台灣產業外籍移工的結核病傳染動態及控制策略研擬

Modeling the transmission dynamics and infection control of tuberculosis in the immigrant worker of Taiwan

王子云(Tzu-Yun Wang)

指導教授 ： 陳詩潔

中山醫學大學/健康管理學院/公共衛生學系/碩士(2021年)

https://doi.org/10.6834/csmu202100102

摘要

結核病(Tuberculosis)為全球性的慢性傳染性疾病，目前仍為全球十大死因之一。台灣位於結核病高發生率的東南亞地區，一年四季都有病例。有研究顯示來自結核病高發生率國家的移民對於低發生率國家的結核病會有影響，近年來，來台的產業外籍移工人數越來越多，從2013年為280,928到2019年上升為458,619人，增加63%，且都是來自高發生率的國家。本論文目的為建構本土族群與產業外籍移工族群的傳染動態模型，且利用模型來模擬未來10年產業外籍移工來台是否會對台灣的結核病造成影響，以及透過模擬實施控制策略找出台灣三個縣市(桃園市、台中市以及新北市)結核病的最佳控制策略方式。本研究建立易感-潛伏-感染-復原(Susceptible-Latent-Tuberculosis Infection-Recovery, SLTR)模型，此模型將分為兩大族群:本地族群(NL)與產業外籍移工族群(NM)，又將族群細分為易感性(Susceptible, S)、潛伏性(受感染但非傳染，Latent, L)、傳染性(開放性，Tuberculosis Infection, T)及復原性(Recovery, R)，共八個區塊(SM, LM, TM, RM和SL, LL, TL, RL)，並透過靈敏度分析評估哪些參數對於模型的輸出影響最為明顯。為了解如何有效減少結核病的傳播與控制，將模擬實施三種控制策略:距離控制(u1)是透過減少傳播率(β)，減少與傳染性族群接觸，盡而減少其他族群的感染;早期篩檢(u2)是透過減少重新激活率(k)，定期篩檢早期發現早期治療，避免發展為開放性結核病；都治計畫(u3)是提高復原率(γ)，確保每位患者能按時服藥，避免產生抗藥性，將三種策略又分為單一控制(u1、u2、u3)、雙重控制(u1+ u2、u1+ u3、u2+ u3)以及三重控制(u1+ u2+ u3)。本研究結果指出，三個縣市在產業外籍移工族群(NM)總數是持續上升，但在本地族群(NL)是呈現下降的，這表明在台灣的本地結核病是有逐步地控制，但要加強對產業外籍移工的防治；在靈敏度分析中，影響模型最為明顯的參數是產業外籍移工之間的傳播率(βM)以及本地之間的傳播率(βL)；在三個縣市的控制策略模擬結果顯示，單一控制策略中以距離控制(u1)為最佳，雙重控制中最佳的為距離控制加早期篩檢(u1+ u2)，三重控制策略比雙重控制策略好，但成效相差不大(1~3%)。本研究提供台灣本地與產業外籍移工的結核病之族群傳輸模型架構，也提供三個縣市最佳的結核病控制策略。

關鍵字

結核病 ； 外籍移工 ； 動態模型 ； 控制策略 ； 參數估計

並列摘要

Tuberculosis (TB) is a global chronic infectious disease, and it is still one of the top ten causes of death in the world. Taiwan is located in Southeast Asia with a high incidence of TB. Studies have shown that immigrants from countries with a high incidence of TB have a key effect on areas with low incidence. Recently, the number of immigrant workers who came to Taiwan increased from 280,928 in 2013 to 458,619 in 2019, with an sharply increase of 63%. Here, the objective of this theme is to construct a dynamic model of TB and compose it with local population (NL) and immigrant workers (NM). Each subpopulation is further subdivided into four classes: the susceptible (S), latently infected (L), infectious TB (T) and recovered (R), SLTR model, which results in eight compartments SM, LM, TM, RM and SL, LL, TL, RL. We used the SLTR model to simulate the TB dynamics in Taoyuan City, Taichung City and New Taipei City from 2018 to 2028. We also conducted a sensitivity analysis of parameters and implemented three control strategies, including social distance control, early screening and Directly Observed Treatment Short-Course (DOTS). Social distance control (u1) is to reduce the transmission rate (β), reduce contact with infectious ethnic groups, and minimize the infection of other ethnic groups. Early screening (u2) is to reduce the reactivation rate (k), and regular screening to detect early treatment early and avoid the development of pulmonary tuberculosis. And, the DOTS (u3) is to increase the recovery rate (γ) to ensure that each patient can take the medicine on time to avoid the development of drug resistance, the three strategies are divided into single control strategies (u1, u2, u3), coupled control strategies (u1 + u2, u1 + u3, u2 + u3) and three control strategies (u1 + u2 + u3). Results indicate that the total number of immigrant workers (NM) in three study areas continues to increase, but the local population (NL) is decline from 2018 to 2028. Hence, we must enhance the ability to provide prevention and treatment services for immigrant workers. Besides, the most sensitive parameter is the transmission rate for local population (βL) and immigrant workers (βM) based on sensitivity analysis. Our finding in this study is that (i) enhancing social distance control (u1) is shown the best strategy than other measures; (ii) the control effectives of u1+ u2 and u1+ u2 + u3 are slightly different (1 – 3%). Hence, this study provides a compartment model and simulate the dynamics modeling for TB infection between local and immigrant workers in Taiwan, as well as the effectiveness by different control strategies in Taoyuan City, Taichung City and New Taipei City.

並列關鍵字

Tuberculosis ； Immigrant workers ； Dynamic model ； Control strategy ； Parameter estimation

參考文獻

Aldridge RW, Yates TA, Zenner D, White PJ, Abubakar I, Hayward AC. 2016. Pre-entry screening programmes for tuberculosis in mi grants to low-incidence countries: a systematic review and meta-analysis. The Lancet Infectious Disease 14(12): 1240-1249.

Bai KJ, Ching CY, Lee CN, Chang JH, Wu LC, Yu MC. 2008. Tuberculosis among foreign-born persons in Taiwan, 2002–2005. Journal of the Formosan Medical Association 107(5): 389-395.

Blower SM, Mclean AR, Porco TC, Small PM, Hopewell PC, Sanchez MA, Moss AR. 1995. The intrinsic transmission dynamics of tuberculosis epidemics. Nature Medicine 1: 815-821.

Chaulet. 1983. Treatment of tuberculosis : case holding until cure. Technical Report: World Health Organization.

Dale KD, Trauer JM, Dodd PJ, Houben RMGJ, Denholm JT. 2020. Estimating long-term tuberculosis reactivation rates in Australian migrants. Clinical Infectious Diseases 70(10): 2111-2118.