Prognostic factor determination mortality of acute glufosinate-poisoned patients.

BACKGROUND: Glufosinate-containing herbicide is increasingly used in agriculture. Its poisoning is a worldwide concern. More and more patients are poisoned by glufosinate. The aim of this study was to determine the factors associated with mortality of patients with acute poisoning of glufosinate.

METHODS: This was a retrospective cohort study conducted from January 1998 to October 2015. Using a multivariate logistic analysis, data for the total population were retrospectively analyzed to determine the factors associated with mortality. Various variables were compared in survivors and non-survivors. Significant predictive variables, Acute Physiology and Chronic Health Evaluation (APACHE) II scoring system, the Simplified Acute Physiology Score (SAPS) II, and Sequential Organ Failure Assessment (SOFA) score were compared by analyzing receiver operating characteristic (ROC) curves.

RESULTS: A total of 253 patients (mean age: 58 years) were enrolled. Of the 253 patients, 219 (86.6%) survived and 34 (13.4%) died. Decreased Glasgow Coma Scale (GCS) and bicarbonate (HCO3 - ), use of mechanical ventilator, and use of vasopressors (dopamine, dobutamine, norepinephrine) were associated with mortality. The areas under the curve in the ROC curve analysis for the predictive variables, SOFA score, APACHE II scoring system, and SAPS II were 0.952, 0.829, 0.927, and 0.944, respectively.

CONCLUSION: Four predictive variables (GCS < 9, HCO3 - < 16.0 mmol/L, mechanical ventilator apply, and use of vasopressors) were associated with mortality in the total population of patients with acute poisoning of glufosinate. These predictive variables had good discriminative power for predicting mortality of patients with acute poisoning of glufosinate-containing herbicide compared to APACHE II, SOFA, and SAPS II.