When the impact is not the same everywhere: The need for regionalized characterization factors in the LCA of plastics.

LCA presents important methodological gaps in the impact assessment phase, especially regarding the characterization factors (CF) applied for plastics. The purpose of this article is to identify methodological barriers and opportunities for improvement in the CFs proposed for plastics-related impacts. It discusses the need to advance toward regionalized characterization factors that better reflect the specific environmental, social, and technological conditions of different geographical contexts. The methodology adopted for this analysis has three phases. First, a systematic review of studies focused on CF for plastics was carried out following PRISMA guidelines, making this the first study to explicitly examine geographical and methodological bias in studies addressing the development of characterization factors for plastics. Second, a triple-axis analytical framework covering thematic, methodological, and CF construction dimensions was applied to identify trends, divergences, and emerging research directions. Third, the findings were synthesized to evaluate the strengths, limitations, and applicability of existing characterization factors across different environmental and geographical contexts. As a result, this article highlights both the strengths and limitations of current CF construction methodologies, Through the application of a triple-axis analytical framework, it provides the first structured synthesis of methodological and geographical patterns in plastic-related characterization factor development, identifying key methodological requirements, research gaps, and future research directions. These findings may support the future development of regionalized characterization factors capable of better reflecting diverse ecological conditions, ecosystem types, impact categories, and LCA modelling approaches, particularly in geographical contexts where environmental and socioeconomic heterogeneity remains underrepresented in current global datasets.