Write a discussion that separates findings, explanations and implications

1. State the principal finding with its limits

Write an opening sentence that answers the stated question using the analysed population, outcome and comparison. Avoid repeating every table entry. If the main result is imprecise or inconclusive, say so before moving to an interesting secondary result. A discussion should not make a failed or unresolved primary question disappear.

ICMJE advises that conclusions remain supported by the data and that new hypotheses be identified as such. Apply that distinction sentence by sentence: did the study observe this, does existing evidence support it, or is it a proposed explanation to investigate?

Reference: ICMJE: manuscript preparation

2. Test claims against a fixed fictional record

Assume a fictional service evaluation reviewed appointment records before and after a scheduling change. Recorded median time from booking to first visit was three days before and two days after. There was no concurrent comparison service, the composition of attendees differed between periods, and clinical outcomes and costs were not collected. The example supplies no uncertainty estimate and no causal analysis.

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

3. Compare studies on the dimensions that matter

Choose relevant original studies and check what each actually measured. A study of time from referral to booking does not measure the same interval as booking to first visit. Explain such differences before calling results consistent or contradictory. Compare populations, settings, definitions, timing, analysis and uncertainty, not merely whether both papers reported a favourable conclusion.

Use a literature comparison sheet with a source location for every factual statement. Do not write “most studies agree” after reading only a convenient subset. Where the search was selective, describe the comparison as contextual rather than presenting it as a systematic synthesis. Keep a source pending when its full result cannot be verified instead of allowing a citation generator to supply the missing finding.

4. Explain how a limitation changes interpretation

STROBE’s explanation asks for cautious interpretation considering limitations, other evidence and the analyses performed. Use that as a prompt to explain consequences rather than listing generic weaknesses.

For the fictional service evaluation, different attendee composition could contribute to the observed difference, and a changing external service environment could also matter. The available record does not establish the size or direction of these contributions. State that uncertainty. “Single centre” is more useful when followed by what differs elsewhere—such as referral pathways or staffing—than when used as a ritual sentence with no implication.

Separate missing measurement from a negative finding. No cost data means economic benefit was not assessed; it does not mean the change had zero cost. Similarly, failure to collect a patient outcome cannot support either improvement or no improvement. The limitation is the unavailable evidence.

Reference: STROBE: explanation and elaboration

5. Match the next step to the evidence gap

Propose a next step that addresses a specific unresolved question: verify consistent time definitions, collect a patient-relevant outcome, examine changed case mix or plan an appropriate comparison. Explain which gap the proposal addresses. Do not promise that a larger sample alone will remove confounding or repair a poorly defined outcome.

Keep planned research separate from approvals, funding and implementation. A proposal can be useful before any of those are secured, but it should be described as a proposal. Avoid announcing routine clinical adoption on the basis of this fictional service record. Actual practice decisions require the responsible team’s broader evidence and context.

6. Audit the conclusion in both languages

Underline causal verbs, absolute claims, claims of novelty and statements about completed actions. Trace each to the evidence. Check whether “may explain” became “explains” during editing or “not assessed” became “no effect”. In Arabic, “كان الزمن المسجل أقصر” preserves an observed comparison; “أثبت النظام فعاليته” adds a conclusion the example cannot support.

Read the final paragraph alongside the abstract conclusion and the objective. They should describe the same result at the same level of certainty. A useful discussion can identify a local signal and a precise next question while acknowledging that causation, clinical benefit and costs remain unestablished. That is a substantive conclusion, not a weakness to hide with stronger language.

Questions for a practical final check

• The opening answers the main question without hiding uncertainty.
• Observations, explanations and proposals are distinct.
• Literature comparisons use compatible definitions and verified sources.
• Limitations explain their consequences without inventing a direction.
• Unmeasured outcomes remain unassessed.
• Arabic, English and the abstract use the same level of certainty.