domingo, 15 de junio de 2025

Systematic reviews cited retracted articles, new study finds Could AI tools address such flaws and make reviews a ‘true gold standard’?

https://www.statnews.com/2025/06/12/researchers-examine-scientific-rigor-of-systematic-reviews-new-ai-tool-may-help/?utm_campaign=rss&_hsenc=p2ANqtz-_lAvvQji5W_eOzRvLKi_f50p8TphnpNQpKOd6WR9QzdDAKLC9aFhHVXUElawP13tAkQI4HEL2JVscgPLoyfzlmiSp9vQ&_hsmi=366507755 Whenever I see a literature review for a topic I’m writing about, I often think, “Oh, great! I can trust this paper.” Based on reporting from STAT’s Anil Oza, I might need to reevaluate this impulse. A new JAMA Network Open paper adds to a body of evidence that not all systematic reviews are created equal and some researchers are trying out alternatives to these analyses. https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2835228?utm_campaign=morning_rounds&utm_medium=email&_hsenc=p2ANqtz-_GkfpeCXOeL4r3K1X5ImRJmgA2QZt-PJTTfOgbYPOMR4uO8v_UGLP63RSLrLJDIt1rw5zOXAoDwv26KyFxd2QA1fBPdA&_hsmi=366507755&utm_content=366507755&utm_source=hs_email Systematic reviews, which involve pooling data from multiple studies and analyzing them together, are increasingly popular as a way to produce more authoritative conclusions than can be derived from the individual smaller papers. A small number of systematic reviews included retracted studies from paper mills, journals that produce low quality research to help researchers game the academic job market. The problems are not limited to academia, either. In recent weeks, systematic reviews have been used to justify policy around gender affirming care as well as in the MAHA report. Media reports quickly followed in their wake showing that these reports were riddled with errors. Read more from Anil. https://www.whitehouse.gov/wp-content/uploads/2025/05/WH-The-MAHA-Report-Assessment.pdf?utm_campaign=morning_rounds&utm_medium=email&_hsenc=p2ANqtz-8UVba8JDtdyuckH_yQm11-7T2QTVRfxW0lmMEEnGMUPslVIonmdm1YW51_2NVogYvGvL-i6fKZjXDj5qvFGhD2WOINAA&_hsmi=366507755&utm_content=366507755&utm_source=hs_email

No hay comentarios: