well saves more time, because it removes the need to hunt through the report repeatedly for information. If the annotations leave the finished outline looking cluttered, an identical version without them can be shared with the authors for approval.
Identifying the key data and messages
Before writing the paper, the first step should be to identify the key data and messages it should communicate. As a starting point, it may be appropriate for the sponsor or an author, or the writer in early discussion with them, to physically highlight the key data, messages and methods in the source document.
The source may be one or more of several documents: a clinical study report and its associated tables, a congress presentation, or a file of tables generated post hoc that may have no explanatory document. Such a file of tables can be particularly confusing without appropriate guidance.
Where the paper is based on a clinical study report, the key data, methods and messages may be clear, but they are still often worth discussing. For example, if the report gives both mean and median values for the primary endpoint, the writer may need guidance on whether one or both belong in the paper. Similarly, even where the primary objective was to study efficacy, it may need discussing whether the paper should consider other important issues, such as cost-effectiveness, in relation to efficacy, even if neither the study nor the report's discussion addressed them.
The writer should use the CONSORT checklist for abstracts together with the clinical study report to identify the methods and results that must be included in both the abstract and the body of the manuscript.
The writer may ask the sponsor or authors for the key papers and, if necessary, to identify the key messages in them that belong in the introduction or discussion.
Deciding with the authors how much of the report to carry
A freelance writer is likely to be directed by the client on how to use the report to outline a manuscript, but under AMWA's Code of Ethics a writer would presumably develop an outline from a clinical study report only in close collaboration with the named author or as a co-author.
How information carries forward from the report to the journal article depends on the study's design and objectives and on the target journal's instructions for authors. The outline is not automatically introduction, methods and materials, statistical analysis, results, discussion, references and tables or figures. Rather, the writer has to evaluate the individual study protocol against the client's intentions and the target journal's article types, styles and instructions.
In the past, companies could and did cherry-pick data from a clinical study report for a journal article, most of the time the investigators were not given a copy of the final report, and the named authors frequently had little or no input to the paper, by their own choice. As of 2017, companies were required to be far more transparent about all clinical trial data, and most of the better medical journals published both in print and online. Extended methods, results, tables and figures could therefore be submitted as online supplementary material, so the writer could consider using a relatively large proportion of the report for the article.
Checking what the authors add
When an author adds or changes data in the paper, the writer should check the source files rather than presume the data are correct, as part of being proactive about completeness, accuracy and transparency in presenting the study's data and methodological details.
Where the writer sees an important issue, they should contact the sponsor or authors and, if they can, suggest possible solutions. For example, if an author gives the absolute numbers of subjects who experienced adverse events in the text without percentages, but the treatment groups were of unequal size, the writer might ask whether the percentages should be included.