Expand a conference abstract by rebuilding its evidence trail

1. Recover what was actually presented

Save the final submitted abstract, any published version, the poster or slides and the record identifying the conference. Note differences among them. An author’s earlier draft may contain a different denominator from the published abstract. Record the difference and investigate it rather than choosing whichever number fits the new narrative.

BMJ’s guidance asks authors to acknowledge previous conference presentation and cite published abstracts. Use the target journal’s current policy to determine how the prior item should be identified in the new submission.

Create a short chronology: when the abstract was submitted, when the relevant data were extracted, what was shown at the meeting and what changed afterwards. Keep a later analysis update visible. More complete follow-up may justify a changed result, but the explanation and actual source output are needed; a later date alone cannot explain it.

Reference: BMJ Author Hub: formatting and prior conference presentation

2. Map each manuscript section to evidence

Start the full report with an evidence map rather than a target word count. For each section, identify the source and the person able to check it. A protocol describes planned methods; the execution record describes what happened. Where they differ, retain both accounts and explain the deviation through the applicable reporting process.

PLOS One’s instructions require enough methodological detail to understand and replicate the work. This is a concrete journal requirement, not a claim that a long Methods section establishes reproducibility.

The map below is a proposed working aid. It is not a complete reporting guideline. Select the guideline appropriate to the actual design and destination, then identify where its relevant items are addressed. Mark an item unavailable when the team has not found its evidence, even if a convincing sentence could be written about it.

Reference: PLOS One: submission and methods requirements

On small screens, scroll the table horizontally, or focus it and use the arrow keys.

3. Reconcile a fictional abstract result

Assume a fictional cross-sectional questionnaire invited 200 adults. There are 150 valid responses to the item about travel barriers, and 60 of those respondents report a barrier. The item result is 60/150, or 40%. The remaining 50 invited adults have no valid response for this item. Reporting 60/200 as a 30% prevalence would silently treat those missing responses as if their status were known.

An earlier abstract merely said “40% reported barriers” without naming the denominator. The full report should supply the observed count and denominator, define the item and explain missing responses. The fictional record does not provide recruitment dates, response reasons, an uncertainty interval or a representative sampling claim. Those details remain to be supplied or assessed; none may be invented for completeness.

4. Draft sections in an order that exposes gaps

A practical sequence is to assemble the participant flow and principal outputs, describe the methods that produced them, then write the interpretation and background. This is a proposed work order, not an experimentally tested productivity method. Its purpose is to make the claims depend on identifiable material while uncertainty is still visible.

Avoid adding analyses merely to fill pages. A new subgroup comparison needs a scientific rationale, suitable data and an actual checked execution. Discuss whether it was prespecified or developed later. If the study cannot support the proposed analysis, explain the limitation rather than imply that the result was negative. “Not assessed” and “no difference found” describe different evidence.

Use a running question list for the contributors: which definition was applied, which exclusions were made and which version produced the table? Assign each question to someone who can inspect the source. A citation to a method paper does not establish that the team followed it.

5. Explain how the manuscript relates to the abstract

Prepare a plain description of the shared question, sample, outcomes and new detail. If the full manuscript includes additional follow-up or a changed analysis, identify what changed and why. Supply the previous abstract and other related reports where the journal requests them. Do not rely on an editor recognising a shortened title.

Check the actual status of any proceedings document. A conference may have published only abstracts, or it may have accepted a complete paper under separate terms. These are different circumstances. If the earlier abstract contains an error, ask the responsible publication office about its correction route; silently changing the full manuscript leaves the old public record unexplained.

6. Define a real completion check

Before calling the manuscript ready to submit, have the relevant contributors check the methods and results against their source versions. Obtain the required final-author confirmations and resolve outstanding policy questions. A polished Word file remains a draft if its analysis or approval record is incomplete.

Read the final abstract against the final main text again. Check every repeated count, unit, design label and conclusion, including the Arabic if provided. Record which version passed this comparison. The next milestone is a complete submission received by the journal; editorial review and acceptance remain later decisions. The team can use the gap map to explain exactly what is complete without overstating that milestone.

Questions for a practical final check

• The actual conference version is preserved.
• Every major claim has an identified source.
• Counts and missing observations reconcile.
• Unperformed analyses remain unperformed.
• The relationship to prior material is disclosed.
• Final files and contributor confirmations match.