ECG changes in patients on chronic psychotropic medication

Objectives. To determine the ECG changes in a group of outpatients on chronic psychotropic medication, and the association, if any, with factors such as gender, age, co- morbid illness and the use of concomitant medication. Methods. Study subjects included patients 18 years and older attending the outpatient departments of Chris Hani Baragwanath and Johannesburg hospitals. The subjects’ demographic and clinical characteristics were obtained and a resting ECG was recorded. Results. Eighty patients were included in the study. The mean age of the subjects was 45.4 (standard deviation (SD) =18.2) years, with a minimum age of 18 and a maximum of 86 years. Fifty-four subjects (67.5%) had evidence of some ECG abnormalities. There was no significant difference between the occurrence of ECG abnormalities and the different age groups (p > 0.05), gender (p > 0.05), and different race groups (p > 0.05). Sixty-one subjects (76.3%) had no co-morbid medical illness and were on psychotropic medication only; of these patients 43 (70.5%) had abnormal ECG tracings (p > 0.05). The ECG abnormalities recorded included abnormal rate (28.8%), abnormal ST segment (20.5%), abnormal QRS complex (17.8%), abnormal T wave (15.4%), prolonged or borderline corrected QT interval (8.2%), irregular rhythm (5.5%) and prolonged PR interval (2.7%). There was a significant positive correlation between the corrected QT interval and age (r = 0.43, p 0.05). Conclusion. The use of psychotropic drugs is associated with ECG changes in ordinary doses. However, this study serves to strengthen previous evidence that, although common, most of these changes are of a benign nature.

Objectives.To determine the ECG changes in a group of outpatients on chronic psychotropic medication, and the association, if any, with factors such as gender, age, comorbid illness and the use of concomitant medication.
Methods.Study subjects included patients 18 years and older attending the outpatient departments of Chris Hani Baragwanath and Johannesburg hospitals.The subjects' demographic and clinical characteristics were obtained and a resting ECG was recorded.
Results.Eighty patients were included in the study.The mean age of the subjects was 45.4 (standard deviation (SD) =18.2) years, with a minimum age of 18 and a maximum of 86 years.Fifty-four subjects (67.5%) had evidence of some ECG abnormalities.There was no significant difference between the occurrence of ECG abnormalities and the different age groups (p > 0.05), gender (p > 0.05), and different race groups (p > 0.05).Sixty-one subjects (76.3%) had no co-morbid medical illness and were on psychotropic medication only; of these patients 43 (70.5%) had abnormal ECG tracings (p > 0.05).The ECG abnormalities recorded included abnormal rate (28.8%), abnormal ST segment (20.5%), abnormal QRS complex (17.8%), abnormal T wave (15.4%), prolonged or borderline corrected QT interval (8.2%), irregular rhythm (5.5%) and prolonged PR interval (2.7%).There was a significant positive correlation between the corrected QT interval and age (r = 0.43, p < 0.05) and between corrected QT interval and female gender (r = 0.31, p < 0.05).There was no correlation between corrected QT interval and treatment of a co-morbid illness (r = 0.13, p > 0.05).

Conclusion. The use of psychotropic drugs is associated with
ECG changes in ordinary doses.However, this study serves to strengthen previous evidence that, although common, most of these changes are of a benign nature.

Volume 12
No. 3 September 2006 -SAJP measurement.However, a consensus appears to be emerging of a normal QTc upper limit of 450 ms for males and 470 ms for females, with a 'red zone' limit of 500 ms for both genders.[14] The risk of drug-induced torsades de pointes is increased under certain conditions, viz.structural heart disease, intracranial lesions, electrolyte abnormalities, hypothyroidism, pre-existing QT prolongation, QT dispersion, sinus bradycardia and polymorphic ventricular premature beats. 15It can also be induced by drugs, viz.antihistamines, antimalarials, antifungals, macrolide antibiotics, prokinetics and psychotropics. 16,17These effects may be indirect through cytochrome P450 enzymes (fluoxetine, fluvoxamine, and ketoconazole), or a direct effect (tricyclic antidepressants, antihistamines and anti-infectives).It is important to note that the correlation between prolonged QTc and torsades de pointes is not always a direct one in that there are a number of medications that prolong QTc but do not cause torsades de pointes.
The prevalence of torsades de pointes in the psychiatric population is unknown, but estimates from antiarrhythmic-induced torsades de pointes in the cardiac population range from 3% to 15%.Although usually self-limiting, torsades de pointes tends to recur, and in 31% of cases progresses to ventricular fibrillation and sudden death. 18Being associated with entirely nonspecific symptoms such as palpitations, dizziness, syncope and seizures, its potential seriousness may easily be misconstrued as primary psychiatric disorder, which can have a fatal outcome.
Antipsychotics differ in their capacity for QT prolongation. 11mong the antipsychotic drugs, the low-potency typical antipsychotics have most often been implicated.The highpotency typical and the atypical antipsychotics are less frequently associated with torsades de pointes; however they (especially ziprasidone) have raised much debate and serious concern, which caused the Food and Drug Administration (FDA) to delay approval in some instances.Goodnick et al. 19 report that the greatest concern is directed at the immediate use of haloperidol, the short-term use of thioridazine, and the long-term use of clozapine and olanzapine.
Among the antidepressants, the tertiary tricyclic antidepressants (imipramine, amitriptyline) appear to have a more general impact, while the secondary tricyclic antidepressants (nortriptyline, desipramine) may impact more on children and the elderly.Among other antidepressants, the only reports of torsades de pointes appear to have occurred with mirtazapine.There are no effects on QTc by sertraline, citalopram, paroxetine and bupropion.Lithium and the benzodiazepines show little effect on the QTc, although there may be effects on other cardiovascular parameters. 19ditional risk factors for QT prolongation and torsades de pointes in the psychiatric population include deliberate or accidental antipsychotic overdose, co-morbid substance misuse and, in particular, the effects of high sympathetic arousal during restraint.
Because of a lack of resources in South Africa it is common for psychotropic medication to be initiated and maintained in an outpatient setting.At most, monitoring of these patients can only be done on a monthly basis, often by a psychiatric nurse.As this group of patients is at high risk for the cardiac side-effects of psychotropic medication and they are not well monitored by trained staff, there is a need to establish the safety of these drugs in our everyday clinical practice.The aim of this study was to determine the ECG changes in a group of outpatients on chronic psychotropic medication and the association, if any, with factors such as gender, age, co-morbid medical illness and concomitant medication.

Subjects
A cross-sectional study of all patients aged 18 years and older attending the outpatient departments of Chris Hani Baragwanath and Johannesburg hospitals was undertaken during the period September 2004 -November 2004.Patients were included if they were psychiatrically stable and had been on psychotropic medication for more than 6 months.Pregnant women were excluded from the study.The University of the Witwatersrand Human Research Ethics Committee approved the study.

Procedures
After obtaining written informed consent, the subjects' demographic data (age, race and gender) were recorded as well as presence of co-morbid medical illness and all medication currently used.The subjects then had a resting ECG recorded, which was analysed by the ECG machine and checked by a physician with respect to the rate, rhythm and other parameters.

Statistical analysis
The outcome variable was an abnormal ECG recording, and the factors considered were age, gender, race and the presence of comorbid illness.Descriptive statistics were computed as mean and articles Volume 12 No. 3 September 2006 -SAJP frequencies (count and percentages).The two-sample t-test was used to compare the continuous characteristics (age) between the groups.Comparisons between the outcome variable with respect to the exposure variables were examined using contingency tables (chi-squared test with Yates's correction and Fisher's exact test).Logistical regression was computed to determine any significant associations between QTc and exposure variables.All analysis was done using Statistical Package for Social Sciences 10.0 for Windows (SPSS Inc., Chicago, Ill.).A value of p < 0.05 was considered significant.

Results
About 150 patients attended the outpatient clinics during this period but only 80 patients volunteered to be included in the study.The mean age of the subjects was 45.4 (standard deviation (SD) 18.2) years, with a minimum age of 18 and a maximum of 86 years.Fifty-four (67.5%) had evidence of some ECG abnormalities (Table I).There was no significant difference between ECG abnormalities and the different age groups (χ 2 = 3.77, p > 0.05), gender (χ 2 = 0.66, p > 0.05), and the different race groups (χ 2 = 1.86; p > 0.05).
The psychotropic medication that the patients were receiving included antipsychotics (haloperidol, trifluoperazine, risperidone, clozapine, olanzapine, quetiapine and sulpiride), antidepressants (amitriptyline, clomipramine, mianserin, fluoxetine, citalopram and venlafaxine), and mood stabilisers (lithium, valproate, carbamazepine and lamotrigine).Patients were either receiving antipsychotics alone or a combination of antipsychotic and/or antidepressant and/or mood stabiliser.
The ECG abnormalities recorded included prolonged or borderline QT interval (8.2%), abnormal rate (28.8%), irregular rhythm (5.5%), prolonged PR interval (2.7%), abnormal QRS complex (17.8%), abnormal T wave (15.4%), and abnormal ST segment (20.5%) (Table II).The ECG abnormalities occurred more frequently in patients on antipsychotic medication alone or if it was combined with an anticonvulsant.The abnormalities were less frequent if the patient was on an antidepressant alone or in combination with an antipsychotic.
There was a significant positive correlation between the corrected QT interval and age (r = 0.43, p = 0.0001) (Fig. 1a) and between corrected QT interval and female gender (r = 0.31, p = 0.006) (Fig. 1b).There was no correlation between corrected QT interval and the treatment of a co-morbid illness (r = -0.13,p > 0.05).

Discussion
In this study the chronic use of common psychotropic medication was associated with abnormalities in the ECG tracings.1][22] However, most of the changes recorded are considered 'benign' 23 and may also be seen in athletes without demonstrable organic heart disease, chronic schizophrenics not receiving any psychotropic medication, physically healthy persons under certain stressful conditions, 10 and patients on placebo therapy. 24like other studies, 25 frequency of ECG changes did not increase with age or concomitant use of other medication.Concern about prescribing psychotropic medication is greatest in the case of the elderly and patients with co-morbid medical illness.Patients with co-morbid illnesses are more susceptible to the side-effects of psychotropics because of disturbed drug distribution and metabolism and because of the likelihood of interactions between psychotropic and non-psychotropic medication.It is possible that some of the patients exhibiting ECG abnormalities may have some as yet undiagnosed and untreated cardiac pathology.
Previous studies have shown that predictors of QTc lengthening include age over 65 years, use of tricyclic antidepressants and antipsychotics, antipsychotic dose, female gender, bradycardia, electrolyte imbalances, cardiac diseases, simultaneous use of multiple drugs prolonging QT interval, and genetic predisposition.This study confirmed a significant positive correlation between the corrected QT interval and age and female gender but did not show any correlation with a bradycardia, simultaneous use of multiple drugs or pre-existing medical illness.It is likely that this is because of the small sample size rather than any fundamental difference in characteristics of this study population.
This study is limited in its generalisability because all subjects were attendees of a tertiary hospital outpatient department.The small sample size may have statistical limitations in ascertaining meaningful differences when comparing groups.Finally, the

Conclusion
Psychotropic drugs have properties that result in ECG changes in ordinary doses and there is much concern about these cardiac effects and their relation to sudden death.This study serves to provide some evidence to mental health care practitioners in limited-resources settings that it is relatively safe to initiate and titrate psychotropic medication in an outpatient setting.However, it would be prudent to ask apparently healthy patients if they have had syncope, if they have relatives with long QT syndrome, or if they have relatives who died suddenly at a young age, before initiating treatment with psychotropic medication.Among older patients, especially those with known heart disease or taking drugs that can prolong QT, a pre-treatment ECG would be appropriate.
Finally, before prescribing a medicinal product that prolongs QT interval, physicians should carefully evaluate not only the disease they want to treat but also the availability of equally effective, alternative drugs.One of the most basic ethical principles of medicine requires that the beneficial effects expected from a therapy should, for each treated patient, outweigh any possible adverse consequence, particularly when the latter could be lethal.
The Division of Psychiatry, University of the Witwatersrand, assisted in the funding of this study.

Table II . Frequency of the various types of ECG changes
articles Volume 12 No. 3 September 2006 -SAJP