以系統動力學模擬登革熱擴散現象與評估防治策略效果－台南市為例

Simulation and Evaluating Epidemic Intervention Policies with System Dynamics: A Case Study of Dengue Fever in Tainan

黃經洲(Jin-Jun Huang) ； 陳加屏(Chia-Ping Chen) ； 艾昌瑞(Chang-Rui Ai)

《台灣公共衛生雜誌》 28卷6期 (2009/12) Pp. 541-551

https://doi.org/10.6288/TJPH2009-28-06-10

摘要

目標：建構一個以台南市人文、氣候、感染人數等實際資料為依據的登革熱擴散因果模式，並用以評估數種防治策略的效果。方法：運用系統動力學，建構感染者中心－雙流巴氏模式(Agent-Dual flow Bass Model)。感染者中心以受感染者的活動半徑進行模擬，取代行政區域的人為劃分。單流式是人傳人的感染擴散型式，雙流式是指人與蚊交互接觸感染的擴散型式。結果：1.以感染者中心方法取代原本以人為區域劃分法，重新描繪台南市2007年登革熱感染人數變化曲線後，發現本模式的模擬行為能彌補人為區域劃分法的缺失。2.試著將氣溫增加2℃、雨量增加0.5倍時，感染人數會比原氣象水準高5.5倍，顯示登革熱暴發率對氣溫與雨量等氣候因素具高敏感性。3.防治策略能抑制感染人數的成長，而五種策略中，以噴殺蟲劑效果(-85%)最好，若合併三項同屬教育性質的策略，其綜合效果(-84%)與噴殺蟲劑旗鼓相當。結論：以氣溫閥值（約24.2℃），取代病媒蚊指數做為登革熱爆發門檻的領先指標，更能提前預防登革熱。

關鍵字

巴氏Bass模式 ； 登革熱 ； 系統動力學 ； 防治策略

並列摘要

Objectives: This research aims were to construct a modified model of dengue fever dispersion, taking into consideration geographic, demographic and climatic variables, and then evaluate the efficacies of intervention policies currently in use. Methods: From a system dynamics perspective, this study constructs a modified Agent-Dual Flow Bass Model. The definition of the variable Agent refers to the infectious population's activity sphere rather than a man-made administrative district. Single Flow indicates human-to-human spread, while Dual Flow refers to interactive infections between humans and mosquitoes. Results: 1. The modified Agent-Dual Flow Bass Model results in a better simulation correlation than the man-made administrative-district simulation by re-coding the 2007 dengue fever cases in Tainan. This result suggests that the new model can overcome the deficiencies in the previous model. 2. With a 2℃ temperature increase and a one-half times greater rainfall, the infectious population increases by 5.5 times, which indicates that a dengue fever outbreak is highly sensitive to temperature and rainfall. 3. The intervention policies are effective for decreasing the spread of dengue fever. Among the five intervention policies, fumigation is the most effective (-85%). However, when the three educational policies are combined, the effect (-84%) is equivalent to fumigation policies. Conclusions: These results suggest that using a temperature threshold (24.2℃) as the leading indicator for dengue fever outbreak provides an earlier warning than the conventional Bretaeu index.

並列關鍵字

Bass Model ； Dengue Fever ； system dynamic ； intervention policies

參考文獻

Barnett WP.(1990).The organizational ecology of a technology system.Adm Sci Q.35,31-60.

Bass FM.(2004).Comments on "a new product growth for model consumer durables".Management Sci.12(Suppl),1833-1840.

Brauer F.(2006).Some simple epidemic models.Math Biosci Eng.3,1-15.

DeMello S.(1993).Systems thinking and strategic planning in healthcare.Syst Thinker.4,3.

Derouicha M,Boutayeb A.(2006).Denguefever: mathematical modeling and computer simulation.Appl Math Comput.177,528-544.