Report the design before procedural detail
Name the study design, setting, dates, population, sampling process, inclusion criteria, and unit of analysis. Then explain measures, interventions or exposures, outcomes, procedures, exclusions, and analytic choices. The order should follow the logic needed to interpret the estimates, not merely the chronology of project administration.
Define who made subjective judgments, whether assessors were masked, and how disagreements were resolved. Passive voice is acceptable for routine procedures, but it should not hide responsibility for decisions.
Connect each choice to its purpose
We used [method] to estimate [target quantity] because [design-relevant reason], with [assumption or limitation] considered in interpretation.
Observations were excluded according to the prespecified criterion [criterion]; [number or proportion] were removed before [analysis stage].
State whether decisions were prespecified or data-informed. Report software and version when it affects reproducibility. Link supplementary detail without omitting information needed to understand the main analysis.
Audit methods against results
Every population count, model, subgroup, sensitivity analysis, and outcome reported later should have a methodological basis. Every method described should either produce a reported result or have a clear role. Use the results guide to check this correspondence.