Evidence explainer
  • Identify whether the source is a topline release or a full report.
  • Read the definition of the analysis behind the percentage.
  • Compare benefits, harms and study design together.

Start with the type of report

A topline announcement is an early summary of selected findings. It may arrive before readers can examine a full paper, protocol or detailed safety tables. Describe such results as company-reported when the company is the source; do not quietly present a press release as an independent assessment.

An observational study of drug-trial reporting found that press releases sometimes preceded journal publication by many months. The authors explained why access to methods and eligibility criteria matters. Peer review adds scrutiny, but publication is still a starting point for appraisal rather than a guarantee that every claim applies to every patient.

Ask exactly what was estimated

An estimand is the treatment effect a study sets out to estimate. The ICH E9(R1) framework asks researchers to specify the population, treatment, outcome and how events such as stopping medication or starting another treatment are handled. Two percentages from one trial may answer different questions.

A hypothetical analysis may estimate what would happen under specified conditions, such as continued treatment. A treatment-policy approach can consider outcomes regardless of a specified event, such as stopping treatment. Read the study’s definition: these labels are not interchangeable, and neither automatically provides a forecast for routine care.

Put the number back into the study

Check who joined, whether they had diabetes, which dose was studied, how long follow-up lasted and what support both groups received. The NIH describes the protocol as the plan defining eligibility, procedures, duration and research questions. Those details help explain why percentages from separate trials cannot simply be arranged into a reliable league table.

Read adverse effects and discontinuation alongside average benefit. Distinguish average change from the proportion reaching a target. Our editorial approach keeps these measures separate and flags what remains unanswered. For a UK reader, promising results must also be separated from MHRA authorisation and NHS funding: each answers a different question.

← More research & news
Follow the evidence

Sources & further reading

These links support the information above; they do not imply endorsement of this website.

  1. FDA / ICH: E9(R1) estimands and sensitivity analysis guidance
  2. Purohit and colleagues: observational study of trial press releases before publication
  3. US National Institutes of Health: clinical trial phases and protocols
  4. MHRA: authorisation and NHS access distinguished in its Wegovy tablet announcement
A note on this guide

Written for a UK audience by the Health Comparisons editorial team. This page is an information resource and has not been presented as a clinician’s personal recommendation. Discuss treatment decisions with your prescriber.

Read our editorial standards →