应用分析流程图法建立流行性感冒风险预报等级

Establishment of influenza risk forecasting model by analytical flow chart method

目的应用分析流程图法建立流行性感冒(流感)风险预报分类树图模型,并指导居民采取相应的预防措施。方法收集浙江省舟山市2015—2018年的流感监测资料,采用分析流程图法,依据是否处在流感流行季节、流感样病例就诊百分比(ILI%)及有无有效的防控和治疗措施进行分层,拟合2018年流感趋势线、警戒线和行动线,评估流感发病风险等级并作为预报模型,采用2018年第1—20周每周流感监测实际资料对预报模型作验证,评价其用于每周流感风险指数预报的效果。结果流程图构建的分类评估预测模型显示,流感发病指数共分为Ⅰ级极高风险(n=0)、Ⅱ级高风险(n=14,ILI%=12.61%±7.54%)、Ⅲ级较高风险(n=16,ILI%=7.76%±3.53%)、Ⅳ级一般风险(n=57,ILI%=6.31%±2.26%)、Ⅴ级较低风险(n=32,ILI%=5.37%±1.96%)和Ⅵ级低风险(n=37,ILI%=4.30%±1.40%) 6个级别。2018年1—20周流感发病的实际风险指数与模型预报的风险指数的一致率为70.00%(14/20)。结论流感发病风险等级评估模型对每周流感发病的分级预报有较好的应用价值。

Objective To apply analytical flow chart method to building a model for forecasting the risk of influenza and providing preventive measures. Methods Surveillance data of influenza in Zhoushan from 2015 to 2018 was collected. Analytical flow chart method was used to evaluate the real-time risk of influenza,which were stratified by the seasonality of influenza,the percentage of influenza-like patients(ILI%) and whether there were effective prevention and treatment measures. The trend-line,warning line and action line of the influenza in 2018 were fit and the risk grades of influenza forecast were defined,which were compared with the actual risk grands during the first 20 weeks in 2018,to evaluate the accuracy of influenza forecast index. Results Through the analysis of the flow chart, the influenza index was divided into six grades, very high risk(gradeⅠ,n=0),high risk(gradeⅡ,n=14,ILI%=12.61%±7.54%),relative high risk(gradeⅢ,n=16,ILI%=7.76%±3.53%),general risk(gradeⅣ,n=57,ILI%=6.31%±2.26%),relative low risk(gradeⅤ,n=32,ILI%=5.37%±1.96%)and low risk(grade Ⅵ,n=37,ILI%=4.30%±1.40%). The concordance rate between the forecast risk and actual risk was70.00%(14/20). Conclusion This influenza forecasting model could well forecast the levels of influenza risk each week.