Mathematics Institute, University of Warwick, Zeeman Building, Coventry, UK.
School of Life Sciences, University of Warwick, Gibbet Hill Campus, Coventry, UK.
Trans R Soc Trop Med Hyg. 2018 Jul 1;112(7):342-348. doi: 10.1093/trstmh/try062.
When tests are used in series to determine individual risk factors and infection status in a mass drug administration (MDA), the diagnostics, test order and subsequent treatment decisions (the testing algorithm) affect population-level treatment coverage and cost, but there is no existing framework for evaluating which algorithm optimizes any given outcome.
We present a mathematical tool (with spreadsheet implementation) to analyse the effect of test ordering, illustrated using treatment for onchocerciasis in an area where high-burden Loa loa co-infections present a known risk factor.
The prevalence of the infection and risk factor have a non-linear impact on the optimal ordering of tests. Testing for the MDA infection first always leaves more infected people untreated but fewer people with the risk factor being misclassified. The cost of the treatment given to infected individuals with the risk factor does not affect which algorithm is more cost effective.
For a given test and treat algorithm and its costs, the correct strategy depends on the expected prevalence. In most cases, when the apparent prevalence of the target infection is greater than the apparent prevalence of the risk factor, it is cheaper to do the risk factor test first, and vice versa.
当在大规模药物治疗（MDA）中使用串联测试来确定个体风险因素和感染状况时，诊断方法、测试顺序以及随后的治疗决策（测试算法）会影响人群的治疗覆盖率和成本，但目前还没有评估哪种算法可以优化任何给定结果的现有框架。
我们提出了一种数学工具（带有电子表格实现）来分析测试顺序的影响，使用在 Loa loa 高负担感染作为已知风险因素的地区治疗盘尾丝虫病进行了说明。
感染和风险因素的流行率对测试的最佳排序有非线性影响。首先对 MDA 感染进行检测始终会使更多受感染的人未得到治疗，但将具有风险因素的人错误分类的人数更少。给予具有风险因素的感染个体的治疗成本不会影响哪种算法更具成本效益。
对于给定的测试和治疗算法及其成本，正确的策略取决于预期的流行率。在大多数情况下，当目标感染的明显流行率大于风险因素的明显流行率时，先进行风险因素测试更便宜，反之亦然。
