A university announces a quit-smoking finding: read beyond the press release

01

Make sure the link belongs to this announcement

Open the institution's original page, rather than a syndicated copy. Note its issue date and follow its paper link. Match the paper's authors, subject and study setting; a journal name alone is insufficient. The release date is not the period when participants were studied. Check the paper record for a later correction too.

If the release links only to a conference summary, preprint or inaccessible article, describe what you actually found. Do not silently upgrade it to a full journal report. You can ask the press office for the public citation or publicly available manuscript without sending your smoking history.

02

One plausible headline, several missing distinctions

The following is entirely fictional, with no real institution, paper or DOI. A release says ‘More people quit with text-message support.’ Its linked report compares 250 adults offered messages plus usual support with 250 offered the same usual support alone. At a six-month visit, 50 and 40 respectively reported no smoking during the previous seven days.

Those counts are 20% and 16%: a four-percentage-point difference in that reported outcome. They do not mean 50 people had continuously stopped for six months. Nor is a group percentage your personal chance. The exercise has not supplied allocation methods, uncertainty or how missing follow-up answers were handled; those are questions to resolve, not blanks to fill with confidence.

Suppose the release also quotes a researcher saying the approach ‘could be inexpensive to scale’. Unless the paper evaluated costs and delivery at scale, that sentence is a proposal about future implementation, not a measured cost result. Preserve the word ‘could’ and identify it as a quotation.

03

Return to the result, not just the abstract's final sentence

Find where the outcome is defined, the results table with both groups, the participant flow and the limitations. Check whether the headline uses the main outcome or one of several secondary analyses. A clinic recruiting phone-owning volunteers is not automatically representative of everyone considering quitting.

If the confidence interval or missing-data explanation is not in the material you can access, stop short of a firm effect claim. State the access limit. It does not prove the study is poor or that the intervention is ineffective; it prevents this reading from answering those questions.

04

Use a news-reading checklist for context

NCCIH's checklist for complementary-health news asks about participants, comparisons, study duration, limitations and funding. These are useful questions here, not an endorsement of the fictional intervention.

[1]

05

Write the report and the ambition as different sentences

A bounded note for the exercise could say: ‘In the fictional comparison, the six-month visit recorded seven-day self-reported non-smoking in 50/250 versus 40/250 adults. I have not checked allocation, missing-data handling or uncertainty. The cost claim is a future possibility in the release, not a reported evaluation.’

For a real story, attach the original release and the actual paper record, keeping the outcome's time window visible. If the paper is unavailable, say so rather than presenting the release as a completed appraisal. Bring questions about your own support or treatment to a qualified local clinician or tobacco-cessation service.

06

What to keep in mind

• Follow the institution's announcement to the actual report.
• Keep the denominator, comparison and outcome window with the result.
• A researcher's future hope is not another measured endpoint.

07

Common questions