To the Editor—We read with interest the impressive study by Hong et al [1] that developed a new clinical prediction model of aminoglycoside (AG)-induced hearing loss among patients initiating drug-resistant tuberculosis (DR-TB) treatment. The results showed reasonable discrimination (area under curve [AUC] = 0.71) and calibration (χ²_[8] = 6.10; P = .636) in the development cohort, and satisfied discrimination (AUC = 0.81) and calibration (χ²_[8] = 6.48; P = .593) in the validation (ultrahigh-frequency hearing loss) cohort. According to these results, we agree that the authors built a simple model for the identification of patients with DR-TB who are at the highest risk of developing AG-induced ototoxicity. However, we think this model should be improved before its clinical application.

中文翻译：

---

预测模型在临床应用之前需要进一步改进

致编辑-我们感兴趣地阅读了Hong等人[1]的一项令人印象深刻的研究，该研究开发了一种新的氨基糖苷（AG）诱发耐药结核病（DR-TB）治疗的患者听力丧失的临床预测模型。结果表明合理歧视（曲线下面积[AUC] = 0.71）和校准（χ ² _[8] = 6.10; P = 0.636）的开发队列中，和满意歧视（AUC = 0.81）和校准（χ ² _{[ 8]}= 6.48; P = .593）在验证（超高频听力损失）队列中。根据这些结果，我们同意作者建立了一个简单的模型，用于识别发生AG诱发的耳毒性最高风险的DR-TB患者。但是，我们认为该模型应在临床应用之前加以改进。