Risk assessment of severe tricyclic antidepressant overdose.

Prognostic factors for severe complications in tricyclic antidepressant (TCA) overdose remain unclear. We therefore evaluated the value of clinical characteristics and electrocardiograph (ECG) parameters to predict serious events (seizures, arrhythmia, death) in severe TCA overdose of 100 patients using logistic regression models for risk assessment. The overall fatality rate was 6%, arrhythmia occurred in 21% and 31% of the patients developed seizures. Using an univariable logistic regression model, the maximal QRS interval (OR 1.22; 95% CI 1.06-1.41; p = .005), the time lag between ingestion and occurrence of first symptoms of overdose (OR 1.13; 95% CI 0.99-1.29; p = .072) and the age (OR 0.73; 95% CI 0.55-0.98; p = .038) were determined as the solely predictive parameters. In the multivariable logistic regression model, the QRS interval could not be established as independent predictor, however, the terminal 40-ms frontal plane QRS vector (T40) reached statistical significance regarding prediction of serious events (odds ration [OR] 1.70; 95% confidence interval [CI] 1.02-2.84; p = .041), along with age and time lag between ingestion and onset of symptoms of overdose with a sensitivity and specificity of 71% and 70%, respectively. Evaluation of both clinical characteristics and ECG-parameters in the early stage of TCA overdose may help to identify those patients who urgently need further aggressive medical observation and management.