文献速递流行病学与生物统计基础与转化研究

尿液-LAM诊断对结核病发病率和死亡率的潜在影响:一项模型分析

The potential impact of urine-LAM diagnostics on tuberculosis incidence and mortality: A modelling analysis.
PubMed 文献 · PLoS Med · 2020-12-11

该建模分析聚焦南非,模拟当前和未来尿液-LAM检测对结核病发病与死亡的影响。结果显示,未来LAM检测若广泛用于所有有结核症状的就诊者,2020-2035年可减少18%发病和30%死亡,主要归因于诊断尚未进入HIV关怀的HIV感染者。

要点

  • 当前LAM检测按WHO指南用于HIV关怀者,对住院结核死亡影响有限。
  • 未来高灵敏度LAM检测扩展至所有结核症状就诊者,可大幅减少发病和死亡。
  • 扩大LAM检测覆盖至未进入HIV关怀的HIV感染者是提升群体影响的关键。

核心结论

  • LAM检测可能对晚期HIV感染者的结核死亡有重要减少作用。
  • 若要在高HIV负担地区实现群体层面影响,未来LAM检测需性能足够并广泛部署。

研究设计与样本

  • 构建结核传播数学模型,聚焦南非,区分当前LAM(灵敏度与LF-LAM一致)和未来LAM(灵敏度与SILVAMP-LAM一致)。
  • 模拟完全采纳2019 WHO指南在HIV关怀者中使用LAM,以及未来LAM用于所有结核症状就诊者的情景。

主要发现

  • 维持现状下,2020-2035年预计累计结核发病2,700,000例(95% CrI 2,000,000-3,600,000),死亡420,000例(95% CrI 350,000-520,000)。
  • 当前和未来LAM在住院患者中分别减少54例(95% CrI 33-86)和90例(95% CrI 55-145)结核死亡,即住院结核死亡率降低5%(95% CrI 4%-6%)和9%(95% CrI 7%-11%)。
  • 扩展至门诊患者时绝对死亡减少翻倍,但仍不足全国结核死亡的1%。
  • 未来LAM用于所有结核症状就诊者,可减少470,000例(95% CrI 220,000-870,000)发病(18%降低,95% CrI 9%-29%)和120,000例(95% CrI 69,000-210,000)死亡(30%降低,95% CrI 18%-44%)。

临床怎么用

  • 当前LAM检测应优先用于HIV关怀中的住院或重症HIV感染者。
  • 未来高灵敏度LAM检测可考虑扩展至所有结核症状就诊者,包括未进入HIV关怀的HIV感染者。

局限性

  • 未分析未来检测的成本或成本效果。
  • 模型忽略耐药,且仅聚焦国家层面流行,忽略HIV和结核负担的次国家差异。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

BACKGROUND: Lateral flow urine lipoarabinomannan (LAM) tests could offer important new opportunities for the early detection of tuberculosis (TB). The currently licensed LAM test, Alere Determine TB LAM Ag ('LF-LAM'), performs best in the sickest people living with HIV (PLHIV). However, the technology continues to improve, with newer LAM tests, such as Fujifilm SILVAMP TB LAM ('SILVAMP-LAM') showing improved sensitivity, including amongst HIV-negative patients. It is important to anticipate the epidemiological impact that current and future LAM tests may have on TB incidence and mortality. METHODS AND FINDINGS: Concentrating on South Africa, we examined the impact that widening LAM test eligibility would have on TB incidence and mortality. We developed a mathematical model of TB transmission to project the impact of LAM tests, distinguishing 'current' tests (with sensitivity consistent with LF-LAM), from hypothetical 'future' tests (having sensitivity consistent with SILVAMP-LAM). We modelled the impact of both tests, assuming full adoption of the 2019 WHO guidelines for the use of these tests amongst those receiving HIV care. We also simulated the hypothetical deployment of future LAM tests for all people presenting to care with TB symptoms, not restricted to PLHIV. Our model projects that 2,700,000 (95% credible interval [CrI] 2,000,000-3,600,000) and 420,000 (95% CrI 350,000-520,000) cumulative TB incident cases and deaths, respectively, would occur between 2020 and 2035 if the status quo is maintained. Relative to this comparator, current and future LAM tests would respectively avert 54 (95% CrI 33-86) and 90 (95% CrI 55-145) TB deaths amongst inpatients between 2020 and 2035, i.e., reductions of 5% (95% CrI 4%-6%) and 9% (95% CrI 7%-11%) in inpatient TB mortality. This impact in absolute deaths averted doubles if testing is expanded to include outpatients, yet remains <1% of country-level TB deaths. Similar patterns apply to incidence results. However, deploying a future LAM test for all people presenting to care with TB symptoms would avert 470,000 (95% CrI 220,000-870,000) incident TB cases (18% reduction, 95% CrI 9%-29%) and 120,000 (95% CrI 69,000-210,000) deaths (30% reduction, 95% CrI 18%-44%) between 2020 and 2035. Notably, this increase in impact arises largely from diagnosis of TB amongst those with HIV who are not yet in HIV care, and who would thus be ineligible for a LAM test under current guidelines. Qualitatively similar results apply under an alternative comparator assuming expanded use of GeneXpert MTB/RIF ('Xpert') for TB diagnosis. Sensitivity analysis demonstrates qualitatively similar results in a setting like Kenya, which also has a generalised HIV epidemic, but a lower burden of HIV/TB coinfection. Amongst limitations of this analysis, we do not address the cost or cost-effectiveness of future tests. Our model neglects drug resistance and focuses on the country-level epidemic, thus ignoring subnational variations in HIV and TB burden. CONCLUSIONS: These results suggest that LAM tests could have an important effect in averting TB deaths amongst PLHIV with advanced disease. However, achieving population-level impact on the TB epidemic, even in high-HIV-burden settings, will require future LAM tests to have sufficient performance to be deployed more broadly than in HIV care.

来源:PLoS Med,查看原文。版权归原作者所有。

引用

  1. Ricks S, Denkinger CM, Schumacher SG, et al. The potential impact of urine-LAM diagnostics on tuberculosis incidence and mortality: A modelling analysis. PLoS Med. 2020. PMID: 33306694. doi:10.1371/journal.pmed.1003466

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。