Evaluation And Optimization Of Evidence Quality Grading Method For Public Health Decision-Making

Objective: Comprehensively review the research status,hot spots and gaps of evidence quality grading evaluation tools and methods,systematically evaluate the content and methods of current evidence quality evaluation tools,select and optimize key items of evidence quality grading,construct evidence quality grading system suitable for public health decision-making,and promote evidence can be used efficiently and scientifically to guide public health practice.In order to provide reference for government decision-making.Methods:(1)Bibliometric analysis: The research literature of evidence quality evaluation was comprehensively searched,BICOMS-2 bibliographic co-occurrence analysis system was used to extract the publication characteristics,research characteristics and keyword information of the included research and collate and code it,and the distribution of publication year,journal,institution,country and author was analyzed descriptively.VOS viewer 1.6.10 software was used for social network analysis and keyword cluster analysis.(2)Qualitative systematic evaluation: Chinese and foreign databases were systematically searched,studies with clear evidence classification items,content and classification methods were included,CASP list and AGREE Ⅱ tool were used to evaluate the quality of the included literatures,and subject synthesis method was used to conduct three-level interpretation analysis to construct a pool of evidence quality classification items for public health decision-making.(3)Delphi survey: The Delphi questionnaire was compiled based on the hierarchical items formed by qualitative theme synthesis.24 domestic and foreign experts were selected for letter consultation,and 2 to 3 rounds of letter consultation were planned.By analyzing the positive coefficient of experts,the degree of concentration of expert opinions,the coefficient of expert authority and the degree of coordination of expert opinions,the key items were selected.Based on the investigation,the consensus was formed on the item content,evaluation details and scoring rules of evidence evaluation,the principles and methods of comprehensive evaluation were determined,and the evidence quality grading system for public health decision-making was finally developed.Results:(1)Analysis of research status of evidence quality evaluation: A total of 913 studies were included.The first study was published in 1980.Since 2013,the number of research papers on evidence classification at home and abroad has shown a rapid growth trend.The study was published in 590 journals,and 5,904 authors from 82 countries participated in the study,with the United States,Canada,the United Kingdom and other developed countries at the center of the collaborative network;The analysis of key words shows that English studies mainly focus on health care,chronic disease management,health manpower,resources and services,public health,preventive health services,nutrition and diet,etc.Chinese studies focus on traditional Chinese medicine therapy(TCM)therapy,TCM clinical evaluation,evidence-based nursing and other fields.(2)Systematic evaluation of the evidence quality evaluation system for public health decision-making: A total of 56 studies were included,including 11 high quality studies,9 medium quality studies and 24 low quality studies.Four of them are Grade B guidelines and two are grade C guidelines.After subject synthesis,19 core topics,63 descriptive topics and 32 analytical topics were obtained,including study design,quality of study execution,accuracy,directness,consistency,publication bias,large effect size,dose-response,confounding factors,robustness,coherence of causal inference,funding bias,resource dependence,extensibility,balance of benefits and benefits,applicability,analogy mechanism,health equity,and barriers and promotion factors of evidence.(3)Establishment of evidence quality grading system for public health decision-making: After two rounds of Delphi survey,the experts’ opinions were basically the same,the recovery rate of two rounds of correspondence was higher than 80%,and the experts’ positive coefficient was good.The authority coefficient was higher than 0.85.Kendall coordination coefficients were 0.227 and 0.494,respectively,showing statistical significance(P<0.05).After reaching a consensus based on expert opinions,a public health decision-making evidence quality grading system consisting of 15 first-level items,55 second-level items and 25 third-level items was finally constructed.Based on the evaluation results of each part of items(15 points is the highest overall score),it was finally transformed into four levels of evidence intensity:high,medium,low and very low.The corresponding intervals of each level were as follows: high quality(>11),moderate quality((8-11]),low quality((4-8]),and very low quality(≤4).Conclusion: Based on literature analysis,subject synthesis method and Delphi survey method,the evidence quality grading system for public health decision-making was initially established.It includes 15 grading items,including study design,quality of study execution,accuracy,directness,consistency,publication bias,large effect size,dose-response,negative bias,robustness,importance of causal inference,funding bias,extensibility,health equity,and barriers and promotion factors of evidence.The introduction of a quantitative scoring system has good applicability and feasibility.Its system performance and scalability need to be further verified.