Study objective: To compare the value of ECG measurements from lead aVR with the QRS-interval duration in predicting seizures and ventricular arrhythmias due to acute tricyclic antidepressant (TCA) toxicity.
Design: Prospective cohort series of referral cells from hospitals to a regional poison control center.
Participants: Seventy-nine patients (mean age, 30 +/- 15 years) who presented within 24 hours of ingestion. Seizures occurred in 16 patients (20%) and ventricular arrhythmias in 5 (6%).
Interventions: The amplitude of the terminal R wave in lead aVR (RaVR), the R-wave/S-wave ratio in lead aVR (R/SaVR), and the maximal limb-lead QRS interval were measured on the initial ECG.
Results: RaVR was greater in those patients who had seizures or arrhythmias than in those who did not (4.4 versus 1.8 mm, P < .001), as was R/SaVR (1.4 versus .5, P < .001). The sensitivity of an RaVR of 3 mm or more was 81% and that of an R/SaVR of .7 or more was 75%, compared with 82% for QRS intervals greater than 100 milliseconds. The positive predictive value (PPV) of an RaVR of 3 mm or more was 43% and that of the R/SaVR of .7 or more 46%, compared with a PPV for QRS interval of 100 milliseconds or more of 35%. Multiple logistic-regression analysis demonstrated that an RaVR of 3 mm or more was the only ECG variable that significantly predicted seizures and arrhythmias (OR, 6.9 [95% CI, 1.2 to 40], P = .03).
Conclusion: RaVR and R/SaVR were greater in patients in whom seizures or arrhythmias developed after an acute TCA overdose. RaVR of 3 mm or more was the only ECG variable that significantly predicted these adverse outcomes.