AIDS 2024丨WHO推荐的HIV连续三次检测程序可显著提高诊断的准确率

图片
编者按:2023年,世界卫生组织(WHO)的建议,低负担国家采用连续三次检测来明确HIV感染阳性诊断。WHO建议HIV感染率低于5%的国家采用连续三次检测来明确HIV阳性诊断。人群中HIV检测的阳性比例会影响诊断的正确率,当HIV检出阳性率低于5%时,则至少需要三次连续检测才能保证99%的诊断正确率。
鉴于此,马拉维修订了其HIV检测指南,成为全球首个采用并过渡到新的HIV连续三次检测的国家,旨在降低因阳性率下降而导致的假阳性诊断风险。这一转变伴随着国家HIV检测服务(HTS)程序的更新,引入“3-test HIV testing algorithm”,即进行三项连续的快速诊断测试(Determine HIV1/2、Uni-Gold HIV1/2和SD Bioline HIV1/2),并同时采用了先进的监测与评估(M&E)工具,以获取关键的HIV检测结果。
图片
该全国性分阶段倡议由马拉维卫生部协调,自2022年11月开始实施。截至2024年1月23日,该程序已成功在全部28个地区的545个设施中启用,覆盖了78%的检测点。马拉维国家卫生质量监督和基于ScanForm技术的每日自动化数据质量评估确保了数据的准确性、算法一致性,在推动各检测点持续质量改进方面发挥了关键作用。在近期举行的第25届国际艾滋病大会(AIDS2024)上,马拉维卫生部Tiwonge Chimpandule教授报告了该项目的最新进展。 
研究人员从2022年11月1日至2024年1月23日,共收集了3,082,406份HIV检测记录。在总检测人数中,“3-test HIV testing algorithm”的一致率为99.96%(2,605,324 / 2,606,280)。其中,371名个体(239名女性和132名男性)在初次和第二次测试均呈阳性,在第三次测试后获得了不确定的结果。而根据之前的两次检测的算法,这些个体很可能被误诊为HIV阳性并立即开始抗逆转录病毒治疗。此外,还有448名个体(281名女性和167名男性)被诊断为HIV阴性,而根据之前的两次检测的算法,他们则需要在两周后进行复检。
这项研究的结果强调了采用WHO推荐的连续三次检测对于减少误诊的重要性。马拉维的经验表明,通过有效的提供者培训和诸如ScanForm等强大的M&E工具,可以迅速实现该算法的准确实施。预计全面过渡将在2024年6月完成,并将持续提供现场支持以确保持续的成功。
该研究的临床意义显著,不仅为其他国家提供了关于如何有效实施HIV检测的实际案例,还强调了采用先进技术进行数据监测和评估的重要性。通过这些措施,马拉维不仅提高了HIV检测的准确性,还减少了因误诊而导致的不必要治疗和潜在的健康风险。这一成功经验有望在全球范围内推广,以进一步优化HIV检测策略并提高患者护理质量。
专家访谈
01
《感染医线》:在您的研究中,371名个体因连续三次检测的算法获得了不确定的结果,这在之前的两次检测的算法下可能会被误诊为HIV阳性。这一发现对于HIV的诊断策略有何具体影响?
Chimpandule教授:非常感谢您的邀请。我来自马拉维卫,它位于非洲南部,是首批采用世卫组织新推出的HIV连续三次检测的国家之一。从我们的实施情况来看,我们从2022年11月开始从两次检测算法过渡到连续三次检测的算法。截至7月份,我们已经实施了21个月。在这21个月内,我们检测了约400万人,发现其中有556人避免了误诊。这意味着,在之前的两次检测的算法下,这些人在检测1呈阳性且检测2呈反应性后,会在第一天就开始接受抗逆转录病毒治疗。而增加了第三次检测后,他们中的大多数人接受了进一步的检测,并得出了阴性的结果。基于该项研究我们深刻认识到,采用WHO实施的连续三次检对个人的生活产生了实际而积极的影响。可以看到有556人避免了错误的阳性诊断。我们为此感到非常高兴。
IIDF: The study found that 371 individuals received inconclusive results under the 3-test algorithm, which would have likely led to misdiagnosis as HIV positive under the previous 2-test algorithm. How does this finding specifically impact HIV diagnostic strategy?
Dr Chimpandule: Thank you for having me. As you said, my name is Tiwonge Chimpandule from Malawi Ministry of Health. Malawi is in Southern Africa, and is actually one of the first countries that has adopted the new three-test algorithm from the WHO, which requires that countries with <5% HIV prevalence transition from the previous two-test algorithm to the three-test algorithm. The three-test algorithm basically just says that for an individual to conclude they are HIV-positive, they need to be tested for HIV using test 1, test 2 and test 3 to confirm that are HIV-positive. From our implementation, we started our transition from the two-test algorithm to the three-test algorithm in November, 2022. Right now, in July, we are at 21 months of implementation. What we have seen is that for around 4 million individuals that we have tested within this 21 months, 556 individuals have been saved from misdiagnosis. What this means is that from the previous two-test algorithm, these individuals would have been started on ART on the first day after testing positive for test 1 and reactive for test 2. With the addition of the third test, most of them went for further testing and turned out to be negative. What we are learning is that adopting the three-test algorithm that has been implemented by the WHO has practical impacts on individual lives. As you can imagine, the 556 individuals is quite a lot that have been saved from a positive diagnosis. So we are happy after adopting, we have actually managed to make an impact on these 500-plus individuals who would otherwise have started on lifelong treatment. We are really happy.
02
《感染医线》:ScanForm技术在数据收集和质量评估中扮演了怎样的角色?它的引入对研究结果的准确性有何影响?
吴博士:非常感谢。此次部署中使用的ScanForm技术是由我们团队研发的。总的来说,马拉维近2500个检测点进行的这次全国性部署工作,得益于ScanForm技术的支持。这项技术非常简单,只需普通的纸张和笔。用户只需填好表格,然后用手机拍照即可提交。我们利用人工智能技术,自动提取并分析全国范围内登记册中的所有数据。这项技术对部署工作产生了深远影响,因为在非洲许多设施都面临电力供应不稳定的问题。其中一些设施在屋顶、水源和其他基本设施方面都存在困难,因此在每个检测点都部署计算机显得非常不切实际。而ScanForm技术的出现,使得只需在每个诊所配备一两部手机,就能在马拉维全国范围内实际实施“3-test HIV testing algorithm”。此外,它的成本也非常低廉,同时我们还能确保人们能够准确地遵循该检测算法,利用人工智能系统就能立即发现错误,并提示他们进行更正。得益于这种自动的、持续的质量改进机制,我们看到的算法正确实施率高达99.96%。这是一种既实用又成本低廉的大规模部署的检测方式,其他国家也可以借鉴并使用。
图片
IIDF: How did ScanForm technology play a role in data collection and quality assessment, and what impact did its introduction have on the accuracy of the study results?
Dr Wu: Thank you very much. My name is William Wu, and I am Chief Executive of QED. We are making the ScanForm technology that was used in this deployment. To summarize, all the data for this national scale deployment, across almost 2500 testing points in Malawi, was deployed using ScanForm. This is regular paper, regular pen. You fill out the forms and take a picture with an Android phone. We use artificial intelligence to automatically extract and automatically analyze all the data in that register at the national scale. It has impacted the deployment because there are many facilities that do not have stable electricity. Some of them struggle with roofs, and water, and other basic things, so it can be very difficult to deploy computers at every testing point. This makes it practical to implement the three-test algorithm on a national scale by simply using pen and paper, which everyone has, and only one or two phones per clinic. Another thing is the cost is very inexpensive. The cost of printing these registers is cheap, and we can also make sure that people follow this three-test algorithm accurately. Anytime they deviate from the protocol, when we take a picture, the AI knows they have made a mistake and sends them an error message saying please correct it. Thanks to this automatic, continuous quality improvement, we see this 99.96% correct implementation of the algorithm. It is a very practical and inexpensive way to deploy this algorithm at scale that other countries can also use.
03
《感染医线》:未来,您计划如何进一步改进HIV测试算法或监测与评估工具,以提高HIV诊断的准确性和效率?
Chimpandule教授:展望未来,我们将密切关注HIV诊断方面的最新证据,探索是否有适合我们采用的新方法。当然,在采纳任何新方法之前,我们都会进行试点,以确保其在我们的环境中切实可行。在这项研究中,ScanForm发挥了重要作用。它不仅帮助我们实现了记录的数字化,还助力我们顺利过渡到“3-test HIV testing algorithm”。基于这一成功经验,我们相信ScanForm的应用可以扩展到其他项目。目前,我们正考虑将其应用于性传播感染项目,因为尽管我们拥有大量数据,但目前仍以纸质形式存在,未能实现数字化。而缺乏ScanForm提供的登记册功能,也导致我们在某些决策点上存在不足。因此,展望未来,我们计划利用ScanForm来收集和处理实验室表格中的实时个性化数据,即使这些表格目前仍以纸质形式存在于各设施中。
吴博士:这实际上是历经大约15年的研发成果,也是我们今天最为自豪的项目,其影响力也最为深远。我们由衷地感谢马拉维艾滋病毒和艾滋病部门,他们具备远见卓识,允许我们在全国范围内进行探索,并真正挖掘出这项技术的潜力。在此,我要特别感谢各方的大力支持和共同努力,才能让这一技术应用到临床,并推动HIV检测技术的不断进步。
IIDF: Moving forward, how do you plan to further improve HIV testing algorithms or monitoring and evaluation tools to enhance the accuracy and efficiency of HIV diagnosis?
Dr Chimpandule: Thank you very much. I think there are two things. Number one is, as I said, we are one of the first countries to have adopted the three-test algorithm in sub-Saharan Africa. We looked at the evidence that was issued by the WHO, and then of course piloted within our setting. Based on the adoption of the three-test algorithm, I can say we managed to save many individuals from starting ART. Moving forward, I think we will be on the look out to see what evidence is there on issues of HIV diagnosis, and see if there are things that we can adopt. Of course, before adopting, you need to pilot, and then see if the program can be implemented within our context. Then you asked about monitoring and evaluation tools. As we have said, ScanForm has really helped us in ensuring we have digitized records, which is very important for individualized data, but also in trying to make sure we have a quality transition to the three-test algorithm. So in this case, our experience suggests we can expand ScanForm to other programs. We are thinking about expanding ScanForm to a sexually transmitted infections program, because right now we are paper-based and not digitized but we have lots of data that we want to be ScanFormed. There are decision points we are failing to make because we don’t have the registries that ScanForm can provide. Moving forward, even for laboratory forms, some are collected in facilities in paper registers, but we could use ScanForm to have real-time individualized data from the various lab forms. That is the plan that the Ministry is trying to look into based on the lessons we have learned so far.
Dr Wu: I wanted to say that this was actually about 15 years of work in actual development. This is the project we are most proud of today. It has the most impact. We really want to thank the Malawi Department of HIV and AIDS. They really believed in this technology. They saw the vision. They allowed the technology to explore on a national scale and really make use of its potential. I just want to express my thanks to our sponsors, Global Fund, and the Malawi Department of HIV for making this such a wonderful experience.
图片