Agree the mandate before you review someone else's text
An instruction to review someone else's text says little about how far the review should go, so its scope is agreed before the review starts: what kind of text it is, the capacity the reviewer reads in, how much authority each aspect of the text gives the reviewer, how much time there is, and what a specialist reviewer will and will not comment on.
Reviewing as part of the medical writer's job
Medical writers often spend as much of their time reviewing documents as writing them, yet the scope of a review is seldom as well defined as the task of writing. An assignment to write a clinical study report is usually understood to mean a final report complete with in-text tables, a hyperlinked table of contents and a list of abbreviations. A review by a medical writer often leaves room for much interpretation, so a priority should be to clarify the scope of the review and the time available.
Because of their close involvement with all of a project's documentation, the medical writers on it are often the only members of the team with an overview across documents. So the writer, in the regulatory area at least, inevitably takes on the function of reviewer and quality-control person, ensuring consistency across documentation and compliance with guidelines.
The kind of text and the reviewer's capacity
One experienced medical writer roughly divides the texts she reviews into these categories:
- an outsourced regulatory document owned by medical writing, such as a clinical study report
- a regulatory document owned by medical writing and written by a colleague
- a regulatory document not owned by medical writing, such as a statistical analysis plan
- a non-regulatory company document requiring official input from medical writing, such as a standard operating procedure
- presentation slides
- a publication
- a website or other text
Whatever the category, the reviewer also needs to define the capacity in which they are reviewing: as a medical writer, as a native English speaker, as a mentor, or perhaps just as a second pair of eyes. The role varies with the capacity, and with it the style and extent of the review. How much of the text the reviewer then permits themselves to correct depends on their role and on what they judge the extent of their responsibility to be.
The aspects of a review
The same writer divides the aspects of a review into what she considers formal requirements, content, consistency and language.
Formal requirements
Any regulatory document coming out of a medical writing department must adhere to the formal requirements of authority guidelines and company standard operating procedures and templates. On this level, medical writers should have the full authority to change the text and bring it in line, and doing so is their job. The correct numbering of paragraphs, tables and figures could be seen as a formal requirement too, but it is often reviewed as part of a technical quality check done separately.
Content
The extent to which medical writers can review the content of a text varies enormously, and hard and fast rules about how far their involvement should go are difficult to set. One experienced reviewer recommends that a medical writer go only as far as they are able and confident to go, and remember that most documents are the result of teamwork and that teams have their specialists. Medical writers can make an important contribution by asking questions of the other specialists, but any interpretation must be made with great care. A reviewer may notice points that trouble them even without the scientific, statistical or medical background to grasp the complexities fully, and this can be very helpful to authors. It is quite possible to perform a detailed review of all the other aspects of a text while understanding little of its content.
Consistency
Reviewers are often the ones to point out inconsistencies, and this aspect of reviewing is one of the most important, whatever the type of text. Inconsistencies look sloppy and can lead to confusion, but ironing them out can be painstaking work that invariably takes much longer than anticipated. A reviewer who has already worked on other documents from the same project has insight the author might not have, and will want to compare the text with those other documents as well as make it consistent in itself. Such a full review needs to be agreed up front if there is to be time to do it thoroughly. It is a job worth doing whether or not the reviewer stands to benefit personally from the harmonisation, as when later writing the clinical summaries, and the project, whatever it is, will benefit.
Language
A request to "take a look at the English" can be interpreted in many ways and, depending on whom it comes from and the type of text, can mean anything from checking that there are no howling mistakes to translating the parts the author could not, rephrasing as necessary and making a rough draft suitable for publication. The language review may be the most difficult one to gauge beforehand, unless the reviewer knows the quality of work the author is likely to produce. Formulating a single sentence satisfactorily can take a long time, and having someone else's imperfect draft to work from does not necessarily make the job easier.
Making language corrections is also a very sensitive issue. Authors do not usually take offence at having the standard operating procedure quoted at them, and many take pleasure in explaining the intricacies of their subject to the uninitiated and are grateful to a nitpicker for pointing out their inconsistencies, but hardly anyone enjoys having their prose pulled apart by someone else. Knowing when it is not appropriate to correct a text, because of the author's individual style or personal preferences or because it is simply outside the scope of the mandate, is essential. A reviewer who starts micro-editing needs to see it through to the end, and after five pages of an 80-page document may realise that they have wildly underestimated the work involved. A reviewer with only half a day for a long document should try to establish what the absolute minimum is. While reading through and checking consistency, the reviewer should get a feeling for the style, and may notice recurring mistakes in vocabulary or grammar and decide to correct them even without time for a full-text review, but it is wise to make the extent of the corrections clear to the author. Otherwise the reviewer meets the familiar complaint "… but I had it checked by a native speaker!".
Rules for honest reviewing
A reviewer who wants to avoid these pitfalls and be appreciated for honest reviewing can keep to rules such as these:
- Clarify the level of review the author expects.
- Establish whether anyone else is doing an official quality control or number check.
- Ensure that the expected feedback fits into the time available.
- Once agreed, do not exceed the mandate.
- Do not attempt to make the text sound like your own.
- If in doubt, leave well alone.
A reviewer who keeps to them may become a favourite reviewer and get more work than they actually want, but will be accused neither of skimming nor of nit-picking.
What a specialist reviewer will and will not comment on
Much time and energy can be wasted in poor review cycles, with non-specialists offering their "wisdom" on sections outside their remit, and it is in everyone's interest to make review cycles as efficient and useful as possible. Statisticians have expertise in statistics, physicians in medicine and medical writers in communication, and if each sticks to their own area of expertise and recognises the expertise of the others, research is written up more efficiently.
In the experience of one statistician who was formerly a medical writer, gratuitous nitpicking is probably what annoys medical writers most about reviewers, and besides annoying them it wastes everyone's time. He will not change a sentence that makes perfect sense, even if he would have written it differently, because deciding how to phrase things is the medical writer's job. There may have been a good reason for the phrasing that the reviewer was unaware of, such as complying with a client's style guide; the reviewer's idea of the phrasing may be odd and the writer's version clearer anyway; and even if the reviewer's way is better, no drug was ever rejected by the FDA because of how a sentence was phrased in a study report.
What this statistician does focus on is the sort of thing for which the writer wanted a statistician to review the document in the first place, such as whether the statistical methods section accurately describes the analyses that were actually done, whether the statistical results described in the results section have been properly interpreted, and whether the conclusions as written follow from the results. There can be a fine line between commenting on what is written and on how it is written, because some statistical concepts require a certain precision in the way they are expressed. He will comment on how a sentence is written where it misses a statistical subtlety, as when a writer states that treatment A was superior to treatment B when a non-inferiority analysis has shown only that A is not inferior to B, and will also comment where text is unclear or misleading.