Objectives Science journalism plays an important role in bridging the knowledge gap between scientists and the general public. By failing to address that a study was undertaken in rodents rather than humans, news outlets may inadvertently contribute to patients and health professionals making ill-informed decisions. In April 2019, James Heathers started the viral @justsaysinmice twitter account to highlight the issue of inaccurate science news reporting. The account highlights tweets of news and media reports of pre-clinical research in mice that do not make it clear in their headlines that the research was conducted in mice. This is done by the simple act of re-tweeting the original tweet with the quote ‘in mice’. The objective of this study is to quantify and evaluate the impact of the @justsaysinmice on reporting accuracy of the headlines and/or lede of media reports about studies carried out in mice.

Method We will perform an interrupted time series analysis on the numbers of records reporting the phrase ‘in mice’ in the ‘Headline’ and ‘Headline and Lead Paragraph’ sections of all reports in the Dow Jones Factiva news database. Searches will be performed in monthly periods between May 2016 and end of March 2022. A separate analysis of reporting of ‘in mice’ in highly accessed publications will be performed. A generalised linear model with binomial error structure accounting for the month-by-month variation in denominators (that is, the number of reports eligible to report ‘in mice’ each month) will be used. Seasonal effects will be accounted for by adjustment for the calendar month. Time will be divided into two segments: before and after the exposure period (April 2019). We will investigate effect modification by the following prespecified factors: Covid-19 reporting, proportion of reports from traditional (print) versus other media types (online).

Results We will present raw counts of reports with ‘in mice’ in the headline and/or lede as an indication of the trends present in news reporting of scientific findings in mice over the course of six years. Results of the interrupted time series analysis will be presented as the ratio of the odds of reporting ‘in mice’ in the headline and/or lede before compared with after April 2019, with a value greater than 1.0 indicating greater odds of reporting ‘in mice’ before compared with after April 2019 and a value below 1.0 indicating lower odds of reporting ‘in mice’.

Conclusions This study will provide information on the association of a corrective social media intervention and subsequent accuracy of health and biomedical media reporting. Improved accuracy of media reporting will ensure users of health research can make better and fully informed healthcare decisions.