A false economy: we cannot afford to be complacent when it comes to tuberculosis control
A false economy: we cannot afford to be complacent when it comes to tuberculosis control
复制标题
虚假经济：在结核病控制方面我们不能自满
DOI：
10.1016/s2468-2667(18)30019-7
复制
发表时间：
2018
期刊：
影响因子：
--
通讯作者：
Kirwan D
中科院分区：
文献类型：
--
作者：
Kirwan D
在19世纪，每四个人中就有一个死于肺结核。随着社会经济的改善和医学的进步，结核病的发病率稳步下降，到20世纪80年代，结核病被认为是过去的疾病，控制措施也有所放松。这种做法为时过早：发病率上升导致世卫组织在1993年宣布结核病为全球紧急情况。控制工作再次加强，尽管发病率现已达到峰值，但世界人口增长意味着绝对数量保持稳定。所涉及的成本是巨大的，2017年估计有69亿美元用于全球结核病控制，但对不同人群可用干预措施的有效性仍然知之甚少。在他们发表在《柳叶刀》公共卫生杂志上的研究中，Kianoush Dehghani和同事2分析了影响美国、加拿大和格陵兰岛六个土著人口结核病发病率的因素，这些人口的结核病发病率和社会剥夺都非常高。从1960年至1980年，所有群体的结核病发病率大幅下降，当时整个区域停止了抗结核病干预措施。然后，发病率继续下降，在三个（nonrecrecordescent）人群，但大大增加了其他三个（recordescent）人群。这些数据提供了两个重要的教训。首先，复发与停止抗结核干预有关，证明了这些干预的有效性。第二，研究人群中有一半复发，另一半复发，这一事实并不意味着两组之间存在根本性差异。有效的干预措施是对婴儿进行卡介苗（BCG）接种，以及对潜伏性结核感染（LTBI）进行检测和治疗，但不是对活动性疾病进行胸部X线筛查。尽管在临床试验中的疗效不同，3，4荟萃分析显示卡介苗接种是具有成本效益的，特别是在高发病率的情况下。5.在结核病发病率下降的地区，国际防痨和肺病联合会建议，如果每年平均涂阳肺结核报告率福尔斯降至每10万人5例以下，5岁以下儿童每年平均结核性脑膜炎报告率连续5年低于每1000万人1例，或年均感染风险低于0.1%。6 Dehghani及其同事发现，当结核病发病率“被认为很低”时，就会停止治疗，但所有组的报告率仍高于建议的阈值。这项研究中LTBI检测和治疗的有效性与南非金矿进行的大规模试验中显示的无效性形成对比。[7]在他们的随机分组试验中，Churchyard及其同事将矿工随机分配接受干预（结核病筛查，然后对活动性病例进行治疗或9个月的异烟肼预防性治疗）或对照（无干预）。虽然在9个月的治疗期间，干预组的结核病发病率下降了58%，但没有观察到整体效益。由于职业和家庭接触异常高而造成的再感染风险，加上高违约率造成的低人口覆盖率，可能完全抵消了干预措施的任何效果。数学模型表明，异烟肼预防性治疗在中等发病率的情况下效果最好，在高发病率的情况下效果较差，在高发病率的情况下，...
During the 19th century a staggering one in four people died from tuberculosis. Incidence steadily declined with socioeconomic improvements and medical advances such that by the 1980s tuberculosis was considered a disease of the past and control measures were relaxed. This approach was premature: rising incidence led WHO to declare tuberculosis a global emergency in 1993. Control efforts were re-intensified and although incidence has now peaked, world population growth means that absolute numbers remain stable. The costs involved are substantial, with an estimated US $6· 9 billion spent on global tuberculosis control in 2017, 1 yet the effectiveness of available interventions across different populations remains poorly understood. In their study published in The Lancet Public Health, Kianoush Dehghani and colleagues2 analysed factors affecting incidence of tuberculosis in six Indigenous populations in the USA, Canada, and Greenland with exceptionally high incidence of tuberculosis and social deprivation. Tuberculosis incidence fell substantially in all groups from 1960 to 1980, when anti-tuberculosis interventions were discontinued across the region. Incidence then continued to decrease in three (nonrecrudescent) populations but substantially increased in the other three (recrudescent) populations. The data impart two important lessons. First, recrudescence was associated with discontinuation of anti-tuberculosis interventions, demonstrating the effectiveness of these interventions. Second, the fact that half the populations studied had recrudescence and half did not implies a fundamental difference between the two groups. The effective interventions were bacillus Calmette-Guérin (BCG) vaccination of infants and testing and treatment for latent tuberculosis infection (LTBI), but not chest radiographic screening for active disease. Despite variable efficacy in clinical trials, 3, 4 a meta-analysis has shown BCG vaccination to be cost-effective, particularly in high-incidence settings. 5 In areas where tuberculosis incidence is declining the International Union Against Tuberculosis and Lung Disease (IUATLD) recommends that BCG vaccination could be discontinued if the average annual smear-positive pulmonary tuberculosis notification rate falls below five per 100 000 population, average annual tuberculosis meningitis notification in children younger than 5 years has been below one per 10 million population for 5 years, or the average annual risk of infection is lower than 0· 1%. 6 Dehghani and colleagues found that discontinuation occurred when tuberculosis incidence was “perceived to be low”, yet notification rates remained above the recommended threshold in all groups.The effectiveness of LTBI testing and treatment in this study contrasts with the absence of effect shown in a large-scale trial done in gold mines in South Africa. 7 In their cluster-randomised trial, Churchyard and colleagues randomly assigned miners to receive an intervention (tuberculosis screening followed by either treatment of active cases or 9 months of isoniazid preventive therapy) or control (no intervention). Although tuberculosis incidence fell by 58% among the intervention group during the 9-month treatment period, no overall benefit was observed. The risk of reinfection resulting from exceptionally high occupational and household exposure, compounded by low population coverage due to high default rates, might have entirely negated any effect of the intervention. Mathematical models have shown that isoniazid preventative therapy has the greatest effect in intermediate-incidence settings, becoming less effective in higher-incidence settings where robust …