背景:在低收入和中等收入国家(LMICs)中,冠状病毒病(COVID)的快速诊断测试(RDTs)用于指导治疗决策。然而,到目前为止,目前尚不清楚这种使用何时具有成本效益。现有的分析仅限于一组狭窄的国家和用途。这项研究的目的是评估COVIDRDT的成本效益,以告知LMIC中严重疾病患者的治疗,考虑现实世界的实践。
结果:我们使用决策树模型评估了不同LMIC的COVID测试的成本效益,按国家收入水平区分结果,严重急性呼吸系统综合症冠状病毒2(SARS-CoV-2)患病率,和测试场景(无,RDT,聚合酶链反应试验-PCR和组合)。LMIC专家定义了现实的护理途径和治疗方案。使用医疗保健提供者的观点和净货币收益方法,我们评估了每种测试方案的预期(COVID症状缓解)和非预期(治疗副作用)健康和经济影响.我们包括皮质类固醇的副作用,这通常是唯一可用的COVID治疗方法。因为副作用取决于治疗和患者的潜在疾病(COVID或COVID样疾病,如流感),我们在分析中考虑了COVID样疾病的患病率。我们发现,对患有严重COVID样疾病的患者进行SARS-CoV-2检测在所有LMIC中都具有成本效益,但只有在某些情况下。疑似COVID病例中的高流感患病率提高了成本效益,因为错误提供皮质类固醇可能会恶化流感结局。在低收入和一些中低收入国家,只有怀疑COVID指数高的患者才应该进行RDT测试,而其他患者应推定为没有COVID。在一些中低收入和中高收入国家,怀疑严重的COVID病例几乎都应该进行检测。Further,在这些设置中,初始怀疑指数较高的患者的阴性检测结果应通过PCR确认,在流感爆发期间,初始怀疑指数较低的患者的阳性结果也应通过PCR确认.使用白细胞介素-6受体阻滞剂,当测试支持时,在高收入的低收入国家中也可能具有成本效益。它们在低收入国家具有成本效益的成本(每个疗程162至406美元)低于当前价格。我们分析的主要局限性是由于有限的数据,我们模型中的一些参数存在很大的不确定性,最值得注意的是,在标准护理下,目前的COVID死亡率,以及糖皮质激素对重症流感患者影响的证据不足。
结论:COVID检测对于治疗患有严重COVID样疾病的LMIC患者具有成本效益。最优算法由国家收入水平和卫生预算驱动,怀疑患者可能患有COVID的程度,和流感流行。进一步的研究,以更好地表征皮质类固醇的非预期作用,特别是流感病例,可以改善LMIC患者在治疗COVID样症状患者方面的决策。
BACKGROUND: Rapid diagnostic tests (RDTs) for coronavirus disease (COVID) are used in low- and middle-income countries (LMICs) to inform treatment decisions. However, to date, it is unclear when this use is cost-effective. Existing analyses are limited to a narrow set of countries and uses. The aim of this study is to assess the cost-effectiveness of COVID RDTs to inform the treatment of patients with severe illness in LMICs, considering real world practice.
RESULTS: We assessed the cost-effectiveness of COVID testing across LMICs using a decision tree model, differentiating results by country income level, Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) prevalence, and testing scenario (none, RDTs, polymerase chain reaction tests-PCRs and combinations). LMIC experts defined realistic care pathways and treatment options. Using a healthcare provider perspective and net monetary benefit approach, we assessed both intended (COVID symptom alleviation) and unintended (treatment side effects) health and economic impacts for each testing scenario. We included the side effects of corticosteroids, which are often the only available treatment for COVID. Because side effects depend both on the treatment and the patient\'s underlying illness (COVID or COVID-like illnesses, such as influenza), we considered the prevalence of COVID-like illnesses in our analyses. We found that SARS-CoV-2 testing of patients with severe COVID-like illness can be cost-effective in all LMICs, though only in some circumstances. High influenza prevalence among suspected COVID cases improves cost-effectiveness, since incorrectly provided corticosteroids may worsen influenza outcomes. In low- and some lower-middle-income countries, only patients with a high index of suspicion for COVID should be tested with RDTs, while other patients should be presumed to not have COVID. In some lower-middle-income and upper-middle-income countries, suspected severe COVID cases should almost always be tested. Further, in these settings, negative test results in patients with a high initial index of suspicion should be confirmed through PCR and, during influenza outbreaks, positive results in patients with a low initial index of suspicion should also be confirmed with a PCR. The use of interleukin-6 receptor blockers, when supported by testing, may also be cost-effective in higher-income LMICs. The cost at which they would be cost-effective in low-income countries ($162 to $406 per treatment course) is below current prices. The primary limitation of our analysis is substantial uncertainty around some of the parameters in our model due to limited data, most notably on current COVID mortality with standard of care, and insufficient evidence on the impact of corticosteroids on patients with severe influenza.
CONCLUSIONS: COVID testing can be cost-effective to inform treatment of LMIC patients with severe COVID-like disease. The optimal algorithm is driven by country income level and health budgets, the level of suspicion that the patient may have COVID, and influenza prevalence. Further research to better characterize the unintended effects of corticosteroids, particularly on influenza cases, could improve decision making around the treatment of those with COVID-like symptoms in LMICs.