Dear Editor, We read your recent article Optimal timing of HIV homebased counselling and testing rounds in Western Kenya [1] with considerable interest. The authors make questionable assumptions about their modelled reference campaign in which 90% of the population is reached, of these 60% of those contacted are linked to care, and of these only 75% receive the CD4 count test required to access ART. Of people on ART, 86% are virally suppressed, and of these 95.5% are retained at one year and only 88% after four years. Therefore, the proportion of people living with HIV (PLHIV) in community who are virally suppressed for the first year can be derived by multiplying the parameters: diagnosed (90%), linked (60%), lab tested (75%), virally suppressed (86%) and retained at one year (95.5%). Using these parameters, the denominator rapidly decreases and the percentage of HIV-positive people that are virally suppressed after a year is about 30% and it declines thereafter. Using these parameters results in an ineffective and inefficient treatment programme at a substantial cost. The article then asks what the optimal timing would be if it were to be repeated? Given the poorly performing programme, one answer is that it should first be significantly improved before considering the question of how often it should be repeated. There are many examples of well-performing programmes and successful interventions. For example, the 2008 privatepublic sector collaboration involving the Kenya Ministry of Health, Centers for Disease Control, CHF International, and Vestergaard Frandsen showed that an effective communitybased multi-disease prevention campaign in Western Kenya was able to reach and test 87% of its target population over a seven days period [2]. By offering long-lasting insecticide-treated bed nets, water filters and rapid HIV testing they reached 47,311 people with a 96% uptake of the multi-disease prevention package with 99.7% accepting HIV testing including 18,101 (38%) men. Of participants, 80% had never been tested and 4% were diagnosed with HIV. The campaign also reached people with higher CD4 cell counts probably earlier in their HIV disease [2,3]. The more recent SEARCH study in Uganda and Kenya has also reached very high proportions of the community with testing of 131,307 (89%) of and successful treatment with minimal loss to follow-up [4,5]. These programmes and recent results from population-based incidence studies offer further evidence that expanding access to welldesigned and implemented HIV services can achieve remarkable impact [6-10]. On closer examination the model in the article examines the utility of repeating a failing home-based counselling and testing (HBCT) programme when compared to more successful models. HBCT programmes that reach 90% of its intended target are expected, but a programme that can only put 60% on treatment clearly has major problems that require urgent redress. The subsequent 55% losses due to unnecessary waits for CD4 cell counts and other service delivery problems further reduce the overall percentage on treatment and suppressed. Perhaps most worrisome about the modelled programme is the assumption that no one who is lost to follow-up returns. Despite considerable and effort to reach people in their homes, the combined result is that the model projects that only around 30% of PLHIV are successfully diagnosed, treated and virally suppressed. Additionally, the cost of treatment is $367 which is very expensive given the poor performance and the current annual costs of around $75 per year for ARVs alone. In many programmes these avoidable high costs and low performance measures would be a sign of failure and indicate the need for significant revision in service delivery to prevent avoidable illnesses, deaths and HIV infections. Specifically, interventions to improve performance that other successful programmes have used include same-day treatment without delays for unnecessary laboratories, immediate social support for patients starting treatment, routine use of viral load for monitoring success and motivating patients, a highly Granich R et al. Journal of the International AIDS Society 2019, 22:e25230 http://onlinelibrary.wiley.com/doi/10.1002/jia2.25230/full | https://doi.org/10.1002/jia2.25230

中文翻译：

---

肯尼亚家庭测试活动建模受到质疑的假设

亲爱的编辑， 我们怀着极大的兴趣阅读了您最近的文章《肯尼亚西部艾滋病毒家庭咨询和检测轮次的最佳时机》[1]。作者对他们的建模参考活动做出了可疑的假设，其中 90% 的人口得到了接触，其中 60% 的接触者与护理有关，而其中只有 75% 的人接受了获得 ART 所需的 CD4 计数测试。在接受 ART 治疗的人中，86% 的病毒得到抑制，其中 95.5% 在一年后得到抑制，而四年后只有 88% 的病毒得到抑制。因此，社区中第一年病毒受到抑制的艾滋病毒感染者 (PLHIV) 的比例可以通过以下参数相乘得出：诊断 (90%)、关联 (60%)、实验室检测 (75%)、病毒检测抑制（86%）并保留一年（95.5%）。使用这些参数，分母迅速下降，一年后病毒受到抑制的 HIV 阳性人群的百分比约为 30%，此后又下降。使用这些参数会导致无效且低效的治疗方案，并且成本高昂。文章接着问，如果重复的话，最佳时机是什么？鉴于该计划表现不佳，一个答案是，首先应对其进行显着改进，然后再考虑应该重复执行的频率问题。有许多表现良好的计划和成功干预措施的例子。例如，2008 年涉及肯尼亚卫生部、疾病控制中心、CHF International 和 Vestergaard Frandsen 的私营公共部门合作表明，肯尼亚西部有效的基于社区的多种疾病预防运动能够达到并测试其目标的 87% 7 天期间的人口数量 [2]。通过提供长效杀虫剂处理的蚊帐、水过滤器和快速艾滋病毒检测，他们覆盖了 47,311 人，其中 96% 的人接受了多种疾病预防方案，其中 99.7% 的人接受了艾滋病毒检测，其中包括 18,101 名男性（38%）。在参与者中，80% 的人从未接受过检测，4% 的人被诊断出感染了艾滋病毒。该活动还惠及了 CD4 细胞计数较高且可能处于 HIV 疾病早期的人群 [2,3]。最近在乌干达和肯尼亚进行的 SEARCH 研究也达到了很高的社区比例，测试了 131,307 人（89%），治疗成功，随访损失最小[4,5]。这些计划和基于人群的发病率研究的最新结果提供了进一步的证据，表明扩大获得精心设计和实施的艾滋病毒服务的机会可以取得显着的影响[6-10]。经过仔细检查，本文中的模型检验了与更成功的模型相比，重复失败的家庭咨询和测试 (HBCT) 计划的效用。HBCT 计划有望达到预期目标的 90%，但只能将 60% 用于治疗的计划显然存在需要紧急纠正的重大问题。随后由于不必要的 CD4 细胞计数等待和其他服务提供问题而造成 55% 的损失，进一步降低了治疗的总体百分比并受到抑制。也许对模型计划最令人担忧的是假设没有人会因为后续行动而失败而返回。尽管付出了巨大的努力来接触人们的家中，但综合结果是，该模型预计只有约 30% 的艾滋病毒感染者得到成功诊断、治疗和病毒抑制。此外，治疗费用为 367 美元，考虑到效果不佳且目前仅抗逆转录病毒药物每年的费用约为 75 美元，这一费用非常昂贵。在许多计划中，这些本可避免的高成本和低绩效措施将是失败的迹象，并表明需要对服务提供进行重大调整，以预防可避免的疾病、死亡和艾滋病毒感染。具体来说，其他成功项目所采用的提高绩效的干预措施包括当天治疗，避免不必要的实验室延误，对开始治疗的患者立即提供社会支持，常规使用病毒载量来监测成功和激励患者，这是 Granich R 等人高度评价的。国际艾滋病协会杂志 2019，22：e25230 http://onlinelibrary.wiley.com/doi/10.1002/jia2.25230/full | https://doi.org/10.1002/jia2.25230