Evaluating algorithmic fairness in the presence of clinical guidelines: the case of atherosclerotic cardiovascular disease risk estimation.
Evaluating algorithmic fairness in the presence of clinical guidelines: the case of atherosclerotic cardiovascular disease risk estimation.
复制标题
DOI：
10.1136/bmjhci-2021-100460
复制
发表时间：
2022-04
影响因子：
4.1
通讯作者：
Shah N
中科院分区：
文献类型：
--
作者：
Foryciarz A;Pfohl SR;Patel B;Shah N
美国心脏病学会和美国心脏协会关于动脉粥样硬化性心血管疾病（ASCVD）一级预防的指南建议使用10年ASCVD风险估计模型来启动他汀类药物治疗。对于与指南一致的决策，需要校准风险估计。然而，现有的模型往往是错误的种族，民族和性别为基础的亚组。本研究评估了两种算法公平性方法，以调整风险估计值（组重新校准和均衡赔率）与支持指南决策规则的假设的兼容性。使用更新的汇总队列数据集，我们推导出无约束，组重新校准和均衡的赔率约束版本的10年ASCVD风险估计，并比较其校准在指南一致的决策阈值。我们发现，与无约束模型相比，组重新校准提高了每个组的相关阈值之一的校准，但加剧了组间假阳性和假阴性率的差异。一个均衡的几率约束，意味着均衡各组的错误率，通过整体和相关决策阈值的模型错误校准来实现。因此，由于诱导的误校准，指导决策的风险估计学习与均衡的赔率公平约束是不一致的，与现有的准则。相反，为每个组单独重新校准模型可以增加指南的兼容性，同时增加错误率的组间差异。因此，当指南建议以固定的决策阈值治疗时，各组之间错误率的比较可能会产生误导。在满足公平标准和保持准则兼容性之间的权衡，强调需要在下游干预的背景下评估模型。
The American College of Cardiology and the American Heart Association guidelines on primary prevention of atherosclerotic cardiovascular disease (ASCVD) recommend using 10-year ASCVD risk estimation models to initiate statin treatment. For guideline-concordant decision-making, risk estimates need to be calibrated. However, existing models are often miscalibrated for race, ethnicity and sex based subgroups. This study evaluates two algorithmic fairness approaches to adjust the risk estimators (group recalibration and equalised odds) for their compatibility with the assumptions underpinning the guidelines’ decision rules. MethodsUsing an updated pooled cohorts data set, we derive unconstrained, group-recalibrated and equalised odds-constrained versions of the 10-year ASCVD risk estimators, and compare their calibration at guideline-concordant decision thresholds. We find that, compared with the unconstrained model, group-recalibration improves calibration at one of the relevant thresholds for each group, but exacerbates differences in false positive and false negative rates between groups. An equalised odds constraint, meant to equalise error rates across groups, does so by miscalibrating the model overall and at relevant decision thresholds. Hence, because of induced miscalibration, decisions guided by risk estimators learned with an equalised odds fairness constraint are not concordant with existing guidelines. Conversely, recalibrating the model separately for each group can increase guideline compatibility, while increasing intergroup differences in error rates. As such, comparisons of error rates across groups can be misleading when guidelines recommend treating at fixed decision thresholds. The illustrated tradeoffs between satisfying a fairness criterion and retaining guideline compatibility underscore the need to evaluate models in the context of downstream interventions.
登录
查看更多内容
影响因子：
37.8
作者：
Churchwell, Keith;Elkind, Mitchell S. V.;Williams, Olajide
通讯作者：
Williams, Olajide
影响因子：
39.2
作者：
Cook, Nancy R.;Ridker, Paul M.
通讯作者：
Ridker, Paul M.
影响因子：
24
作者：
Lloyd-Jones, Donald M.;Braun, Lynne T.;Blumenthal, Roger S.
通讯作者：
Blumenthal, Roger S.
影响因子：
158.5
作者：
PAUKER, SG;KASSIRER, JP
通讯作者：
KASSIRER, JP
影响因子：
39.2
作者：
Yadlowsky, Steve;Hayward, Rodney A.;Basu, Sanjay
通讯作者：
Basu, Sanjay