A review article or book chapter can help get your name knownReview articles and book chapters
All of us rely to some extent on review articles and books to keep up to date. The scientific literature is so vast that no one could possibly read everything relevant to their subject. The problem is most acute for people working in clinical medicine, who are fully stretched just seeing their patients without trying to read everything that could possibly be relevant to them. Review articles and review chapters in books therefore play an essential role in keeping clinicians and researchers up to date, and alerting us to important shifts in opinion or practice.
A review article or book chapter can help get your name known
A good review article or book chapter not only benefits your fellow scientists – if you are a young scientist, it also benefits your career. It can bring your name to the attention of a much wider audience than would a single piece of original research, and helps to establish you as an expert in your field.
Do not underestimate the amount of work involved
Writing a review article or book chapter is very worthwhile, provided it is approached in a systematic and objective way. But it is not usually a quick or easy task. However well you know your subject, it is a mistake to think that a review article or book chapter can be ‘dashed off’ in a few days. Generally speaking, it is considerably more difficult and time-consuming than writing an original research paper.
If the review is commissioned, get a good brief
Usually, a review will be commissioned by the editor of a journal or book. If you receive a letter asking you to write a review, you may well find that an outline of the proposed content is enclosed.
If in doubt, ask plenty of questions
If you have not been thoroughly briefed by a commissioning letter or email, it is your job to ask some important questions before you start to write.
• Who is the intended audience?
• Which topics should you include?
• Which topics should you exclude?
• Should you stress a particular point of view?
• How long should the review be?
• How many figures and tables are required?
• How many references would be appropriate?
• Will there be an electronic or online version?
• When is your deadline?
Of course, the editor may say that you have a free hand – it is up to you. But often you will find that they do have a preconceived idea of what your review should cover, and in how much detail.
If other authors are involved, beware of overlap
In a multiauthor book, there is often potential for overlap between the chapters. Check with the editor who else is contributing, and whether there are likely to be any areas of overlap. If so, ask how this should be managed. The editor may lay down some boundaries, or invite you to contact the other author(s) and work with them to ‘carve up’ the subject between you. On the other hand, some editors may take the view that more than one viewpoint on the same subject will add something useful to the book.
If the review is your idea, check it out with your intended journal
Some journals only accept commissioned reviews. Others sometimes accept reviews submitted speculatively. However, it would be a big gamble to take the time and trouble required to write a good review, without knowing whether it fits in with the journal’s publication policy. If you have an idea for a review, it is a good plan to phone the editor of the journal to see whether he or she would be interested. The editor may ask you to submit a detailed outline. Of course, they are unlikely to agree definitely to publish without seeing the completed review – among the top journals, peer review is usually just as important for review articles as it is for original research papers.
Decide what kind of article you are going to write
Reviews and book chapters come in many different shapes and sizes. For example, they may be:
• comprehensive – everything there is to know about a particular topic
• systematic – analysing multiple studies using a predetermined method to answer a question
• selective – deliberately excluding particular topics or approaches
• descriptive – simply outlining what has been done
• evaluative – highlighting ‘good’ studies
• argumentative – supporting one point of view against another.
Decide where you are going to get your information
Many reviews and book chapters are based exclusively on material available in the ‘scientific literature’ – published papers, abstracts and theses. Others include unpublished data from the author’s own studies (sometimes used rather cynically as a way of publishing data that would not make a paper on its own).
Consider whether a meta-analysis is possible …
Meta-analyses are considered the ultimate supporting evidence for guidelines and decision making in evidence-based medicine and are therefore likely to be well received. However, they require a lot of work and statistical expertise.
… Or whether a less statistically rigorous approach is required
Even in reviews that do not include meta-analysis, a non-quantitative but systematic review may be required – showing that you have looked at all the published studies, excluded those that do not meet specified criteria, and objectively assessed the implications of the rest.
Search the literature carefully
For a review to be authoritative, it is important to consider all the relevant papers. To do this, you need to conduct a careful literature search. Depending on the topic, you may need to use a variety of search methods. You will probably find it helpful to consult a scientific information specialist in order to devise a robust search strategy.
Get copies of all the references and read the relevant parts
As mentioned in Chapter 19, it is your responsibility to check that each reference says what you say it says. Inadvertent misrepresentation of the contents of scientific papers is quite common, and can be perpetuated from one review article to another. Make sure that this does not apply to you. You do not usually have to read all of every paper – often just the abstract will give you a good idea of the main points. Then you can read the whole of just the most relevant papers.
Have a ‘big idea’
Every piece of writing should have a ‘big idea’. Nowhere is this more important than in a review. Sometimes your idea may be very big, e.g. ‘Everything known about the causes, diagnosis and treatment of X’. More often, however, it will be something more specific, e.g. ‘Based on the available evidence, the best way of treating X in elderly patients is the ABC method.’ If you cannot write your big idea down in a single sentence, you are not ready to start writing the article.
Draw up a detailed plan
Planning is all-important when writing long and complex documents. If you are working to a detailed outline, you will be confident that you are not missing out essential parts of your argument or story. You will also be able to draw up a timetable ensuring that you meet your deadline. Surprise your publisher by being one of the few authors who send their manuscripts in on time! To draw up your plan, try the techniques listed in Chapter 29.
Use standard structures where appropriate
Some review articles have standardised structures. For example, if you were writing a comprehensive review of a new drug, it might follow this sort of structure:
• the disease area
• previous treatments
• chemistry
• pharmacology
• pharmacokinetics
• toxicology
• animal models of disease
• dose- ranging studies
• efficacy in humans (vs placebo and vs comparators)
• adverse effects in humans
• conclusions and recommendations.
Include a statement of how you got your information
Nowadays, it is customary to give some information in your review about how you obtained the references, e.g. ‘We conducted a literature search of PubMed, Google Scholar, the Cochrane Library and ClinicalTrials.gov for studies focused on type 2 diabetes and Asian diet published up to 01 November 2015. We used the search terms …’
Think carefully about which studies you include
To paraphrase George Orwell, all studies are equal, but some are more equal than others. Usually, you will want to ascribe more weight to the ‘best’ studies. You may decide not to mention seriously flawed studies at all, or to mention them but not discuss them in detail. Just as you would describe how you selected subjects or patients for a study, you should include inclusion and exclusion criteria for papers in your review. For example, did you only evaluate blinded studies, or did you include open studies as well? Why?
Within sections and paragraphs, double-check for logical order
Readers expect to see logical order in your writing. Note that you may need to use different kinds of logical order in different parts of your review. For example:
• chronological – early studies to recent studies
• order of importance – best studies first, minor studies last
• structural – by body system or geographical area
• deductive – e.g. trans fats are bad for us, deep-fried foods are rich in trans fats, therefore we should avoid eating deep-fried food.
Structure your arguments carefully
Still on the subject of logic, remember that readers must be able to follow the development of your argument. Make sure that every statement you make is supported by the evidence. Never assume that it is obvious. You also need to make sure that readers do not lose sight of the argument – a summary sentence or paragraph every now and then is a good idea.
Prepare yourself for peer review
Even if your review is commissioned, it may still be subject to peer review. Here is a checklist of criteria used to examine review articles.
✓ Was the purpose of the review stated?
✓ How was the literature selected?
✓ How were the articles assessed?
✓ Were differences in study findings analysed and explained?
✓ Were studies grouped appropriately?
✓ Were the conclusions supported by data?
✓ Were directions for future research specified?
Write an informative abstract or summary, if the format allows
For readers, an abstract or summary that outlines the main points of the article or chapter is very useful. When a review-article title is downloaded from an online database, searchers tend to feel cheated if they do not find an abstract attached. So, assuming that the format of the journal or book allows it, provide an informative abstract or summary. Try to avoid the sort of summary that says ‘Recommendations are made for approaches to …’ This leaves readers still wondering what the recommendations are.
It is harder to write an informative summary of a long article, but very worthwhile.