Clinical Rehabilitation, the journal I have edited from 1994 to 2021, rejects about 88% of all submissions. Rejections fall into two groups. Many papers are not worth publishing or are entirely outside the scope of a rehabilitation journal. For the remainder, it is a question of relative priority. Some articles are better than others, and the editor makes a reasonably arbitrary decision. About 3% of papers submitted are important, and the main concern is to improve them before publication. This blog post is a reflection on how to reduce the risk of an editor rejecting your paper. I hope it helps you to succeed.
An hour in the life of an editor.
One theme that will run through this post is, put yourself in the position of the person reading your submission. And the first person to read your submission is an editor. I make an initial decision on a paper within one minute. I will either reject it or consider it worth reading a little more.
The first thing I will read is the title. The title immediately puts me into a certain mindset, which may well determine the fate of your paper. If the title suggests immediate rejection (after five seconds), I will scan the abstract to ensure nothing changes my mind. By 30 seconds, I will have rejected the paper, and I will dictate by reason onto a computer. I do not waste time reading the article.
In case you think this is unfair, consider:
- we receive 1,500 submissions a year, and increasing. That is four every day.
- I, like most editors, have many other jobs and can only devote limited time. It is a job done for love, not money – and I have loved it.
- I wish to spend the bulk of my available time helping authors to improve papers that should be published.
- If a paper is obviously not going to be published, I stop to allow more time for more productive editorial activities.
- research has shown that decisions made on the title and abstract are rarely changed on reading the paper.
In this hypothetical hour, I will also:
- reject two or three other papers
- review a paper that might be publishable, and give the author feedback to improve their paper
- read the reviews returned on papers sent for review
- select up to 16 reviewers for a paper worth reviewing, and invite two (sometimes 20+ people need to be approached before two agree and provide a review).
- go through papers which need me to invite another reviewer because someone has declined, or simply not responded
- accept a paper that is satisfactory and transfer details to a spreadsheet
- answer emails inquiring about the suitability of a paper for submission (100-200 each year)
- reply to people who want the decision to reject their paper reviewed (about 10% succeed)
- select papers for the next issue
Make life easy for the editor by writing a clear, understandable title that communicates the purpose and nature of your study and paper and writing a clear, informative abstract with as much data and information in it as possible, using the recommended structured headings.
Is the topic suitable?
Clinical Rehabilitation publishes papers that inform decisions and actions made by clinicians undertaking rehabilitation. Papers reporting studies on medical diagnostic procedures, animal studies, healthy people, English Literature (we have had one!), philosophy unrelated to rehabilitation, technical details of robots and so on, I reject immediately.
It is more difficult to draw boundaries around precisely what might or might not be within scope. The Cochrane Rehabilitation group is attempting to do this. (here) I think it is a futile exercise. (here) The criteria that guide me (but do not dictate decisions) are as follows:
- are the main measures relevant to patients? All measures of activities, all measures of pain or distress, some measures of satisfaction and patient-reported measures of quality of life all are relevant.
- studies with measures restricted to impairment, physiology or disease are generally of low priority.
- is the study focused on the process of rehabilitation? Studies on data-collection tools (assessments and measures), formulation, setting goals and making plans, and any intervention or strategy fall into this group.
- studies focused on inter-relationships between variables or on prognosis and natural history have a low priority.
- does the study involve people with a disability? The study must focus on a patient population
- studies involving healthy people who have not presented to a healthcare professional with a problem generally have a low priority.
- if a study is investigating prevention of disabilty it may be considered.
- could or should the study alter clinical rehabilitation practice? This is, perhaps, the fundamental question. I know that very few papers will, on their own, alter clinical practice. – or should do so. However a published paper must contribute something towards developing better rehabilitation. (See next section)
Look at published papers to see if your article is likely to be within the journal’s scope. If you are doubtful, ask. When asking, write a focused email or letter, and include an abstract.
Is the topic or question of interest?
Study design, clarity of writing and many other factors are essential when considering a paper. These factors only come into play if the paper’s subject is sufficiently interesting to be worth more thought. The largest study, well-designed, with a clearly written paper, will not be accepted if the question answered in the study is irrelevant.
I rarely say to an author that the research undertaken was, in my opinion, pointless or of little clinical relevance – that would be unfair and unhelpful. Nonetheless, I reject a significant number of papers for that reason. Authors can only avoid this rejection by considering whether the research is needed and relevant before starting. Some examples might illustrate what I mean:
- Descriptive studies on a cohort of patients seen in a stroke unit or more broadly. I published some studies like this in the early 1980s, and they were only accepted because they were larger and better documented that earlier studies. Hundreds of such studies have been published. We do not want to add one more.
- Systematic reviews that look at some sub-question, such as whether exercise after myocardial infarction is beneficial in people aged over 70 years. If it has already been esablished that the intervention is beneficial across all patients, there must be some extremely strong, evidence-based reason for look at a sub-group. Many submitted systematic reviews are answering questions that are of no importance.
- Any trial on an intervention where a systematic review has already answered the question.
- Any trial of an intervention where there are already many small trials but still no answer; the need is for a large trial, and publishing yet another small trial will not help.
This criterion might be called the “So What?” test. The authors and researchers must be able to explain how their study changes something. What would you answer if I said – sound design and data, but so what?
Design, numbers, etc
You will have noticed that methodological considerations come pretty far down the list. Coming further down the list does not imply that they are less critical. They are essential but often require considerable work to understand because the authors describe the methods so poorly. Therefore, it is more efficient to check on more easily established criteria for rejection first.
The number of patients involved in a study is a crucial determinant of acceptance. More patients general translate into a more precise estimate of the outcome studied. The magnitude of a difference between two groups, the size of a change over time, the degree of association between two variables, and the sensitivity of a diagnostic test will all be more precise if more patients are used.
Most researchers now appreciate that power calculation is required and offer one within the paper. Unfortunately, the calculation of power requires estimates of the difference expected or the degree of variation expected, and if unrealistic estimates are used, the result carries no meaning. The underlying estimates, and the source of those estimates, are rarely given.
Rehabilitation research is replete with small studies. Many randomised trials are submitted, with only 20 or 30 patients in total. Sometimes a trial has three or four arms, with 10-20 people in each group. Often many statistical comparisons or calculations are made, perhaps 20-40. Then the researcher concludes that a strengthening programme, for example, affects the left brachioradialis muscle because it was ‘statistically significant’. This is simply gross overanalysis with no sense.
Many papers discuss the problem of false-positive statistical significance associated with small numbers of patients in a study. Three easily accessible ones are here, here, and here.
Many studies also involve unrepresentative populations. Unless there are practical reasons for having an unrepresentative sample, or some persuasive reason is given, an unrepresentative sample will reduce the priority of a paper.
Almost all research’s logical and practical design is now covered by statements of the standards expected when publishing. Most of them can be found at the Equator Network website here. Failure to follow a published standard will usually give the paper a low priority. Conversely, completing and attaching the associated checklist will overcome many editorial concerns.
Other negative factors
Many other factors may not lead to immediate rejection if single, but if many of them are present, they will often tip the balance. About 70% of accepted papers are not very different from an equivalent number of papers that are rejected. Small irritating aspects of a presented paper may tip the decision towards rejection, especially at the end of a long day.
Authors who make claims and draw conclusions not justified by the design or patient numbers will reduce a paper’s priority. Failure to discuss any weaknesses will also cause an editor great concern.
Careless mistakes, such as getting the numbers wrong, giving an incorrect reference or quoting a reference incorrectly, leaving tracked changes visible and so on raise great concern. If authors cannot be bothered to check and double-check their submitted paper, they will probably have an equally slapdash and/or disrespectful approach to their research.
Submitting a poorly presented article with mistakes, overanalysing and overinterpreting results and writing poorly all lead an editor to have major worries about the standard of the underlying research. Someone who obviously is careless and slapdash about their paper, and who has no understanding of the limitations of their research is likely to be a researcher with similar characteristics.
How to avoid immediate rejection
Obviously, an author should avoid all the errors given above. The authors can use several techniques to achieve this:
- Always, always put yourself in the position of the anticipated reader – editor, reviewer, and hopefully a person who wants to read it.
- read the paper aloud to someone.
- ask a friend to read it and tell you what is not clear.
- put the paper aside for three days, and then read it.
- double-check all tables and figures.
- read and follow any guidance given by the journal.
- remember, no study or paper is perfect nor will any study give a definitive answer. Be humble.
- imagine that your life depends upon your paper’s acceptance, and you have to be in the top 20% of all submitted papers.
If you have just had a paper rejected, it is likely that one or more of the many avoidable faults identified in this blog post were present. Read your paper carefully, read any comments made, revise your paper thoroughly and try again. Sometimes, however, you may realise that the study was too flawed to be publishable. Learn from that, and start a new project! Do not give up. You can learn more from failure. Do not waste the opportunity.