Researchers analyzing a cross-section of EHRs from one large New England health system found that the antidepressants citalopram, escitalopram and amitriptyline, as well as methadone contribute to long QT interval, a marker for heart rhythm abnormalities. The study was published online Jan. 29 in BMJ.
In March 2012, the FDA issued a revised drug label for citalopram hydrobromide (Celexa) and its generic equivalents, warning that the drug was associated with prolonged QT intervals and was not recommended for people with congenital long QT syndrome. The label also changed the maximum recommended dose for citalopram to 20 mg/day for patients 60 years of age or older.
Roy H. Perlis, MD, of the psychiatry department at Massachusetts General Hospital in Boston, and colleagues attempted to quantify the association between elongated QT and various selective serotonin reuptake inhibitors (SSRI) in a diverse population. Using the EHR of the Partners HealthCare System, which encompasses records of more than 4 million people, the researchers used a pharmacovigilance approach to examine the QT intervals of patients who had been prescribed SSRI. They also examined the records of patients who received methadone as a measure of assay sensitivity, because methadone is known to contribute to QT elongation.
The researchers identified 38,397 people who had been given at least one prescription for an SSRI or methadone between February 1990, and August 2011, and who had an electrocardiogram between 14 and 90 days after receiving the prescription. The researchers excluded patients who did not have a follow-up prescription after their initial prescription, and those who received prescriptions for antipsychotic drugs and other drugs known to elongate QT intervals within one year of the date of the electrocardiogram.
The researchers conducted linear regression models using dose as well as covariates including age, gender, ethnicity and potential confounders such as cardiovascular disease history. When the researchers found a significant dose effect in linear regression models they conducted post hoc pairwise comparisons to determine whether specific dosing increments contributed to QT elongation. For some patients, electrocardiograms were available at escalating doses and these were examined to estimate a dose response.
The researchers found “statistically significant evidence of modest QT prolongation for … methadone and the tricyclic antidepressant amitriptyline, as well as for citalopram and escitalopram. However, the sizes of these effects were small, and the proportion of individuals with abnormal QTc intervals were broadly similar across individual antidepressant treatment.”
The authors concluded that for patients with cardiovascular risk factors, SSRIs that are not associated with QT prolongation, such as sertraline, may be preferred over citalopram. They noted as well that bupropion is often used to augment antidepressant treatment, that bupropion has the capacity to shorten the QT interval, and therefore “bupropion treatment might be a reasonable next step for patients partially responsive to citalopram who would otherwise require a dose increase.”
Perlis et al also discussed the benefits of conducting pharmacovigilance using EHR data, explaining that this method allows researchers to detect changes that are subtle and unlikely to have been evident in postmarketing surveillance. They also pointed out that the subject population in this type of study is more diverse than is found in randomized trials. “For a drug that has already been marketed, this method may be the only feasible way to systematically examine electrocardiographic changes without requiring a crossover trial,” they wrote.
The authors noted that as there was no random assignment of treatment and dose, there is an increased risk of confounding, which they described as a “key limitation” of their study. They also pointed out that most patients who receive prescriptions for antidepressants do not receive electrocardiograms, which would indicate a bias toward greater mean QT. The results of the study are therefore most applicable to an older and sicker population, the authors said.