How to write a medical case report (CARE guidelines)

What is a case report, and what makes one publishable?

A medical case report describes the presentation, diagnosis, treatment and outcome of one patient (or a small number of patients) in enough detail that another clinician can learn from it. Case reports are often the first signal of a new disease, an unexpected adverse drug effect or an unusual response to treatment, which is why journals still publish them despite their place at the bottom of the evidence hierarchy.

Editors usually ask one question first: what does this case teach that the reader did not already know? A case tends to be publishable when it offers at least one of the following:

• An unexpected association between a disease and a symptom, finding or exposure
• An adverse or beneficial effect of a drug, device or procedure that is not well described
• A rare presentation of a common disease, or a common presentation that was misdiagnosed and holds a lesson
• A diagnostic or management challenge solved in a way others can reuse
• A new or modified technique, with honest limits

"First reported case" claims need a documented literature search (databases, dates, search terms). Reviewers often test this claim themselves, and a thin search undermines the whole paper.

What is the difference between a case study and a case report? In clinical publishing the terms are often used interchangeably. "Case report" is the usual term for a journal article about one patient; "case study" is broader and is also used in education, business and social-science research. If you are writing for a medical journal, use the journal's own article-type name.

What are the CARE guidelines?

CARE (CAse REport) is the reporting guideline for case reports. It was developed through a consensus process and published in 2013 in several journals at once, including Global Advances in Health and Medicine, Journal of Medical Case Reports and BMJ Case Reports (Gagnier et al., 2013). An explanation and elaboration paper followed in the Journal of Clinical Epidemiology in 2017 (Riley et al.). The CARE checklist (2013 version) has 13 items and is listed by the EQUATOR Network as the guideline for case reports. If you are not sure CARE is the right guideline for your paper (for example, a series of consecutive patients may be better reported as an observational study), use our guide on which reporting guideline your study needs.

Some case-report journals ask you to upload a completed CARE checklist with page numbers. Even when they do not, writing to CARE makes the reviewer's job easier.

What is the case report format? (CARE section by section)

The table below summarises each CARE item in our own words. Download the official checklist from care-statement.org for the exact wording.

How do I write each section of a case report?

Write the sections in this order. It is not the order they appear in the paper, but it stops the introduction and abstract from promising more than the case delivers.

1. Build the timeline first (item 7). List every relevant date and event from the medical record: onset of symptoms, presentations, tests, diagnoses, treatments, changes and follow-up visits. Use relative time (day 0, week 6, month 12) rather than calendar dates, which also helps de-identification. Every later section must agree with this table.
2. Patient information and clinical findings (items 5 and 6). Give only what a reader needs to interpret the case: age range or age, sex, relevant history and exposures. Leave out occupation, ethnicity, location or family details unless they are clinically essential.
3. Diagnostic assessment (item 8). Report the tests with units and reference ranges, explain what made the diagnosis difficult, and name the differential diagnoses you considered and how you excluded them. This is the section reviewers read most closely.
4. Interventions and follow-up (items 9 and 10). Give drug names, doses, routes and durations, every change and why it was made, how adherence and tolerability were assessed, and any adverse events. State how long the follow-up lasted.
5. Discussion (item 11). Compare your case with previously published cases, explain what differs and why it matters, state the limits plainly (one patient, no control, possible alternative explanations), and end with one paragraph of take-away lessons with no references.
6. Introduction (item 4). Two or three short paragraphs: the condition, what is known, and the gap this case fills.
7. Abstract, title and keywords (items 1 to 3). Write these last, so they match what the case actually shows. Check whether the target journal wants a structured or unstructured abstract and its word limit.
8. Patient perspective and consent (items 12 and 13). Ask the patient for a few sentences in their own words when you seek consent. Some journals have a dedicated section for it.

Format references in the journal's style; most case-report journals use a numbered style such as Vancouver referencing. Verify every reference against PubMed or the DOI before you submit, since misattributed or non-existent references are an easy way to lose a reviewer's trust.

Worked example: a CARE timeline table (synthetic case)

The case below is synthetic, invented for illustration only. It does not describe a real patient.

A table like this answers item 7, gives reviewers a single source of truth, and makes inconsistencies obvious (for example, a discussion that says "fever resolved within 24 hours" when the table says 48). CliniPaper checks dates, doses and durations across the abstract, timeline and text automatically and flags any that disagree.

Do I need patient permission to write a case report?

In almost all cases, yes. Assume you need written informed consent for publication from the patient, or from a legal guardian or proxy if the patient is a child, lacks capacity or has died. The main sources agree:

• ICMJE: patients have a right to privacy that should not be breached without informed consent. Identifying details (names, initials, hospital numbers, photographs, pedigrees) should be published only when they are essential for scientific purposes and the patient has given written consent. The patient should be shown the manuscript before publication, and told whether it may appear online as well as in print. Masking the eyes in a photograph is not adequate anonymisation.
• COPE (guidance for journals, 2016): a consent form should be required for any case report in which an individual or a group can be identified, and this also applies to deceased patients. Case histories, photographs, x-rays and genetic pedigrees all count as potentially identifying. COPE advises journals not to hold signed forms themselves; instead, authors attest that the form was signed and that the original is kept by the treating institution.
• Journals: most case-report journals provide their own consent form or require one that includes all the elements of theirs. Use the target journal's form if it has one, since a generic hospital consent for treatment is not consent to publish.
• CARE item 13 asks you to state that informed consent was obtained and to provide it if requested.

Ethics committee or IRB approval is a separate question. Policies on whether a single case report needs approval differ by institution and country, and COPE notes that it may be appropriate for institutions to require approval before a case is published. Check with your institution before you start writing, not after acceptance.

How do I de-identify a case report?

Consent does not replace de-identification. Do both.

1. Remove names, initials, record numbers, exact dates, the hospital name, and the town or region unless it is clinically relevant.
2. Use relative time (day 0, month 3) instead of calendar dates.
3. Give age, or an age range if exact age adds nothing.
4. Crop images to the region of interest and remove burned-in text and labels from imaging files. Do not rely on a black bar over the eyes.
5. Ask whether a colleague, a relative or the patient's neighbours could recognise the person from the combination of details. Rare diseases in small communities are identifiable even without a name.
6. If you change any detail to protect anonymity, tell the editor and confirm the change does not alter the scientific meaning, as ICMJE asks.

Never upload identifiable patient records to any tool, including AI tools. Work from de-identified notes.

Where should I submit a case report?

Start with the journals your specialty reads. Many general and specialty journals accept case reports or similar formats (clinical images, "lessons from practice"). There are also dedicated case-report journals indexed in PubMed, for example:

Access models, fees and requirements (such as subscription or membership to submit, or article processing charges) differ and change over time. Check each journal's current author instructions before you start formatting, and confirm it is indexed in the databases your institution counts. When you submit, a short cover letter for your journal submission should state in one or two sentences what is new about the case.

Common mistakes reviewers flag

• No clear teaching point. The discussion summarises the case again instead of stating what readers should do differently.
• Unsupported "first case" or "rare" claims, with no search strategy or a search of a single database.
• Missing differential diagnosis, or no explanation of how alternatives were excluded.
• Inconsistent numbers between abstract, timeline, text and figure legends (doses, days, laboratory values).
• Identifiable details or images left in, or consent described vaguely ("the patient agreed").
• Overreaching conclusions: claiming that a treatment works on the basis of one patient.
• No follow-up, or follow-up too short to support the outcome claimed.
• Undisclosed AI assistance. If you used generative AI to draft or edit, disclose it per the journal's policy; see our guide to the AI disclosure statement for medical journals. The authors remain responsible for every word and number.