Article, Toxicology

Electrocardiographic manifestations of tramadol toxicity with special reference to their ability for prediction of seizures

electrocardiographic manifestations of tramadol toxicity with special reference to their ability for prediction

of seizures?

Mohammadali Emamhadi MD a, Hossein Sanaei-Zadeh MD b,?, Masoumeh Nikniya MD b,

Nasim Zamani MD b, Richard C. Dart MD, PhD c

aDepartment of Forensic Medicine and Toxicology, Loghman Hakim Poison Hospital,

Shaheed-Beheshti University of Medical Sciences, Tehran, Iran

bDepartment of Forensic Medicine and Toxicology, Tehran University of Medical Sciences, Tehran, Iran cRocky Mountain Poison and Drug Center, Denver Health and Department of Emergency Medicine, University of Colorado-Denver, Denver, CO, USA

Received 27 October 2011; revised 9 December 2011; accepted 9 December 2011


Aim: The aims of this study are to determine the electrocardiographic manifestations of the symptomatic patients with isolated tramadol toxicity and to predict seizures based on ECG parameters. Methods: Medical charts of a total of 479 patients with isolated tramadol toxicity were retrospectively evaluated. Their clinical manifestations were recorded, and their ECG parameters including rate, PR interval, QRS duration, corrected QT interval, terminal 40-millisecond frontal plane QRS axis, and the height of R wave and R/S ratio in the Lead aVR were measured. The data were analyzed using Kolmogorov-Smirnov test, Mann-Whitney U test, Pearson ?2, Pearson correlation coefficient (r), and the Student t test.

Results: Electrocardiographic heart rate more than 100 beats per minute in 30.6%, QRS 120 milliseconds or more in 7.5%, corrected QT interval more than 440 milliseconds in 24.6%, height of R wave more than 1 mm in lead aVR in 22.1%, R/S ratio more than 0 in lead aVR in 23.5%, terminal 40-millisecond frontal plane QRS axis greater than 120? in 31.7%, and complete or incomplete right bundle-branch block in 4.6% of the patients were detected. There were no statistically significant differences between the patients who had not convulsed and those who had convulsed after admission regarding age, sex, vital signs, and ECG findings at presentation (all P values were N.05).

Conclusions: Tramadol toxicity shows ECG changes consistent with Sodium channel blockade and potassium channel blockage effects. The risk of development of seizures cannot be predicted based on the changes of ECG parameters at presentation.

(C) 2012

? Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of this article.

* Corresponding author. Department of Forensic Medicine and

Toxicology, Tehran University of Medical Sciences, Hazrat Rasoul Akram Hospital, Tehran 1445613131, Iran. Tel./fax: +98 21 66551201.

E-mail address: [email protected] (H. Sanaei-Zadeh).


Tramadol is a widely used, synthetic opioid analgesic [1-3]. It has a weak u-receptor agonist activity that blocks the pain pathways as well as the inhibition of the reuptake of the biogenic amines especially serotonin and norepinephrine

0735-6757/$ – see front matter (C) 2012 doi:10.1016/j.ajem.2011.12.009

in central nervous system. The latter mechanism results in an increment in the pain threshold [4]. Tramadol toxicity can cause nausea and vomiting, hypertension, tachycardia, central nervous system depression, respiratory depression, agitation, and seizures [4-6]. electrocardiographic changes in opioid overdose are well described with propoxyphene [7-10] and heroin [11-14] including QRS prolongation, nonspecific ST segment and T-wave changes, first-degree atrioventricular block, atrial fibrillation, pro- longed corrected QT (QTc) intervals, and ventricular dysrhythmias. Furthermore, a specific ECG pattern, the Brugada pattern, has previously been reported with isolated tramadol overdose [15]. However, except for the previously mentioned case of Brugada pattern in tramadol toxicity, the ECG findings of this toxicity have not yet been reported. This retrospective study aimed to determine the ECG manifestations of the symptomatic patients with isolated tramadol toxicity and the probability of prediction of seizures by the application of these manifestations.


In this retrospective study, the medical charts of all patients with tramadol toxicity who had been hospitalized in Loghman-Hakim poison hospital in Tehran, Iran, between March 2009 and March 2011 were identified by the computerized discharge diagnosis (International Sta- tistical Classification of Diseases, 10th Revision, codes). Inclusion criteria were (1) patient age older than 12 years;

(2) a history of tramadol ingestion; (3) clinical manifes- tations of tramadol ingestion including hypertension (>=140/90 mm Hg), tachycardia (pulse rate >=100/min), central nervous system depression (Glasgow Coma Scale [GCS], <=13), respiratory depression (respiratory rate, b12/min), and seizure (before or after presentation); and

(4) a 12-lead ECG available in the medical record. Diagnosis of seizures before presentation had been confirmed by an accurate medical history taken from the people accompanying the patient or the report of the personnel of emergency medical service. Exclusion criteria were (1) multiple drug ingestion, (2) underlying heart disease, (3) patients with solely subjective complaints, and

(4) asymptomatic patients. Multiple drug ingestion was excluded based on a history with or without a confirmed urine or serum drug screen test. Serum tramadol levels were not available. All patients had received therapeutic interventions including gut decontamination and Standard supportive care when indicated. The patients’ data including age, sex, blood pressure, pulse rate, respiratory rate, GCS, and occurrence of seizure before presentation or after admission were extracted from the medical charts and entered into the standardized data abstraction forms. The data were double extracted. The ECG parameters were manually measured, and characteristics including rate, PR interval, maximal limb-lead QRS duration, QTc, terminal

40-millisecond (T40-ms) frontal plane QRS axis [16], and the height of R wave and R/S ratio in the lead aVR were recorded. Findings such as dysrhythmia, right bundle- branch block (RBBB), and early repolarization and Brugada patterns were also noted. Electrocardiographic interpretation was supervised by 2 cardiologists blind to the topic of the study.

Statistical analysis

Statistical analysis was done using SPSS (Statistical Package for Social Sciences) software (version 17; SPSS, Inc, Chicago, IL) and application of descriptive statistics, Kolmogorov-Smirnov test, Mann-Whitney U test, Pearson ?2, Pearson correlation coefficient (r), and the Student t test. In addition, an interrater reliability analysis (K-statistic) [17] was done to determine the consistency of the evaluation of the ECGs. P b .05 was considered to be statistically significant. Our study was approved by the regional ethics committee.


In this retrospective study, a total of 479 patients met our inclusion criteria and were enrolled into the study. Of them, 107 (22.3%) were female, and 372 (77.7%) were male. Mean age of the patients was 22.6 +- 6 years (range, 12-60 years). The patients’ vital signs at presentation are shown in Table 1. The mean GCS of the patients at presentation was 14 +- 1. In total, 265 patients (55.3%) had experienced seizures. From these, seizure had been developed before presentation in 245 (51.1%), after admission in 14 (2.9%) and before presentation and after admission in 6 (2.1%) patients. The patients’ ECG parameters at presentation are shown in Table 1. Electrocardiographic heart rate more than 100 beats per minute in 147 (30.6%), QRS 120 milliseconds or more in

36 (7.5%), QTc more than 440 milliseconds in 118 (24.6%), height of R wave more than 1 mm in lead aVR in 106 (22.1%), R/S ratio more than 0 in lead aVR in 113 (23.5%), and T40-ms frontal plane QRS axis greater than 120? in 152 (31.7%) patients were detected. In addition, complete or incomplete RBBB in 22 (4.6%), early repolarization in 3 (0.6%), Brugada pattern in 1 (0.2%), premature ventricular contractions in 2 (0.4%), and premature atrial contractions in 1 (0.2%) patients were seen. None of the patients had developed dysrhythmias.

The ? for the interrater reliability analysis for ECG interpretation between the cardiologists was 0.84 (95% confidence interval, 0.67-1.00) with P b .001. There were no statistically significant differences between the patients who had not convulsed and those who had convulsed after admission regarding age, sex, vital signs, and ECG findings at presentation (Table 1). Such statistically insignificant differences were also obtained in the comparison of the


Group I

Group II

P-value (applied statistical test)

No. of patients




Age (y)

22.6 (+-6)

22 (+-6) a

22 (+-5)

0.763 (Student t test) b

Sex (male/female)




0.313 (?2) b

Rate (beats per min)

91 (+-21)

90 (+-21)

102 (+-21)

0.052 (MWU) b

PR interval (s)

0.16 (+-0.03)

0.15 (+-0.03)

0.16 (+-0.02)

0.307 (MWU) b

QRS interval (s)

0.08 (+-0.01)

0.08 (+-0.01)

0.08 (+-0.01)

0.072 (MWU) b

QTc interval (s)

0.42 (+-0.04)

0.42 (+-0.04)

0. 44 (+-0.03)

0.103 (MWU) b

T40-ms axis (?)

59 (+-85)

54 (+-82)

68 (+-95)

0.583 (MWU) b

Height of R wave in aVR (mm)

0.6 (+-1)

0.5 (+-1)

1 (+-1.5)

0.089 (MWU) b

R/S ratio

0.12 (+-0.3)

0.1 (+-0.24)

0.32 (+-0.6)

0. 101 (MWU) b

Respiratory rate (breaths per min)

17 (+-5)

17 (+-4)

21 (+-20)

0.684 (MWU) b

Pulse rate (beats per min)

90 (+-17)

91 (+-18)

91 (+-21)

0.720 (MWU) b

Systolic BP (mm Hg)

119 (+-16)

120 (+-17)

117 (+-11)

0.895 (MWU) b

Diastolic BP (mm Hg)

75 (+-10)

75 (+-10)

76 (+-8)

0.605 (MWU) b


14 +- 1

13 (+-1)

12 (+-3)

0.867 (MWU) b

MWU indicates Mann-Whitney U test; BP, blood pressure.

a Data are presented as mean value (+-SD).

b Nonsignificant.

patients who had convulsed before presentation with those whose convulsion had occurred after admission and those who had never convulsed except for the level of conscious- ness (GCS) at presentation (Student t test, P b .001) and the height of R wave in the lead aVR (Mann-Whitney U test, P =

Table 1 electrocardiographic findings and vital signs at presentation in all patients and the groups without seizure (group I) and those with seizure after admission (group II)

.046). No statistical correlation was found between the ECG parameters and GCS as well (using Pearson correlation coefficient, all P values were N.05).


In vitro blockade of sodium channels by tramadol has previously been demonstrated at high concentration [18]. In addition, evidence supporting that tramadol causes this effect includes the reduction of compound action potential as shown in 1 experimental study [19]. Sodium channel blockade has also been described to have local anesthetic effect in human beings [20]. Our study showed that ECG changes in up to 73.4% of our symptomatic isolated, tramadol-Intoxicated patients were consistent with the sodium channel blockade effect (“Results” section). Inter- estingly, T40-ms frontal plane QRS axis greater than 120? was detected in almost one third of our patients, which is again an indicator of the fast sodium channel blockade effect of tramadol toxicity (Fig. 1). This is while, to date, T40-ms frontal plane QRS axis deviation has only been described in cyclic antidepressant (CA) poisoning [21-23].

One of our patients had Brugada pattern. This pattern has previously been documented in only 1 case of isolated tramadol overdose [15]. This pattern is also attributed to sodium channel blockade, as previously reported with tricyclic antidepressant overdose [24-26].

In xenobiotics that cause sodium channel blockade, prolongation of QT interval is due to the prolongation of the QRS complex duration [27]. The present study shows that almost one fourth of our patients had QTc prolongation (QTc N0.44 seconds). However, only 7.5% of our patients had QRS prolongation. This may be due to a probable potassium channel blockade effect [28]. This needs to be further evaluated in the future studies.

It has been suggested that Repolarization abnormalities have association with hypotension in tricyclic antidepressant toxicity [29]. Because our patients with early repolari- zation were few, we could not withdraw a statistically significant correlation between this ECG pattern and the clinical manifestations.

Previous studies have shown that some ECG changes might occur in the postictal phase of the seizures [30]. Almost half of our patients had convulsed before hospital presentation. Therefore, the ECG changes of these patients could occur during the postictal phase and be seizure related and not due to the tramadol toxicity, per se. To reject or accept such conclusion, we compared the patients who had convulsed before presentation with those who had convulsed after admission and those who had never convulsed regarding age, vital signs, and ECG parameters. We found that these 2 groups were significantly different in GCS and height of R wave in the lead aVR. It, then, can be concluded that the ECG changes except for the height of R wave are only due to tramadol toxicity.

We hypothesized that maybe–similar to CA poisoning [31-35]–we could predict seizure development based on the ECG parameters of the tramadol-intoxicated patients. Therefore, initial ECGs (ECGs at presentation) of the patients without any seizure in the course of toxicity and

Fig. 1 Electrocardiogram of one of the patients. The Prominent S wave in lead I and R wave in aVR demonstrate the T40-ms rightward axis shift (255?) [16]. The ECG also shows an incomplete RBBB.

those who had developed seizures after admission were compared (Table 1). Our study showed that based on the ECG parameters, we cannot predict seizures in these patients. The most important limitation of the current study was the diagnosis of isolated tramadol toxicity without its confirma- tory laboratory tests being available. To compensate such

limitation, we only selected the symptomatic patients. However, it should not be forgotten that, in the preliminary studies that have been performed on the clinical manifesta- tions of tramadol toxicity and have been cited in the literature, serum tramadol levels have not been available as well [5,6]. Another limitation of our study is lack of serum

level of tramadol, and therefore, we could not comment on its correlation with Abnormal ECG parameters. The probable electrolyte abnormalities that can affect the ECG parameters

–and have not been considered in the course of the study–

may be another limitation of the present study.


According to our study, tramadol toxicity shows ECG changes that are consistent with sodium channel blockade effects. However, the potassium channel blockage effect may also be suggested for it. In addition, this type of toxicity causes T40-ms frontal plane QRS axis deviation, which, to date, has only been shown in the toxicity induced by CAs. In addition, it was concluded that the risk of development of seizures cannot be predicted based on the changes of ECG parameters at presentation. These findings are of particular clinical relevance given that tramadol is a frequent cause of drug toxicity [4,36].


The authors thank Dr Rahbar and Dr Amirfarhangi, cardiologists in the cardiology department of Hazrat Rasoul Akram Hospital, Tehran, Iran, for their kind cooperation in supervising the interpretation of the ECGs.


  1. Raffa RB, Stone Jr DJ. Unexceptional seizure potential of tramadol or its enantiomers or metabolites in mice. J Pharmacol Exp Ther 2008;325:500-6.
  2. Woody GE, Senay EC, Geller A, et al. An independent assessment of MEDWatch reporting for abuse/dependence and withdrawal from Ultram (tramadol hydrochloride). Drug Alcohol Depend 2003;72: 163-8.
  3. Shadnia S, Soltaninejad K, Heydari K, et al. tramadol intoxication: a review of 114 cases. Hum Exp Toxicol 2008;27:201-5.
  4. Tobias JD. Seizure after overdose of tramadol. South Med J 1997;90: 826-7.
  5. Marquardt KA, Alsop JA, Albertson TE. Tramadol exposures reported to statewide poison control system. Ann Pharmacother 2005;39:1039-44.
  6. Spiller HA, Gorman SE, Villalobos D, et al. Prospective multicenter evaluation of tramadol exposure. J Toxicol Clin Toxicol 1997;35:361-4.
  7. Stork CM, Redd JT, Fine K, et al. Propoxyphene-induced wide QRS complex dysrhythmia responsive to sodium bicarbonate–a case report. J Toxicol Clin Toxicol 1995;33:179-83.
  8. Lund-Jacobsen H. Cardio-respiratory toxicity of propoxyphene and norpropoxyphene in conscious rabbits. Acta Pharmacol Toxicol (Copenh) 1978;42:171-8.
  9. Gary NE, Maher JF, DeMyttenaere MH, et al. Acute propoxyphene hydrochloride intoxication. Arch Intern Med 1968;121:453-7.
  10. Gustafson A, Gustafsson B. Acute poisoning with dextropropoxy- phene. Clinical symptoms and plasma concentrations. Acta Med Scand 1976;200:241-8.
  11. Glauser FL, Downie RL, Smith WR. Electrocardiographic abnormal- ities in acute heroin overdosage. Bull Narc 1977;29:85-9.
  12. Labi M. Paroxysmal atrial fibrillation in heroin intoxication. Ann Intern Med 1969;71:951-9.
  13. Lipski J, Stimmel B, Donoso E. The effect of heroin and multiple drug abuse on the electrocardiogram. Am Heart J 1973;86:663-8.
  14. Stimmel B, Lipski J, Swartz M, et al. Electrocardiographic changes in heroins, methadone and multiple drug abuse: a postulated mechanism of sudden death in narcotic addicts. Proc Natl Conf Methadone Treat 1973;1:706-10.
  15. Cole JB, Sattiraju S, Bilden EF, et al. Isolated tramadol overdose associated with Brugada ECG pattern. Pacing Clin Electrophysiol 2010, doi:10.1111/j.1540-8159.2010.02924.x.
  16. Sanaei-Zadeh H, Zamani N, Shahmohammadi F. Methods for the measurement of the terminal 40-millisecond (T40-ms) frontal plane axis in Tricyclic antidepressant poisoning. Resuscitation 2011;82:1255-6.
  17. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33:159-74.
  18. Haeseler G, Foadi N, Ahrens J, et al. Tramadol, fentanyl and sufentanil but not morphine block voltage-operated sodium channels. Pain 2006;126:234-44.
  19. Katsuki R, Fujita T, Koga A, et al. Tramadol, but not its major metabolite (mono-O-demethyl tramadol) depresses compound action potentials in frog sciatic nerves. Br J Pharmacol 2006;149:319-27.
  20. Altunkaya H, Ozer Y, Kargi E, et al. The postoperative analgesic effect of tramadol when used as subcutaneous local anesthetic. Anesth Analg 2004;99:1461-4.
  21. Groleau G, Jotte R, Barish R. The electrocardiographic manifestations of cyclic antidepressant therapy and overdose: a review. J Emerg Med 1990;8:597-605.
  22. Niemann JT, Bessen HA, Rothstein RJ, et al. electrocardiographic criteria for tricyclic antidepressant cardiotoxicity. Am J Cardiol 1986;57:1154-9.
  23. Wolfe TR, Caravati EM, Rollins DE. Terminal 40-ms frontal plane QRS axis as a marker for tricyclic antidepressant overdose. Ann Emerg Med 1989;18:348-51.
  24. Goldgran-Toledano D, Sideris G, Kevorkian JP. Overdose of Cyclic antidepressants and the Brugada syndrome. N Engl J Med 2002;346: 1591-2.
  25. Monteban-Kooistra WE, van den Berg MP, Tulleken JE, et al. Brugada electrocardiographic pattern elicited by cyclic antidepressants over- dose. Intensive Care Med 2006;32:281-5.
  26. Akhtar M, Goldschlager NF. Brugada electrocardiographic pattern due to tricyclic antidepressant overdose. J Electrocardiol 2006;39:336-9.
  27. Nelson LS. Toxicologic myocardial sensitization. J Toxicol Clin Toxicol 2002;40:867-79.
  28. Roden DM. Drug-induced prolongation of the QT interval. N Engl J Med 2004;350:1013-22.
  29. Pellinen TJ, Farkkila M, Heikkila J, et al. Electrocardiographic and clinical features of tricyclic antidepressant intoxication. A survey of 88 cases and outlines of therapy. Ann Clin Res 1987;19:12-7.
  30. Opherk C, Coromilas J, Hirsch LJ. Heart rate and EKG changes in 102 seizures: analysis of influencing factors. Epilepsy Res 2002;52:117-27.
  31. Bailey B, Buckley NA, Amre DK. A meta-analysis of prognostic indicators to predict seizures, arrhythmias or death after tricyclic antidepressant overdose. J Toxicol Clin Toxicol 2004;42:877-88.
  32. Boehnert MT, Lovejoy Jr FH. Value of the QRS duration versus the serum drug level in predicting seizures and ventricular arrhythmias after an acute overdose of Tricyclic antidepressants. N Engl J Med 1985;313:474-9.
  33. Caravati EM, Bossart PJ. Demographic and electrocardiographic factors associated with severe tricyclic antidepressant toxicity. J Toxicol Clin Toxicol 1991;29:31-43.
  34. Liebelt EL, Francis PD, Woolf AD. ECG lead aVR versus QRS interval in predicting seizures and arrhythmias in acute tricyclic antidepressant toxicity. Ann Emerg Med 1995;26:195-201.
  35. Harrigan RA, Brady WJ. ECG abnormalities in tricyclic antidepressant ingestion. Am J Emerg Med 1999;17:387-93.
  36. Afshari R, Afshar R, Megarbane B. Tramadol overdose: review of the literature. Reanimation 2011;20:436-41.