On this page
A press release can announce a result quickly, but it is selected and framed by the organisation issuing it. A peer-reviewed paper exposes much more of the protocol, analysis, uncertainty and safety data to scrutiny. Neither format should be accepted on reputation alone: the fastest reliable check connects the headline to the trial registry, prespecified outcomes and the fullest available results.
The seven-check evidence test
- Identify the exact study. Find the NCT or ACTRN number, sponsor, intervention, population and trial phase.
- Match the primary outcome. Compare the headline with the registry’s outcome measure and timeframe.
- Find the denominator. How many were randomised, treated, completed and analysed?
- Read the effect in absolute terms. Look for group values, difference, uncertainty and follow-up—not only a relative percentage.
- Check discontinuations and missing data. Attrition can change the apparent result.
- Inspect harms with the right denominator. Separate common adverse events, serious events and discontinuations.
- Locate the fullest evidence. Registry results, regulator review and a complete peer-reviewed report usually provide more context than a topline announcement.
Evidence formats are not interchangeable
| Format | Useful for | Main limitation |
|---|---|---|
| Sponsor press release | Fast notice of a topline result or milestone | Selective detail, promotional framing and little methodological depth |
| Conference abstract | Early numerical results and scientific discussion | Short format; methods and harms may be incomplete |
| Registry results | Structured outcomes, participant flow and adverse events | Submitted by the responsible party; not equivalent to journal peer review |
| Preprint | A fuller report available before journal review | Not yet peer reviewed and may change |
| Peer-reviewed paper | Methods, analyses, tables, limitations and editorial scrutiny | Peer review is quality control, not a guarantee; selective reporting can remain |
| Regulator assessment | Integrated benefit–risk and product-quality evaluation for a defined application | May appear later and can be highly technical |
1. Connect the headline to a registered trial
A credible clinical-trial announcement should make the exact study findable. Search the registry identifier. Confirm the intervention, comparison, participant group and primary completion date. Similar study names can cover different doses, indications or populations.
Trial registration matters because it creates a dated record of planned primary and secondary outcomes. The ICMJE requires prospective public registration as a condition of publication in its member journals, although registry compliance does not by itself establish that a study was well designed or honestly interpreted.
2. Ask what “met the primary endpoint” actually means
The phrase usually means the prespecified statistical criterion was reached. It does not tell you whether the effect was large, clinically meaningful, durable or balanced by harms. Find the group estimates, between-group difference, confidence interval and exact assessment time.
| Headline phrase | Missing information to seek |
|---|---|
| “Significant improvement” | Effect size, confidence interval, scale and clinical importance |
| “Reduced risk by 40%” | Absolute risk in each group and number of events |
| “Well tolerated” | Adverse-event rates, serious events, withdrawals and exposure time |
| “No new safety signals” | What was monitored, for how long and in how many participants |
| “Consistent across subgroups” | Prespecification, interaction tests, sample sizes and multiplicity |
3. Trace every percentage back to people and time
Relative effects can look dramatic when baseline risk is low. If an event falls from 2 in 100 people to 1 in 100, that is a 50% relative reduction and a 1 percentage-point absolute reduction. Both are correct; only together show scale.
For continuous outcomes such as body weight or a biomarker, determine whether the release reports a mean, median, model-estimated change, responder threshold or “up to” result from a selected subgroup. Check the analysis population and whether values assume participants remained on treatment.
4. Look for estimands, missing data and discontinuation
Modern trials may answer different questions about events after randomisation. One analysis can estimate the effect regardless of treatment discontinuation; another can estimate what happened while participants remained on therapy. Both may be useful, but they are not the same claim.
High or unequal discontinuation creates uncertainty. A full paper should explain missing-data assumptions and sensitivity analyses. A release that reports only the most favourable analysis deserves a cautious label such as “topline sponsor result”.
5. Read safety separately from efficacy
Safety cannot be compressed into “generally well tolerated”. Ask how many people received each intervention, how long they were exposed, which adverse events were common, which were serious, and how many stopped treatment because of an event. Small or short studies are poorly equipped to detect uncommon or delayed harms.
6. Treat subgroup claims as new questions
Subgroups create many chances for an eye-catching pattern. Stronger claims require prior specification, adequate sample size, a formal test showing effects differ between subgroups, and ideally replication. A result being significant in one subgroup but not another does not prove the groups differ.
7. Update the verdict as fuller evidence appears
A responsible evidence page can change its label over time:
| Available evidence | Appropriate wording |
|---|---|
| Press release only | “The sponsor reported…” and “full methods and results are not yet available.” |
| Conference abstract | Describe it as preliminary and note the limited report. |
| Registry results | Report structured outcomes while noting the source and absence of journal review if applicable. |
| Full peer-reviewed paper | Appraise the methods, effect, uncertainty, harms and limitations. |
| Regulator decision | State the exact product, indication, conditions and decision date. |
Red flags worth slowing down for
- No registry identifier or enough detail to locate the study.
- A secondary, exploratory or subgroup result presented as the main finding.
- Relative percentages without absolute group results.
- No participant flow, discontinuation or missing-data information.
- Safety summarised with adjectives but no counts or exposure time.
- Changes to primary outcomes after recruitment or data availability without explanation.
- A claim about approval that does not name the regulator, product and indication.
Frequently asked questions
Is a press release evidence?
It is evidence that an organisation made a claim, and it may contain genuine topline data. It is not a substitute for a complete report that allows methods and results to be checked.
Does peer review prove a result is true?
No. Peer review can improve reporting and identify problems, but readers should still examine design, prespecification, analysis, uncertainty, conflicts and reproducibility.
What does “topline result” mean?
It usually means a summary of key outcomes prepared before the full dataset and methods are publicly reported. Treat it as provisional.
Can a registry have results before a paper?
Yes. Structured results may be posted to ClinicalTrials.gov without, or before, a journal publication.
What is the best single check?
Match the headline to the prospectively registered primary outcome, then compare the complete effect estimate and safety data with that plan.
Primary standards and tools
- ICMJE: clinical-trial registration policy
- CONSORT and SPIRIT reporting standards
- ClinicalTrials.gov: how to read a study record
- ClinicalTrials.gov: study-data structure
- NHMRC: open science and research access
Reviewed 30 September 2026. General research education only; not medical advice.
How to cite this article
Press Release vs Peer-Reviewed Paper: A Six-Question Evidence Check. Elite Peps Australia. Published 14 September 2026; updated 30 September 2026. https://elitepepsaustralia.com/press-release-vs-peer-reviewed-paper/
Source checks: Consult the primary studies, trial records and regulator documents linked in this article for current evidence and status. Publication and update dates do not by themselves indicate an evidence review.