AIM. To describe and compare the representation of epidemiological research in 2009 in the articles of the online editions of The Irish Times (IT) and Postimees (PM) and, proceeding from the results, to discuss the possibilities to make media representation more reader friendly.
METHODS. Content analysis was used to analyse 269 articles selected by similar principles from the online archives of the IT and PM online editions.
RESULTS. IT published more often the results of local studies (IT 59%, PM 9%), and used them more in discussions related to health policy (IT 18%, PM 8%) as well as gave health recommendations (IT 15%, PM 3%). PM publ ished more often the results of foreign studies (PM 78%, IT 36%), including detailed information, and used the reporting style to present the results (PM 87%, IT 58%). The topics covered by IT are based more on the national statistics of public health than the topics covered by PM. However, the most frequently discussed research topics are similar both in PM and IT: nutrition and diet, socioeconomic and psychological health determinants, and mental health. When presenting the results of the studies both newspapers used most often numerical indicators supplying them with verbal interpretations (73%). The most often used statistical indicator was prevalence proportion (51%). In one fifth of the articles parameters of relative risk and absolute numbers were used. IT provided more frequently absolute numbers (IT 30%, PM 19%), which can be considered good practice in making the results easily understandable. For creating a broader context, IT presented more frequently the results of more than one study (IT 58%, PM 33%) in an article, with more expert comments (IT 46%, PM 21%).
CONCLUSION. The articles of IT reflected less problematic issues of the scientific literature. Introduction of local studies aimed at practical goals (more common practice in IT) is more valuable from the perspective of readers than reporting the results of foreign studies (more common practice in PM).