CONSORT Statement & Trial Reporting

CONSORT Statement & Trial Reporting

1. Definition & scope — what CONSORT is

• CONSORT = CONsolidated Standards Of Reporting Trials — "an evidence-based, minimum set of recommendations for reporting randomised controlled trials (RCTs)" (CONSORT 2010, BMJ 2010;340:c332).
  • It is a reporting standard, not a methodological or quality-of-conduct standard. CONSORT tells authors what to write up about an RCT that was done — it does not prescribe how to design, randomise, or analyse the trial (the conduct of randomisation, allocation and blinding is the domain of trial methodology — see §10 on the distinction). (CONSORT 2010, BMJ 2010;340:c332)
  • The core deliverables are two: a 25-item checklist (CONSORT 2010) of essential elements that a trial report must contain, and a participant flow diagram that tracks subjects through the four trial stages. (CONSORT 2010, PMC2844940)
• Target design. The base CONSORT statement applies to the commonest design — the individually randomised, two-group, parallel-group RCT (the standard MD-thesis interventional design). Other designs (cluster, non-inferiority, crossover, etc.) are covered by official extensions (§7). (CONSORT 2010, PMC2844940)
• Why a reporting standard is needed. A trial that is well conducted but poorly reported is, for the reader, indistinguishable from a poorly conducted trial — the reader cannot judge validity (is the result believable?) or applicability (does it apply to my patient?) from an incomplete report. Inadequate reporting also makes a trial unusable for systematic reviews and meta-analyses, the apex of the evidence pyramid. (CONSORT 2010, BMJ 2010;340:c332)
• Empirical rationale — reporting maps to bias. CONSORT items are not arbitrary; many were selected because methodological studies link the omitted/poorly-reported item to biased (usually exaggerated) treatment-effect estimates — e.g. trials with inadequate or unclear allocation concealment and lack of blinding report, on average, larger effect sizes than trials that report these safeguards adequately. Transparent reporting is therefore a tool for bias detection by the reader. (CONSORT 2010, BMJ 2010;340:c332)
• Reproducibility. Sufficient reporting detail (especially of the intervention, item 5, and outcomes, item 6) is what allows another team to replicate the trial — a prerequisite for confirmatory evidence. The G&G regulatory standard of "adequate and well-controlled investigations" is generally interpreted as needing two replicate trials, which presupposes that the first was reported in enough detail to be repeated. (G&G 14e Ch.1, p.15)