Abstract
Clozapine is a highly effective atypical antipsychotic in treatment of patients with schizophrenia. Acute clozapine poisoning is commonly reported worldwide. This cohort study aimed to assess the serum glucose/potassium ratio and leucocyte glucose index as predictors of poor outcomes in acute clozapine poisoning with a construction of a risk score model. Medical records of 215 acute clozapine poisoned patients from the start of January 2020 to the end of December 2023 were reviewed. Personal and toxicological data, findings of clinical examination, electrocardiography (ECG), and results of laboratory investigations were reported. Mortality, need for intensive care unit admission, and/or mechanical ventilation were recorded as poor outcomes. Patients with poor outcomes had significantly higher GPR and LGI than patients with good outcomes (P < 0.001). Both parameters had a significant weak negative correlation with Glasgow coma scale (GCS) and a significant weak positive correlation with each of length of hospital stay, poisoning severity score, and ECG. The developed risk score model included diastolic blood pressure, GCS, pH, ECG, GPR, and LGI with 50.91% sensitivity and 93.12% specificity. Areas under the curve of the model, GPR, and LGI were 0.844, 0.729, and 0.665 respectively. Pairwise comparisons revealed that the developed model was a significant predictor of poor outcomes better than GPR and LGI individually (P = 0.023 and 0.001, respectively). Both GPR and LGI offer a rapid bedside assessment of patient risk and potentially facilitate timely clinical decision-making in acute clozapine poisoning. A comprehensive risk score model incorporating these parameters alongside other clinical variables would significantly enhance prognostic accuracy in acute clozapine poisoning.