Most case reports that get rejected are not rejected because of the case. They are rejected because the manuscript never answers the one question an editor asks on opening it: what does this teach someone who was not in the room? A case that is interesting to you is not yet a publishable case, and the distance between those two things is almost everything that can be learned.
This guide covers the four things that decide whether it goes through: which case is worth writing, how it is structured under the CARE guidelines, what consent journals require — where most manuscripts die — and where to submit it.
Which case report is worth writing
A publishable case has to add something that is not already in the literature. In practice that comes down to five situations:
An unexpected presentation of something known. The disorder is textbook, but it showed up in a way that is not described, or in a population where it is not described.
An association not previously described. Two conditions appearing together that the literature had not connected.
An adverse or unexpected effect of a treatment. This is the category with the most real value: pharmacovigilance and the safety record of psychological interventions are built in large part on published cases.
A treatment that worked where the standard protocol had failed, with enough detail for someone to replicate it.
A diagnostic error with a clear lesson. They are hard to write and among the most useful there are.
What does not work: "a typical case of X, well managed". If the management followed the protocol and the outcome was the expected one, the case teaches nothing that is not already in the clinical guideline. Neither does "the first case described in my country" when hundreds are described elsewhere: geographical novelty on its own is not novelty.
Before writing anything, run the search. If PubMed already holds twenty similar cases, your manuscript has to state explicitly what it adds to those twenty, and that sentence goes in the introduction.
The CARE guidelines: the skeleton journals expect
Since 2013 there has been a reporting guideline specifically for case reports, the CARE guidelines (Gagnier et al., 2013), and it sits in the EQUATOR network alongside CONSORT and PRISMA. More and more journals require it and ask for the completed checklist as a supplementary file.
The structure it asks for is this:
Title. It has to carry the words "case report" and the main phenomenon. No teaser titles.
Keywords. Two to five, with the diagnosis and the intervention.
Structured abstract. Background, case presentation, conclusions. Usually without references and capped at 250 words.
Introduction. Short — two or three paragraphs — with a single job: to place the case in the literature and say which gap it fills.
Patient information. De-identified demographics, main symptoms, relevant medical, family and psychosocial history, and previous interventions with their outcomes.
Clinical findings. Examination, scales administered with their scores.
Timeline. This is the most distinctive part of CARE and the most frequently omitted: a figure or table with the dates of the case milestones. A case report without a timeline reads badly and many editors send it back on that alone.
Diagnostic assessment. Methods, results, difficulties reaching the diagnosis, and differential reasoning.
Therapeutic intervention. Type, dose or format, duration, changes and why.
Follow-up and outcomes. Using the same measures you used at baseline.
Discussion. Strengths, limitations, relevant literature and the lesson the reader takes away.
Patient perspective. A first-person paragraph, where possible. CARE asks for it, almost nobody includes it, and it adds a great deal.
Informed consent. An explicit statement that it was obtained.
Consent: where half the manuscripts die
This is not paperwork, and it is worth saying plainly: if you do not have written informed consent from the patient to publish, you do not have a case report. No serious journal will accept it, and the ones that accept it without asking are precisely the ones you do not want.
Three things to know before you start writing:
Consent to treatment does not count. You need consent specific to publication, stating that clinical data will be published and that the patient has read the manuscript or a faithful summary of it.
De-identification does not replace consent. This is a common myth. Removing the name and changing the age does not exempt you from asking permission: in a rare case, the combination of diagnosis, age, occupation and center can identify the person to anyone who knows them.
If the patient has died or lacks capacity, you need consent from the legal representative or the family, and in some countries ethics committee approval. Check this before, not after.
On de-identification: remove the name, the record number, exact dates (keep months and years), the specific center, and any professional or family detail that is identifiable. If the case includes images, mask them where needed and say so in the caption. And do not alter clinical data to throw readers off: falsifying age or sex changes the case, and that is not de-identification any more, it is invention.
If your manuscript is going through an ethics committee, the ethics committee simulator anticipates the objections you will get and the paragraphs missing from the application.
How the case presentation is written
The most common writing mistake is telling the case like a medical record: chronologically, everything in, no hierarchy. A medical record is written for the next clinician who will see that patient. A published case is written for someone who will never see them and who needs to understand one specific point.
The practical consequence is that you have to select. Anything that does not support the lesson of the case is surplus, even if it was in the record. If the case is about an adverse drug effect, the detail of the family history only goes in if it bears on that effect.
Two more rules that work:
Scale scores go with the full instrument name, its version and its range the first time. "She scored 32 on the BDI-II (range 0-63, cut-off 29 for severe depression)" is readable; "she scored high on depression" is not.
The outcome is told with the same measures as the baseline. If you measured with the BDI-II at the start, the follow-up goes in BDI-II. Switching instrument halfway through makes the change impossible to judge, and it is a guaranteed objection.
Where to submit it
Many high-impact general journals have stopped publishing case reports because they are not cited and drag the impact factor down. That leaves three routes:
Case-specific journals. There are several indexed and serious ones in medicine and clinical psychology. They usually charge an APC and it is worth checking first.
The case section of a specialty journal. The best option when the case matters to a specific subdiscipline: it gets read more and cited more.
The journal of the relevant scientific society. Less global visibility, more readership among exactly the people you care about.
And a warning worth giving without hedging: case reports are the favorite bait of predatory journals. They know that whoever publishes a case is often a resident or a young clinician, in a hurry to add publications and short on editorial experience. If an unsolicited email offers to publish your case in two weeks, that is what it is. Before submitting anything, check the journal in DOAJ, in Scopus and on your specialty society's list. If you want to see what a desk rejection looks like from the other side, why Q1 journals reject you covers the editor's filter in detail.
Why case reports get desk-rejected
Desk rejection in case reports has a very repetitive set of causes:
It is not clear what it adds. If the introduction does not state explicitly which gap this case fills, the editor will not infer it.
The consent statement is missing. Automatic rejection at most journals.
There is no timeline and the case is hard to follow.
The discussion is a literature review. Three pages on the condition in general and two sentences on the case. It is the other way round: the discussion exists to interpret this case.
Causal conclusions drawn from one case. "This case demonstrates that X causes Y" does not hold with n = 1. A case generates hypotheses; it does not confirm them. Write it that way and you save yourself the objection.
The format is not the journal's. Each one has its own word, reference and figure limits, and case reports usually have strict ones.
If your case is in fact a single-case design with repeated measures — baseline, intervention, withdrawal — then you are not writing a case report but an N=1 study, and it has its own analysis and reporting rules. That is covered in meta-analysis of single-case designs.
Before you submit
A case report is short, and that is deceptive: there is less room to hide a flaw. The missing consent statement, the absent timeline, the causal conclusion that does not hold — these are objections visible on one reading, and they cost a professional review round.
If you want to see them before the editor does, upload the manuscript to the AI paper reviewer. It is free, needs no sign-up, and hands back the critiques Reviewer 2 would most likely make, pinned to the sentence that triggers them.
References
Gagnier, J. J., Kienle, G., Altman, D. G., Moher, D., Sox, H., & Riley, D. (2013). The CARE guidelines: Consensus-based clinical case reporting guideline development. BMJ Case Reports, 2013, bcr2013201554.
Riley, D. S., Barber, M. S., Kienle, G. S., Aronson, J. K., von Schoen-Angerer, T., Tugwell, P., Kiene, H., Helfand, M., Altman, D. G., Sox, H., Werthmann, P. G., Moher, D., Rison, R. A., Shamseer, L., Koch, C. A., Sun, G. H., Hanaway, P., Sudak, N. L., Kaszkin-Bettag, M., … Gagnier, J. J. (2017). CARE guidelines for case reports: Explanation and elaboration document. Journal of Clinical Epidemiology, 89, 218-235.