Tests to screen for certain diseases—for example, thyroid cancer screening, screening mammography, and screening of high blood pressure for hypertension—are increasingly common in medical practice. However, guidelines for routine screening are contentious for many disorders and often fluctuate over time. Some tests are over‐ or underused compared to available evidence that justifies their use, with clinical practice persistently deviating from evidence‐based guidelines. Here we develop an integrated, broad boundary feedback theory and formal model to explain the dynamics of routine population screening including fluctuations in policy‐decision thresholds and the expansion of selection criteria which may lead to inappropriate use. We present a behaviorally realistic, boundedly rational model of detection and selection for medical screening that explains the potential of endogenous oscillations in practice guidelines as decision‐makers—including epidemiologists, clinicians, and patients, or policymakers from guideline issuing organizations, perceive harms and benefits from potential outcomes and make trade‐offs between sensitivity and specificity by altering the existing guidelines and actual practice. The model endogenously generates fluctuations in screening indications, test thresholds, test efficiency, and the target screening population, leading to long periods during which practice guidelines are suboptimal even if the underlying evidence base is constant. We use cancer screening as a motivating example, but the model is generic with a wide range of potential applications for important managerial problems in medical contexts, such as screening for hypertension, hypercholesterolemia, autism spectrum disorder, Alzheimer's disease, and related dementia. It also applies to other managerial problems in nonmedical contexts, such as airport screening, background checks, tax audits, automotive emission tests, contentious jurisdiction, or to consumers of other kinds of information who need to make a decision—on behalf of an individual, or for the whole population. © 2020 System Dynamics Society

中文翻译：

---

“挽救生命或损害健康？” 常规医学筛查中的过度使用和波动

在医学实践中，用于筛查某些疾病的测试（例如，甲状腺癌筛查，乳腺钼靶筛查和高血压高血压筛查）越来越普遍。但是，常规筛查指南对许多疾病都有争议，并且通常会随时间波动。与可证明其使用合理的可用证据相比，有些测试被过度使用或使用不足，并且临床实践始终偏离基于证据的准则。在这里，我们开发了一个集成的，广泛的边界反馈理论和形式模型，以解释常规人群筛查的动态，包括政策决策阈值的波动和选择标准的扩展（可能会导致使用不当）。我们提出一种行为现实的方法，用于医学筛查的有限合理的检测和选择模型，解释了实践指南中决策者（包括流行病学家，临床医生和患者，或准则发布组织的政策制定者）​​内源性振荡的可能性，他们意识到潜在结果的危害和益处并进行贸易-通过更改现有准则和实际做法，在敏感性和特异性之间进行权衡。该模型会内生地产生筛选指示，测试阈值，测试效率和目标筛选人群的波动，从而导致即使指导依据是恒定的，实践指南也不理想。我们以癌症筛查为例，但是该模型具有通用性，在医学领域中对重要管理问题具有广泛的潜在应用，例如筛查高血压，高胆固醇血症，自闭症谱系障碍，阿尔茨海默氏病和相关的痴呆。它还适用于非医疗环境中的其他管理问题，例如机场检查，背景检查，税务审计，汽车排放测试，有争议的管辖权，或需要代表个人做出决定的其他类型信息的消费者，或针对整个人口 ©2020系统动力学学会 背景调查，税收审核，汽车尾气排放测试，有争议的管辖权，或需要个人（或代表整个人群）做出决定的其他类型信息的消费者。©2020系统动力学学会 背景调查，税收审核，汽车尾气排放测试，有争议的管辖权，或需要个人（或代表整个人群）做出决定的其他类型信息的消费者。©2020系统动力学学会