Division of Nephrology, Department of Internal Medicine, Yongin Severance Hospital, Yonsei University College of Medicine, Seoul, South Korea.
Graduate School of AI, College of Engineering, Korea Advanced Institute of Science and Technology (KAIST), Daejeon, South Korea.
Comput Biol Med. 2021 Oct;137:104718. doi: 10.1016/j.compbiomed.2021.104718. Epub 2021 Jul 31.
In patients with kidney failure with replacement therapy (KFRT), optimizing anemia management in these patients is a challenging problem because of the complexities of the underlying diseases and heterogeneous responses to erythropoiesis-stimulating agents (ESAs). Therefore, we propose a ESA dose recommendation model based on sequential awareness neural networks. Data from 466 KFRT patients (12,907 dialysis sessions) in seven tertiary-care general hospitals were included in the experiment. First, a Hb prediction model was developed to simulate longitudinal heterogeneous ESA and Hb interactions. Based on the prediction model as a prospective study simulator, we built an ESA dose recommendation model to predict the required amount of ESA dose to reach a target hemoglobin level after 30 days. Each model's performance was evaluated in the mean absolute error (MAE). The MAEs presenting the best results of the prediction and recommendation model were 0.59 (95% confidence interval: 0.56-0.62) g/dL and 43.2 μg (ESAs dose), respectively. Compared to the results in the real-world clinical data, the recommendation model achieved a reduction of ESA dose (Algorithm: 140 vs. Human: 150 μg/month, P < 0.001), a more stable monthly Hb difference (Algorithm: 0.6 vs. Human: 0.8 g/dL, P < 0.001), and an improved target Hb success rate (Algorithm: 79.5% vs. Human: 62.9% for previous month's Hb < 10.0 g/dL; Algorithm: 95.7% vs. Human:73.0% for previous month's Hb 10.0-12.0 g/dL). We developed an ESA dose recommendation model for optimizing anemia management in patients with KFRT and showed its potential effectiveness in a simulated prospective study.
在接受肾脏替代治疗（KFRT）的肾衰竭患者中，优化这些患者的贫血管理是一个具有挑战性的问题，因为潜在疾病的复杂性和对红细胞生成刺激剂（ESA）的异质性反应。因此，我们提出了一种基于序贯感知神经网络的 ESA 剂量推荐模型。该实验纳入了来自七家三级综合医院的 466 名 KFRT 患者（12907 次透析）的数据。首先，建立了一个 Hb 预测模型，以模拟纵向异质性 ESA 和 Hb 相互作用。基于预测模型作为前瞻性研究模拟器，我们构建了一个 ESA 剂量推荐模型，以预测 30 天后达到目标血红蛋白水平所需的 ESA 剂量。每个模型的性能都以平均绝对误差（MAE）进行评估。预测和推荐模型表现最好的 MAE 分别为 0.59（95%置信区间：0.56-0.62）g/dL 和 43.2μg（ESA 剂量）。与真实世界临床数据的结果相比，推荐模型实现了 ESA 剂量的减少（算法：140 对人类：150μg/月，P<0.001），每月 Hb 差异更稳定（算法：0.6 对人类：0.8g/dL，P<0.001），以及提高了目标 Hb 成功率（算法：前一个月 Hb<10.0g/dL 时为 79.5%，人类为 62.9%；算法：前一个月 Hb 为 10.0-12.0g/dL 时为 95.7%，人类为 73.0%）。我们开发了一种 ESA 剂量推荐模型，用于优化 KFRT 患者的贫血管理，并在模拟前瞻性研究中显示了其潜在有效性。
