Fairness in the prediction of acute postoperative pain using machine learning models.
Fairness in the prediction of acute postoperative pain using machine learning models.
复制标题
使用机器学习模型预测急性术后疼痛的公平性。
DOI：
10.3389/fdgth.2022.970281
复制
发表时间：
2022
影响因子：
--
通讯作者：
Tighe, Patrick J
中科院分区：
文献类型：
--
作者：
Davoudi, Anis;Sajdeya, Ruba;Ison, Ron;Hagen, Jennifer;Rashidi, Parisa;Price, Catherine C;Tighe, Patrick J
基于机器学习的预测模型的整体性能是有前景的；然而，必须大力研究它们的普遍性和公平性，以确保它们对所有患者都表现良好。本研究旨在评估用于预测术后急性疼痛的机器学习模型的预测偏差。我们对 2011 年 6 月 1 日至 2019 年 6 月 30 日在佛罗里达大学卫生系统/尚兹医院接受骨科手术的患者的电子健康记录进行了回顾性审查。 CatBoost 机器学习模型经过训练，用于预测低痛 (≤4) 和高痛 (>4) 的二元结果。根据年龄、性别、种族、地区剥夺指数 (ADI)、口语、健康素养和保险类型等七个受保护属性评估模型偏差。研究了受保护属性的重新权衡，以减少与基本模型相比的模型偏差。对机会平等、预测平等、预测平等、统计平等和总体准确性平等的公平性指标进行了检查。最终数据集包括 14,263 名患者 [年龄：60.72 (16.03) 岁，53.87% 女性，39.13% 术后急性低痛]。机器学习模型（曲线下面积，0.71）在年龄、种族、ADI 和保险类型方面存在偏差，但在性别、语言和健康素养方面没有偏差。尽管在预测急性术后疼痛方面总体表现良好，但基于机器学习的预测模型可能在受保护的属性方面存在偏差。这些发现表明，在将机器学习模型用作临床决策支持工具之前，需要评估涉及围手术期疼痛的机器学习模型的公平性。
Overall performance of machine learning-based prediction models is promising; however, their generalizability and fairness must be vigorously investigated to ensure they perform sufficiently well for all patients. This study aimed to evaluate prediction bias in machine learning models used for predicting acute postoperative pain. We conducted a retrospective review of electronic health records for patients undergoing orthopedic surgery from June 1, 2011, to June 30, 2019, at the University of Florida Health system/Shands Hospital. CatBoost machine learning models were trained for predicting the binary outcome of low (≤4) and high pain (>4). Model biases were assessed against seven protected attributes of age, sex, race, area deprivation index (ADI), speaking language, health literacy, and insurance type. Reweighing of protected attributes was investigated for reducing model bias compared with base models. Fairness metrics of equal opportunity, predictive parity, predictive equality, statistical parity, and overall accuracy equality were examined. The final dataset included 14,263 patients [age: 60.72 (16.03) years, 53.87% female, 39.13% low acute postoperative pain]. The machine learning model (area under the curve, 0.71) was biased in terms of age, race, ADI, and insurance type, but not in terms of sex, language, and health literacy. Despite promising overall performance in predicting acute postoperative pain, machine learning-based prediction models may be biased with respect to protected attributes. These findings show the need to evaluate fairness in machine learning models involved in perioperative pain before they are implemented as clinical decision support tools.