ABSTRACT
OBJECTIVE： To provide experience and reference for the study of medical insurance budget impact analysis （BIA） in China. METHODS： Retrieved from PubMed， ProQuest， CNKI， Wanfang database and CBM， related literatures about medical insurance BIA research in China and the United States were collected since the establishment of the database. The basic information， analysis results and data sources were summarized and sorted out， and descriptive analysis of the included literature was carried out on basis of seven key elements such as model design， research perspective， treatment cost， reference scenario， target population， research time limit and discount/inflation， sensitivity analysis. RESULTS： A total of 72 literatures were included in this study， involving 24 （33.33％） studies in China， 48 （66.67％） studies in the United States； the indications of 45 studies were chronic diseases （62.50％）， and those of 21 studies were acute diseases （37.50％）. Among the research methods， 49 studies （68.06％） used BIA alone and 23 studies （31.94％） adopted BIA combined with pharmaceutical economics. In terms of model design， 50 studies （69.44％） adopted cost calculation models. In terms of research perspective， 60 studies （81.94％） were based on the perspective of medical insurance department research. In the calculation of treatment cost， 69 studies （95.84％） included drug cost. In terms of reference scenarios， 61 studies （84.72％） compared the economics of different drug-based treatment groups. For target population， only 31 （43.06％） studies used real world data. In terms of research duration and discount/inflation， 14 studies （19.44％） used treatment or length of hospitalization to indicate research duration， and 19 studies （26.39％） used discount rate or inflation rate to adjust costs. As for sensitivity analysis， 62 studies （86.11％） conducted sensitivity analysis， of which 49 （68.06％） used single factor sensitivity analysis. CONCLUSIONS： There are still some limitations in medical insurance BIA research literature in China and the United States， such as unreasonable use of data， incomplete coverage of the cost， and unreasonable setting of sensitivity analysis variables. It is recommended that BIA research should standardize data sources to improve the quality of budget evidence quality， reasonably evaluate market size to improve the authenticity of prediction， scientifically set variables and their scope of change to improve the stability of results， establish BIA research paradigms or evaluating standards so as to guide BIA research scientifically.