以數理模式評估危險因子防治對臺灣結核病疫情的影響

The Effect of Risk Factor Control on Tuberculosis in Taiwan: A Modelling Study

王稚慧(Chih-Hui Wang)

指導教授 ： 林先和

國立臺灣大學/公共衛生學院/流行病學與預防醫學研究所/碩士(2016年)

https://doi.org/10.6342/NTU201602123

摘要

背景:
過去十年，臺灣的結核病通報率持續地下降，但結核病仍然是不可忽視的公共衛生議題之一。為了更有效控制結核病疫情，疾病管制署配合世界衛生組織的消除結核策略訂定了新的一期結核病防治計畫，期望能將臺灣的結核病發生率在2035年之前降低至每十萬人十個個案。而控制結核病相關的危險因子是減少結核病疫情的一種介入方式，但過去少有研究評估預防危險因子對控制一個國家的結核病疫情有多少影響。
方法:
本研究建立了一個結核病數理模型，並將其調整為符合臺灣過去的結核病疫情，以評估各個危險因子的介入對台灣結核病疫情的影響。根據過去的系統性文獻回顧，我們假設抽菸、飲酒、糖尿病以及體重過輕會影響模型中的參數。我們根據世界衛生組織的慢性病防治目標以及聯合國提出的永續發展目標訂定未來的危險因子防治目標，並比較達到不同的防治目標下，2035年的結核病發生率將比2015年的基準值改變多少百分比，以此指標評估預防危險因子將如何影響臺灣結核病疫情。
結果:
假設現有的結核病防治策略不變，以及所有的危險因子皆隨著現在的趨勢變化，即抽菸和體重過輕的盛行率下降、飲酒盛行率不變、糖尿病盛行率增加，2035年的結核病發生率將會比2015年減少4.6% (95% Uncertainty interval (UI): -20.3 to 2.3)。將此改變量作為基準，比較達到不同的危險因子介入目標下結核病發生率的改變量與基準值的差。若是達到飲酒盛行率減少10%的目標，結核病累積發生率將比基準值多減少2.9% (95% UI: -4.1 to -1.7)。糖尿病盛行率停止上升則能讓2035年的結核病發生率再多降低9.7% (95% UI: -13.5 to -6.1)。若體重過輕的盛行率於2030年降至零，結核病累績發生率將比基準值減少0.8% (95% UI: -1.2 to -0.5)。而吸菸盛行率停止下降則會使得結核病累積發生率比基準值增加5.1% (95% UI: 2.5 to 7.6)。同時達到飲酒、糖尿病及體重過輕的介入目標，結核病發生率將比基準值減少11.6% (95% UI: -15.3 to -7.8)。在最理想的情況下，即所有危險因子的盛行率都在2030年降至零，結核病發生率將比基準值減少36.0% (95% UI: -47.1 to 24.0)。
結論:
本研究顯示藉由危險因子的控制，能降低結核病的發生率。而在這四個危險因子當中，因為糖尿病大幅上升的趨勢，糖尿病的控制與預防對結核病疫情的影響是最大的。結核病防治與慢性病防治的整合應有利於達到消除結核的目標。

關鍵字

結核病 ； 數理模型 ； 吸菸 ； 喝酒 ； 糖尿病 ； 體重過輕

並列摘要

Background
In the new End TB Strategy, management of TB risk factors, including smoking, alcohol use, diabetes, and underweight, is considered to be a critical component. Few studies have assessed the impact of reducing multiple risk factors on future trend of TB incidence at the national level. The aim of this study is to evaluate the effect of risk factors control on TB incidence in Taiwan.
Methods
We constructed a compartmental model with age structure for tuberculosis transmission, and calibrated the model to the trend of TB incidence in Taiwan. Smoking, alcohol use, diabetes, and underweight were assumed to affect the parameters related to tuberculosis natural history in the model, following the results of previous systematic reviews. We constructed future scenarios of risk factor control based on the World Health Organization’s Noncommunicable Disease (NCD) control targets and the Sustainable Development Goals. The effect of risk factors control was evaluated by comparing the percentage of reduction in TB incidence under different scenarios between 2015 and 2035.
Results
In base case scenario, if current tuberculosis control measures remain unchanged and all the risk factors follow the current trend (decreasing prevalence of smoking and underweight, stable trend of alcohol use, and increasing prevalence of diabetes), the cumulative reduction of TB incidence would only be 4.6% (95% UI: -20.3 to 2.3) in 2015-2035. Compared with the base case scenario, the TB incidence would be further reduced by 2.9% (95% UI: -4.1 to -1.7) if the target of alcohol use reduction is achieved. If diabetes stops rising after 2015, the relative reduction in cumulative TB incidence would be 9.7% (95% UI: -13.5 to -6.1). In the scenario of end hunger, the relative reduction would be 0.8% (95% UI: -1.2 to -0.5) if the prevalence of underweight reduces to zero by 2030. However, if the prevalence of smoking stops declining, the relative increase in cumulative TB incidence would be 5.1% (95% UI: 2.5 to 7.6). For the interventions on multiple risk factors, achieving the targets of alcohol use reduction, diabetes stop rising, and ending hunger simultaneously would result in further 11.6% (95% UI: -15.3 to -7.8) reduction in cumulative TB incidence compared with the base case scenario. If all risk factors reduce to zero by 2030, the relative reduction would be 36.0% (95% UI: -47.1 to 24.0) by 2035.
Conclusions
Our study indicates that controlling the risk factors at the population level could reduce TB burden. Among the four risk factors considered, the impact of diabetes control would be the greatest given to the observed rising trend of diabetes. Coordinated efforts between the national TB program and the NCD program will be required to help achieve the target of the End TB strategy.

並列關鍵字

tuberculosis ； dynamic model ； smoking ； alcohol ； diabetes ； underweight

參考文獻

5. 行政院衛生署, 結核病十年減半全民動員計畫. 2006.

7. Su-Hua Huang, C.-B.H., Kwei-Feng Wang, Shiang-Lin Yang, Yen-Fang Huang, Chang-Hsun Chan, Progress and Implementation of Surveillance of Directly Observed Treatment, Short-Course (DOTS) for Tuberculosis in Taiwan Taiwan Epidemiol Bull 2014. 30(5): p. 92-100.

48. Liang, C.-Y., et al., Prevalence Rates of Alcohol Drinking in Taiwan. Taiwan J Oral Med Health Sci, 2004. 20(2): p. 91-104.

3. Chan, P.C., L.M. Huang, and H.S. Kuo, Is neonatal Bacillus Calmette-Guerin vaccination protective in Taiwan? J Formos Med Assoc, 2008. 107(3): p. 195-7.

8. Lonnroth, K., et al., Drivers of tuberculosis epidemics: the role of risk factors and social determinants. Soc Sci Med, 2009. 68(12): p. 2240-6.