Ahn Seong-Jun, Lee Jun-Ho, Kim Yong-Hwan, Lee Dong-Woo, Hwang Seong-Youn, Park Jong-Yoon
Department of Emergency Medicine, Samsung Changwon Hospital, Sungkyunkwan University School of Medicine, Changwon 630-723, Korea.
J Clin Med. 2022 Aug 17;11(16):4824. doi: 10.3390/jcm11164824.
Background: Glyphosate herbicide (GH) is widely used worldwide. It has a higher fatality rate than expected. GH-poisoned cases are increasingly reported. Acute kidney injury in poisoned patients is one of several predictors of GH mortality. The aim of this study was to determine whether estimated glomerular filtration rate (eGFR) could predict kidney injury in GH intoxication. Methods: This was a retrospective study conducted at the emergency department (ED) of a single hospital between January 2004 and December 2021. A total of 434 patients presented with GH intoxication via oral ingestion, and 424 were enrolled. Their demographic characteristics, laboratory variables, complications, and mortality were analyzed to determine clinical predictors associated with GH-induced mortality using a logistic regression analysis. The relationship between GH intoxication and eGFR was determined based on the results of dominance analysis. Additionally, the comparison of creatinine and eGFR was performed through receiver operating characteristic (ROC) curves. Results: A total of 424 GH-poisoned patients were enrolled. Of them, 43 (10.1%) died. In the multivariable analysis, initial GCS (OR: 0.874; 95% CI: 0.765−0.998, p = 0.047), albumin (OR: 0.874; 95% CI: 0.765−0.998, p = 0.027), pH (OR: 0.002; 95% CI: 0.000−0.037, p < 0.001), QTc interval (OR: 1.018; 95% CI: 1.007−1.029, p = 0.001), and eGFR (OR: 0.969; 95% CI: 0.95−0.989, p = 0.003) were independent factors for predicting in-hospital mortality. In the dominance analysis of the relative importance of the predictive factors, pH accounted for the largest proportion at 35.8%, followed by QTc (20.0%), GCS (17.3%), eGFR (17.0%), and albumin (9.9%). Additionally, eGFR had a larger area under the ROC curve (0.846; 95% CI, 0.809−0.879) than that of creatinine (0.811; 95% CI, 0.771−0.848, p = 0.033). Conclusion: In sum, eGFR, considered a surrogate of renal function, was a useful prognostic factor for mortality in glyphosate herbicide-poisoned patients.
草甘膦除草剂（GH）在全球广泛使用。其致死率高于预期。草甘膦中毒病例报告日益增多。中毒患者的急性肾损伤是草甘膦致死的几个预测指标之一。本研究旨在确定估算肾小球滤过率（eGFR）是否可预测草甘膦中毒所致的肾损伤。方法：这是一项在一家医院急诊科于2004年1月至2021年12月进行的回顾性研究。共有434例经口服摄入草甘膦中毒的患者，纳入424例。分析他们的人口统计学特征、实验室变量、并发症和死亡率，采用逻辑回归分析确定与草甘膦诱导死亡相关的临床预测因素。根据优势分析结果确定草甘膦中毒与eGFR之间的关系。此外，通过受试者工作特征（ROC）曲线对肌酐和eGFR进行比较。结果：共纳入424例草甘膦中毒患者。其中43例（10.1%）死亡。在多变量分析中，初始格拉斯哥昏迷量表（GCS）（比值比：0.874；95%置信区间：0.765 - 0.998，p = 0.047）、白蛋白（比值比：0.874；95%置信区间：0.765 - 0.998，p = 0.027）、pH值（比值比：0.002；95%置信区间：0.000 - 0.037，p < 0.001）、QTc间期（比值比：1.018；95%置信区间：1.007 - 1.029，p = 0.001）和eGFR（比值比：0.969；95%置信区间：0.95 - 0.989，p = 0.003）是预测院内死亡的独立因素。在预测因素相对重要性的优势分析中，pH值占比最大，为35.8%，其次是QTc（20.0%）、GCS（17.3%）、eGFR（17.0%）和白蛋白（9.9%）。此外，eGFR的ROC曲线下面积（0.846；95%置信区间，0.809 - 0.879）大于肌酐（0.811；95%置信区间，0.771 - 0.848，p = 0.033）。结论：总之，eGFR作为肾功能的替代指标，是草甘膦除草剂中毒患者死亡率的有用预后因素。
