藥癮減害對注射藥癮傳播愛滋疫情的防治效果:數理模式研究

Abstract

背景: 注射藥癮族群常伴隨著共用針具行為，因此是感染HIV的高危險族群， 這是全球重要的公共衛生議題之一。因此，世界衛生組織推薦藥癮減害 (Harm Reduction Program, HR)，欲減緩注射藥癮族群感染HIV的風險。藥癮減害主要包含了發放清潔針具計畫 (needle-syringe programs, NSPs) 和鴉片類藥物替代療法 (opioid substitution therapy, OST) 。然而，關於在闡述藥癮減害成效的文獻中，大多都受限在注射藥癮族群較差的追蹤率，而在大規模的全人口研究中，藥癮減害的成效又受限於並未考慮到HIV疫情的下降是不是因為疫情自然演變 (natural course) 的關係? (疫情自然演變: 疫情傳播使未感染的危險族群逐漸減少而自然導致的結果)。台灣自從2004年開始，經歷過一波嚴重的藥癮愛滋疫情，而當時的疾病管制署開始舉辦藥癮減害，而後疫情迅速地受到控制，儘管如此，在國內藥癮減害對於藥癮愛滋疫情的貢獻仍然未有實際的數據可以表達。

目的: 本研究目的為建構一個真實的HIV動態傳播的模式，而整體的模式要能解釋現實生活中注射藥癮族群的HIV新發個案數的趨勢，然後運用此模式去評估藥癮減害的成效 (以疫情自然演變為對照組) 。

方法: 本研究是一個傳播動態的數理模式研究，裡面的參數是根據實證資料的數據，這些實證資料包含相關的文獻回顧，還有一些行政上的資料 (台灣疾病管制署的HIV監測資料以及台灣美沙冬資料庫)。我們以數理模擬方式主要分析：沒有藥癮減害介入的疫情自然演變狀況，以及有藥癮減害介入的策略下，對於消除藥癮愛滋的疫情效果是如何，另外，本研究的次要分析為: 分別單獨使用NSP、OST、擴大篩檢、衛教對消除藥癮愛滋的疫情效果是如何，以及民國96年實施減刑政策，對於疫情的影響。使用的軟體為Stella Professional。

結果: 本研究用真實世界中所取得的相關防治指標進行配適，建構出疫情曲線，並進一步比較有無介入措施前後的發生率變化，藉此推論介入措施的效果。主要結果顯示: 在考慮到IDU入監、擴大篩檢、衛教、減刑等真實世界狀況後，藥癮減害 (NSP+OST) 的效果為68.56%，且到2018Q1為止能預防10930位藥癮者感染HIV；次要結果顯示: 單獨使用NSP的效果為19.92%，且能預防1701位藥癮者感染HIV；單獨使用OST效果為28.9%，且能預防2779位藥癮者感染HIV; 單獨有擴大篩檢的效果為35%，且能預防3682位藥癮者感染HIV；單獨有衛教效果為10.77%，且能預防825位藥癮者感染HIV；民國96年的減刑政策效果為-1.1%，對疫情防治有負面影響，會多增加74位藥癮者感染HIV。

結論: 本研究確認藥癮減害對於台灣藥癮愛滋疫情的貢獻很大，並且發現NSP及OST一起實施時會有加成作用。因此，我們建議在台灣應該繼續推行藥癮減害。

研究優勢與限制: 本研究的優勢是利用台灣的實證資料去估計數理模式裡的參數，因此可以對重要的公共衛生議題去做一個真實性的模擬。另外，本篇的侷限在於真實世界太過複雜，一些不重要的細節必須要簡化才能做數理模型的研究。

Background: Globally, injecting drug users are at high-risk of acquiring HIV infection due to their needle-sharing behaviors. The World Health Organization recommends harm reduction services, esp. needle-syringe programs (NSPs) and opioid substitution therapy (OST), to mitigate HIV risk in this vulnerable population. However, data on the effectiveness of harm reduction is limited by suboptimal follow-up rate. Efficacy in population level is particularly weak due to lack of consideration of natural course of epidemic. Taiwan experienced a nationwide large injecting-drug-use (IDU)-transmitted HIV outbreak since 2004. Centers for Disease Control launched National Harm Reduction (HR) Program, and the epidemic rapidly decreased. Nevertheless, the relative contribution of National HR program on the rapid control of IDU-HIV epidemic remains uncertain. Aims: We aims to construct a realistic HIV dynamic transmission model to explain the IDU-HIV surveillance data, as well as evaluate the effect of National HR program on IDU-HIV epidemic, using natural course of epidemic as the reference. Materials and Methods: This is a transmission dynamic model study. We estimate the parameters in the model from empirical data, including national HIV/AIDS surveillance database, methadone database and Taiwan vital statistics. The main analysis: the natural evolution of the IDU-HIV epidemic without harm reduction, and the effect of eliminating the IDU-HIV epidemic with harm reduction; The secondary analysis: the effect of eliminating the IDU-HIV epidemic with alone NSP, alone OST, alone expanding HIV testing and counselling (HTC), alone education and alone amnesty policy. The software used is Stella Professional. Results: We constructs the epidemic curve with data in real world, and further compares the incidence changes before and after intervention measures, to evaluate the effect of intervention measures. The main result shows: HR program (NSP+OST) has prevented 10,930 (68.56% of what would have occurred in the absence of HR) of IDU-HIV cases by 2018 Q1. The secondary results show: alone NSP has prevented 1,701 (19.92% of what would have occurred in the absence of NSP) of IDU-HIV cases by 2018 Q1; alone OST has prevented 2,779 (28.9% of what would have occurred in the absence of OST) of IDU-HIV cases by 2018 Q1; alone HTC has prevented 3,682 (35% of what would have occurred in the absence of HTC) of IDU-HIV cases by 2018 Q1; alone education has prevented 825 (10.77% of what would have occurred in the absence of education) of IDU-HIV cases by 2018 Q1; alone amnesty policy has increased 74 (-1.1% of what would have occurred in the absence of amnesty policy) of IDU-HIV cases by 2018 Q1. Conclusion: We confirms the harm reduction has contributed a lot to IDU-HIV epidemic in Taiwan. and we found that NSP and OST have an additive effect when implemented together. Therefore, we suggest that harm reduction should be continued implemented in Taiwan. Strength and Limitation: This proposal uses empirical data in Taiwan to estimate parameters in the model, and therefore provides realistic simulations for an important public health issue. One limitation in our study is that some details in the real world need to be simplified for the purpose of mathematical modeling.