ISSN-registered · Peer-reviewed · Open Access
JournalsAboutContact
Research Integrity

"The First Reported Case": How to Search Properly, and What to Write If You Cannot Prove It

DE By Directive Editorial Team, Directive Publications ·9 Sep 2026 ·6 min read
"The First Reported Case": How to Search Properly, and What to Write If You Cannot Prove It

Your case report is finished. The title reads "Spontaneous X of the Y: The First Reported Case", and it is the strongest sentence in the manuscript — which is also the only sentence a stranger with a browser can disprove in four minutes.

"First" is a claim about the entire published literature, in every language. You have made it on one PubMed search in English, run once, on a Tuesday.

The fix is not to soften the wording. It is to change what the sentence claims: not the state of the world, which you cannot verify, but the search you ran, which you can.

Why "the first reported case" is the one sentence in your manuscript a stranger can disprove

Everything else in the report is anchored to your patient, which a reader cannot look up and check. Novelty is the exception. It lives outside your data, in a literature anyone can query — a reader finds a 1987 report in a Japanese journal and has a citable objection, sitting in your title.

Claim your search, not the world: the reframe that turns novelty into something checkable

Compare two sentences about the same case.

This is the first reported case of [condition] in [population].

To our knowledge, and following a search of [databases] on [date] using the strings listed below, no previous report of [condition] in [population] has been published in the literature we were able to search.

The first is a claim about all publishing, everywhere; the second is a claim about a search — bounded, dated, and true if you ran it. A reader who finds an earlier report has not caught an error; they have extended a search whose limits you declared.

Run the search with free tools first: PubMed, Europe PMC, Google Scholar, OpenAlex and regional indexes

You do not need a subscription to run a defensible search. Record what each returns:

  • PubMed — search MeSH terms and free text separately; they return different sets.
  • Europe PMC — a wider net, including preprints, and it searches the full text of the articles it holds in full, not just their abstracts.
  • Google Scholar — poor precision, good recall; it reaches theses and unindexed journals.
  • OpenAlex — an open index covering far more journals than the subscription databases.
  • Regional indexes — LILACS, SciELO, African Journals Online, KoreaMed, and WHO's Global Index Medicus, which searches its five regional indexes in one query. A claim never tested outside English is a weak claim.
  • Citation chasing — read the reference lists of the closest reports, then what has cited them since; this finds cases keyword searching misses.

Search synonyms, eponyms and obsolete names: the 1980s literature is filed under the 1980s vocabulary, and choosing keywords and titles works in reverse here.

No database is complete — including for journals like ours that PubMed does not index

Every search protocol has a hole in it, and naming the hole is what makes the rest credible.

A PubMed search returns PubMed's index, and journals it does not index will not appear. That includes our own titles: Directive journals are not indexed in PubMed, PubMed Central, MEDLINE, Scopus or Web of Science. A report published with us surfaces through Crossref DOI metadata, OpenAlex, Google Scholar and plain web search — which is why those belong in your protocol.

Non-English journals, undigitised material and society journals outside the aggregators sit in similar blind spots. You cannot close them all; you can name them.

Write the four-line search statement: databases, exact strings, date run, language limits

Put this in the Discussion, immediately before the table it refers to. Fill every bracket.

We searched [list every database and index, by name] on [date, as DD Month YYYY] using the search strings [reproduce the exact strings, including Boolean operators and field tags], without language restriction [or: restricted to publications in English, which we note as a limitation of this review]. We did not search [name the subscription databases you had no access to — for example Embase, Scopus, Web of Science]. We identified [n] previously reported cases meeting [state your inclusion criterion in one clause], summarised in Table [X].

Four elements do the work: which databases, the exact strings, the date, and what you could not search. Without the date it is not reproducible.

Build the table of previously reported cases, and make each column earn its place

The table turns a novelty claim into an argument. Seven columns is the full version:

Author, yearCountryAge/sexPresentationInterventionOutcomeFollow-up
[Author] et al, [year][country][nn] / [M or F][presenting features, one clause][procedure or treatment][resolved / recurred / died][duration]
Illustrative row only — Doe et al, 2019Illustrative47 / FPainless swelling, six months, no systemic featuresWide local excisionResolved, no recurrence18 months
Present case[country][nn] / [M or F][your presentation][your intervention][your outcome][your follow-up]

The illustrative row is a fictional placeholder, not a real patient or a real published case. Delete it, keep your own case in the final row, and drop any column where every cell reads "not stated".

Fit the table inside 2,000 words, four display items and twenty references

Case reports are short. Our Author Guidelines set the limits for a case report — up to 2,000 words, up to four figures or tables and up to 20 references, with tables excluded from the word count but counted against the four display items — so a seven-column table of fifteen prior cases will not fit alongside your imaging. Use the trimmed five-column variant:

Author, yearAge/sexPresentationInterventionOutcome
[Author] et al, [year][nn] / [M or F][presenting features][procedure or treatment][resolved / recurred / died]
Present case[nn] / [M or F][your presentation][your intervention][your outcome]

Where the seven-column version adds something, place it as a supplementary file and cite it as "Supplementary Table S1"; see preparing supplementary materials for how to label it.

If the search will not support "first", climb down the ladder: four fallback claims ranked by the evidence each needs

Every rung below is a statement about the scope and date of your search, never a synonym for "first". A claim that cannot be phrased that way does not belong in the manuscript.

RungClaimWhat your search must support
1No previously reported case anywhereMulti-database, multi-language, citation-chased, with the gaps named
2No previous report in this population, age group or anatomical variantThe same search narrowed by one stated attribute — far easier to defend
3No previous report of this intervention or this outcomeCases of the condition exist; you tabulate them and show what was not tried
4A rare presentation adding to a small published seriesA countable set of prior cases, tabulated, with your case as the next row
5No novelty claim at all — teaching value onlyNothing beyond an accurate account of what happened

A model sentence for each rung:

1. Following a search of [databases] on [date] without language restriction, we identified no previous report of [condition].

2. [n] cases of [condition] have been reported (Table [X]). Following the search described above, we identified none in [patients under 18 / this anatomical site / this population].

3. [n] previously reported cases were identified, of which [n] were managed by [alternative]. We identified no previous report of management by [your intervention].

4. [Condition] is uncommon, with [n] cases identified by the search described above. We report a further case presenting with [feature not described in the tabulated reports].

5. We report a case of [condition] in which [the diagnostic or management problem] delayed recognition, and describe how it was resolved.

A title that cannot support rung 1 belongs on rung 2: a smaller claim, defensible on the search you have already run.

A case with no defensible novelty claim can still be worth publishing on teaching value alone

A case is worth writing up when it changes what a clinician would do: a missed diagnosis and the feature that would have caught it; an adverse effect not on the differential; a decision taken under uncertainty, with the reasoning explicit and the outcome reported honestly.

Strip the novelty claim out. If the case still teaches something, build the Discussion around the lesson rather than the search.

Our standards for honest reporting are set out in Research Integrity, and the scope of each format in Article Types. See also our guides to writing a case report, writing a discussion section and preparing supplementary materials.

Frequently Asked Questions

Do I need Scopus, Embase or Web of Science access before I can say a case is unreported?
No. A search built from PubMed, Europe PMC, Google Scholar, OpenAlex and the relevant regional indexes, extended by citation chasing, is defensible on its own. What matters is that you say which databases you searched, on what date, with which strings, and which subscription databases you could not reach. Naming the gap is what makes the rest of the statement credible; pretending there is no gap is what makes it fragile.
Where in the manuscript do the search statement and the comparison table actually go?
Both belong in the Discussion, not the Methods, because a case report has no methods section in the usual sense. Put the four-line search statement immediately before the comparison table it refers to, and number the table as one of your display items so it is cited in the text as Table 1 or Table 2. If you use the longer seven-column version, cite it as a supplementary table instead and keep the trimmed version in the manuscript.
My search found thirty similar cases. Do all of them belong in the table?
Thirty prior reports is itself the answer to your novelty question, so the table stops being a novelty argument and becomes a summary of what is known. Set an explicit inclusion criterion — the same anatomical site, the same age group, the same intervention — state it in the search statement, and tabulate only the cases that meet it. Thirty references would also exceed the twenty allowed in a case report, so a summarised supplementary table with the closest handful in the manuscript is usually the workable form.
If my case is not the first of anything, is it still worth writing up?
Yes. A case that changes what a clinician would do has value whether or not anyone has reported it before: a diagnosis that was missed and the feature that would have caught it, an unexpected adverse effect, or a decision taken under uncertainty with the reasoning made explicit. Drop the novelty claim entirely rather than stretching it, and build the discussion around the lesson. A report with no novelty claim and a clear teaching point is a stronger paper than one with a contested first in the title.
DE
Directive Editorial Team
Directive Publications

The editorial team at Directive Publications — an international open-access publisher of peer-reviewed medical and scientific journals.

Publishing your research?

Directive Publications is an open-access publisher — every article peer-reviewed, Crossref-registered, and free to read under CC BY 4.0.

Submit a manuscript →Read our Research Integrity policy →