摘要: The rapid growth in scientific publications has been accompanied by an increase in fake or poor-quality papers, threatening the integrity of scientific literature. To address this issue, I propose the Publication Health Index (PHI), a novel metric that evaluates the health status of a set of articles based on retraction rates. Using Nature and Science as benchmarks, I suggest that a healthy PHI threshold should be established at 7.2, equivalent to a retraction rate of 0.07%. The PHI is then applied to evaluate the publication health of journals, countries, and institutions, demonstrating its broad applicability and practical feasibility. This metric shows potential to help governments shape science policies and publishers improve the quality of scientific literature.