Objectives: To describe the changes over time of the QRS interval and terminal 40-msec QRS frontal axis (T40-ms) in patients with acute tricyclic antidepressant poisoning, to identify clinical factors and treatment associated with these changes, and to determine if patients with tricyclic antidepressant-related complications (seizures and/or arrhythmias) had differences in such serial electrocardiogram (ECG) changes when compared with patients without complications.
Design: Prospective, observational, cohort study.
Setting: Emergency departments of community and university-based hospitals in Massachusetts that consulted a large regional poison center.
Patients: Thirty-six patients who presented with an acute ingestion (< 24 hrs) of a tricyclic antidepressant, who had at least three electrocardiograms in the first 8 hrs and serial ECGs until discharge, and who had a peak tricyclic antidepressant concentration of > 300 ng/mL.
Measurements and main results: The maximal limb-lead QRS interval and T40-ms axis were measured manually in all ECGs. The maximum recorded QRS interval occurred at the time of presentation for 24(80%) of 30 patients whose QRS was > or = 100 msecs and a median time of 3 hrs (range 1 to 9) for the other six patients. The maximum recorded T40-ms axis occurred at the time of presentation for 31(86%) of 36 patients and at a median time of 3 hrs (range 1 to 5) for the remaining five patients. The minimum QRS interval observed remained > or = 100 msecs in 15 patients (range 100 to 140 msecs) and decreased to < 100 msecs in 15 patients. The median time from presentation to the first ECG with a QRS < 100 msecs was 20 hrs (range 1 to 153) in those 15 patients. There were no significant differences in clinical characteristics and treatment (including sodium bicarbonate therapy) between the two groups. The minimum recorded T40-ms remained > or = 120 degrees in 30 patients and decreased to < 120 degrees in six patients. The median time from presentation until the first ECG with a T40-ms axis < 120 degrees was 13 hrs (range 2 to 30) for the six patients. All ECG measurements were greater and remained abnormal for a significantly longer duration in those patients who developed seizures and/or ventricular arrhythmias. These two ECG parameters demonstrated ongoing changes and persistent abnormalities despite clinical improvement in all patients except one.
Conclusions: The conduction abnormalities seen in severe tricyclic antidepressant toxicity vary widely in the time observed for resolution of these abnormalities and sometimes remain persistently abnormal. All ECG parameters were significantly more abnormal in those patients who developed seizures and/or arrhythmias. Clinical improvement occurred both before and during these ECG changes.