BACKGROUND Tuberculosis preventive therapy (TPT) reduces TB-related morbidity and mortality in people living with HIV (PLHIV). Cascade-of-care analyses help identify gaps and barriers in care and develop targeted solutions. A previous latent tuberculosis infection (LTBI) cascade-of-care analysis showed only 18% of persons in at-risk populations complete TPT, but a similar analysis for TPT among PLHIV has not been completed. We conducted a meta-analysis to provide this evidence. METHODS AND FINDINGS We first screened potential articles from a LTBI cascade-of-care systematic review published in 2016. From this study, we included cohorts that reported a minimum of 25 PLHIV. To identify new cohorts, we used a similar search strategy restricted to PLHIV. The search was conducted in Medline, Embase, Health Star, and LILACS, from January 2014 to February 2021. Two authors independently screened titles and full text and assessed risk of bias using the Newcastle-Ottawa Scale for cohorts and Cochrane Risk of Bias for cluster randomized trials. We meta-analyzed the proportion of PLHIV completing each step of the LTBI cascade-of-care and estimated the cumulative proportion retained. These results were stratified based on cascades-of-care that used or did not use LTBI testing to determine eligibility for TPT. We also performed a narrative synthesis of enablers and barriers of the cascade-of-care identified at different steps of the cascade. A total of 71 cohorts were included, and 70 were meta-analyzed, comprising 94,011 PLHIV. Among the PLHIV included, 35.3% (33,139/94,011) were from the Americas and 29.2% (27,460/94,011) from Africa. Overall, 49.9% (46,903/94,011) from low- and middle-income countries, median age was 38.0 [interquartile range (IQR) 34.0;43.6], and 65.9% (46,328/70,297) were men, 43.6% (29,629/67,947) were treated with antiretroviral therapy (ART), and the median CD4 count was 390 cell/mm3 (IQR 312;458). Among the cohorts that did not use LTBI tests, the cumulative proportion of PLHIV starting and completing TPT were 40.9% (95% CI: 39.3% to 42.7%) and 33.2% (95% CI: 31.6% to 34.9%). Among cohorts that used LTBI tests, the cumulative proportions of PLHIV starting and completing TPT were 60.4% (95% CI: 58.1% to 62.6%) and 41.9% (95% CI:39.6% to 44.2%), respectively. Completion of TPT was not significantly different in high- compared to low- and middle-income countries. Regardless of LTBI test use, substantial losses in the cascade-of-care occurred before treatment initiation. The integration of HIV and TB care was considered an enabler of the cascade-of-care in multiple cohorts. Key limitations of this systematic review are the observational nature of the included studies, potential selection bias in the population selection, only 14 cohorts reported all steps of the cascade-of-care, and barriers/facilitators were not systematically reported in all cohorts. CONCLUSIONS Although substantial losses were seen in multiple stages of the cascade-of-care, the cumulative proportion of PLHIV completing TPT was higher than previously reported among other at-risk populations. The use of LTBI testing in PLHIV in low- and middle-income countries was associated with higher proportion of the cohorts initiating TPT and with similar rates of completion of TPT.

中文翻译：

---

HIV 感染者中潜伏的结核病级联护理：系统评价和荟萃分析。

背景结核病预防治疗 (TPT) 降低了 HIV 感染者 (PLHIV) 与结核病相关的发病率和死亡率。级联护理分析有助于确定护理中的差距和障碍，并制定有针对性的解决方案。先前的潜伏性结核病感染 (LTBI) 级联护理分析显示，只有 18% 的高危人群完成了 TPT，但尚未完成对 PLHIV 中 TPT 的类似分析。我们进行了一项荟萃分析来提供这一证据。方法和结果 我们首先筛选了 2016 年发表的 LTBI 级联护理系统评价中的潜在文章。从这项研究中，我们纳入了报告至少 25 名 PLHIV 的队列。为了识别新的队列，我们​​使用了一种仅限于 PLHIV 的类似搜索策略。搜索是在 Medline、Embase、Health Star 和 LILACS 进行的，2014 年 1 月至 2021 年 2 月。两位作者独立筛选标题和全文，并使用 Newcastle-Ottawa Scale for Cohorts 和 Cochrane Risk of Bias 评估偏倚风险用于整群随机试验。我们对完成 LTBI 级联护理每个步骤的 PLHIV 的比例进行了元分析，并估计了保留的累积比例。这些结果根据使用或不使用 LTBI 测试来确定 TPT 资格的级联护理进行分层。我们还对在级联不同步骤中确定的级联护理的促成因素和障碍进行了叙述性综合。共纳入了 71 个队列，对 70 个进行了荟萃分析，包括 94,011 名 PLHIV。在包括的 PLHIV 中，35.3% (33,139/94,011) 来自美洲，29.2% (27,460/94,011) 来自非洲。总体而言，49.9% (46,903/94, 011) 来自低收入和中等收入国家，中位年龄为 38.0 [四分位距 (IQR) 34.0;43.6]，65.9% (46,328/70,297) 为男性，43.6% (29,629/67,947) 接受抗逆转录病毒治疗ART)，CD4 计数中位数为 390 细胞/mm3 (IQR 312;458)。在未使用 LTBI 测试的队列中，PLHIV 开始和完成 TPT 的累积比例分别为 40.9%（95% CI：39.3% 至 42.7%）和 33.2%（95% CI：31.6% 至 34.9%）。在使用 LTBI 测试的队列中，PLHIV 开始和完成 TPT 的累积比例分别为 60.4%（95% CI：58.1% 至 62.6%）和 41.9%（95% CI：39.6% 至 44.2%）。与低收入和中等收入国家相比，高收入国家的 TPT 完成情况没有显着差异。无论使用何种 LTBI 测试，治疗开始前都发生了级联护理的重大损失。HIV 和 TB 护理的整合被认为是多个队列中级联护理的推动因素。本系统评价的主要局限性是纳入研究的观察性质、人群选择中的潜在选择偏倚、只有 14 个队列报告了级联护理的所有步骤，并且所有队列中都没有系统地报告障碍/促进因素。结论 尽管在级联护理的多个阶段都看到了大量损失，但在其他高危人群中，PLHIV 完成 TPT 的累积比例高于先前报告的比例。在低收入和中等收入国家，在 PLHIV 中使用 LTBI 检测与更高比例的发起 TPT 的队列和完成 TPT 的比例相似。