利用整合式傳染病監測軟體以協助流行性感冒的監測並提供公衛參考

Abstract

台灣疾病管制局的疫情報告顯示，由於H5N1高病原性禽流感病毒的高致死率和快速變種，促使衛生政策決策當局將流感疫苗、抗病毒藥劑及傳染阻絕列為新型流感三大防疫策略。為了評估新型流感流行之可能性與傳播動態，並探討現行公共衛生政策在遏止新型流感擴散的成效，本研究建立一個結合人際接觸社會網路與多代理人系統的新型流感傳播動態之模擬模型。 我們的模型共分為三個部份:社會網路、個體差異和疾病特徵。在社會網路層面，我們利用多個彼此連通的行政區域與人口族群來表現符合現代生活模式的空間群聚、區域間往來的交通現象。在個體層面，我們考慮個體的屬性如:超級感染源、免疫狀態、傳染力和行動力等等。在疾病特徵層面，我們以(SEIR)來表現疾病感染進程。 我們利用模型來探討當疫情從台灣不同縣市爆發時，會造成的全台疫情的散佈情形為何。並且從各個地區的感染情形、疾病爆發的高峰期、疾病的傳染率等等，分別探討新型流感可能的傳播動態。在防疫政策方面，我們將探討三種措施：疫苗、交通管制和特效藥。在流感疫苗方面，我們發現疫苗將可大幅控制疫情。在交通管制方面，我們也發現對疫區的交通管制，其效果並不會比對全國的交通管制所得的效果來的差。在新型流感的特效藥方面，我們針對藥效、施打地點、藥劑量等這些細節上去做模擬，發現預防性投藥在疾病爆發初期的確會有其效果，但若沒有控制好，最後感染人數仍可能會比沒有預防性感染人數來的高。 我們以此模型來探討新型流感在全台各區可能的傳播動態；並模擬現行各種公衛政策在疫情爆發在各分區的施行效果。

According to the Taiwan Center for Disease Control (CDC), the highly pathogenic avian influenza virus H5N1 has evolved so quickly and poses such a threat that Taiwanese health authorities have already announced three major control strategies: a vaccine, antiviral drugs, and isolating infected individuals. We constructed an influenza simulation model based on daily-contact social networks and a multi-agent system to study transmission dynamics and to investigate the potential efficacies of various public health policies. Our model has three major parts: social networks, the individual diversity and epidemiological factors. Daily-contact social networks (composed of multiple interconnecting municipalities and individuals) can be used to represent such social phenomena as long-distance movement, daily visits to fixed locations, different population densities, and inter-area transportation. We analyzed the dynamic of epidemic situation in 22 areas in Taiwan, including the high peak of the burst of the epidemic and infection rate. Our model has three major part: vaccine, traffic control and medication. We found that the vaccine can get best results. The traffic control in epidemic area can get results as good as the global traffic control in whole Taiwan. Considering special remedy, we simulated in the effect of medication, injection location and total doses. We found that protective inoculation has a good effect in the beginning of the burst of the epidemic, but probably an opposite effect if total doses are not enough. The model was used to simulate a novel influenza outbreak and other scenarios involving different seeding events and interventions – for example, antiviral prophylaxis, social distancing policies or closures for schools and workplaces, and control of inter-area movement.