5.13.5 News Releases
Many journals issue news releases on selected articles determined by the editors to be of interest to the public. For JAMA and the Archives Journals, experienced science writers prepare the news releases, which are reviewed by the editors to ensure accuracy and objectivity. News releases of journal content should be under the authority of the editor, not the journal’s publisher or owner (see also 5.10, Editorial Freedom and Integrity).
News editors, writers, and producers receive hundreds of news releases a week. Thus, a news release must attract attention, but it also must conform to a familiar format and style (see Box). Journalists are taught to present facts accurately, but they may not know how to interpret biomedical statistics or understand the specific context of new scientific information. In news releases and news stories, research findings and statistics are often cited inaccurately or out of context to support an exaggerated medical claim.14,15,31,32 To help prevent exaggerated or misleading claims, news releases must include accurate and clearly stated statistics33 (see 20.1, Study Design and Statistics, The Manuscript: Presenting Study Design, Rationale, and Statistical Analysis). In addition, research findings must be placed in proper context and should include important background, summary of study methods, limitations of the methods, and information on study sponsorship and relevant conflicts of interests of authors (see 5.5, Conflicts of Interest). Care should be taken to provide balance (eg, citing a related editorial) and to avoid sensationalism (eg, use of terms like breakthrough). Examples of common problems to avoid in news releases are listed below:
Common Problems to Avoid in News Releases.
▪ Unfamiliar mathematical and statistical terms and numbers that are difficult to interpret should be avoided; do not confuse association and correlation with causation.
▪ Results should be reported in context, including locations and dates of the study, representativeness of the sample, and limitations of the study. Risks of events should be acknowledged to be common (eg, common cold) or rare (eg, being hit by lightning).18
▪ If the results of a survey are reported, the response rate should be provided along with a caveat that the results may not be generalizable if the response rate is low.
▪ If a news release mentions a specific sample that was studied or a specific number of cases, whether the number is large or small, information about the size of the total population from which the sample or cases were drawn should be included.
▪ Statements about statistical significance should not be quoted from an article out of context or without an explanation. Reporters and readers do not necessarily know the difference between statistical significance and clinical significance. For example, quoting a statement that there was a trend toward a statistically significant association between treatment X and outcome Y may give undue importance to a treatment that has no real clinical value.
▪ Absolute event rates should be reported. Care should be taken to avoid confusing absolute and relative risks because relative risks are often erroneously translated to specific risks. For example, a decrease from 2.5% to 2.0% should not be reported as a 20% reduction in risk, but could be reported as a 0.5% absolute risk reduction and 20% relative risk reduction. It is also helpful to report excess or decreased risk in terms of numbers per 1000 or 10 000.
▪ Avoid reporting odds ratios, especially for common outcomes, which may overstate a relative risk.18
▪ If reporting the results of a study about an intervention, event rates for benefits and harms should be reported equally and in a balanced manner.18
Before news releases are distributed, they should be proofread and the content should be reviewed by a professional familiar with the article or report covered in the release, or by the editor.