Antipsychotic Drug Administration Does Not Correlate with Prolonged Rate-Corrected QT Interval in Children and Adolescents: Results from a Nested Case–Control Study

Introduction

A ntipsychotic drugs are effective in children and adolescents with schizophrenia, bipolar disorder, autism, aggression, and tics (Varley and McClellan 2009). These drugs have a dose-dependent effect on the myocardial repolarization that may lead to prolongation of the rate-corrected QT interval (QTc), torsades de pointes, and sudden cardiac death (Vieweg 2002). The estimated incidence of sudden cardiac death in children and adolescents, excluding sudden infant death syndrome, is in the range 0.6–6.2/100,000 in the United States, but the precise contribution of the long QT syndrome (congenital or acquired) to these events is not known (Gajewski and Saul 2010).

The benefit of electrocardiographic screening for QTc prolongation in pediatric populations requiring antipsychotics is not known. In this study we evaluated a cohort of pediatric psychiatric inpatients to determine the prevalence of QTc prolongation and to investigate associations between QTc duration and the utilization of antipsychotics.

Methods

Data for this study were collected from the records of 826 individual patients aged 7–18 years who were consecutively evaluated in the admissions unit and hospitalized over a 10-year period (2000–2009) to a single 230-bed acute care psychiatric hospital. The hospital policy requires a 12-lead electrocardiogram (ECG) on the day of admission. All ECGs were recorded at the same paper speed (25 mm/sec) with the same type of instrument (ELI 1200; Mortara Instruments, Milwaukee, WI). The study was approved by the Institutional Review Board of the North Shore-Long Island Jewish Health System, Manhasset, NY.

We identified for each patient the earliest available ECG. Two medical consultants with extensive clinical and research experience in ECG interpretation measured the QT interval and calculated the QTc according to the Bazett's formula (Haddad and Anderson 2002). These physicians were provided numerically coded ECG's and had no access to demographic or clinical data. The QT was measured from the onset of QRS complex to the end of the T wave at the junction of the T-P line in at least three different leads of 12-lead ECGs recorded at a speed of 25 mm/sec. Tracings for 11 patients were not suitable for accurate interpretation and were excluded from the sample. During the period of clinical observation, none of these patients had evidence of symptomatic arrhythmias and none died suddenly.

The QTc was considered prolonged if >440 msec, regardless of gender, as recommended in the cardiology and pharmacological literature (Moss 1993; Haddad and Anderson 2002) and used in studies of children treated with tricyclic antidepressants (Wilens et al. 1996) and ziprasidone (Blair et al. 2005). This unisex threshold is at variance with findings from a community-based representative North American population that demonstrated a drop of 20 msec in QTc duration in adolescent males after puberty (Rautarhaju et al. 1992), but was used by us to allow comparisons with previous studies. In a nested case–control design, each patient with prolonged QTc was matched for sex and age, with five patients selected at random from among the 680 cases whose QTc was <420 msec.

The psychiatric (diagnoses, antipsychotic drug treatment, chlorpromazine equivalent of atypical and/or typical antipsychotic, and other psychotropic medications), anthropometric (body mass index), and biochemical (serum potassium) characteristics of patients with prolonged QTc were compared with those of the nested control subjects. The chlorpromazine equivalent of antipsychotics was calculated as 100 mg chlorpromazine equal 2 mg of risperidone, 5 mg of olanzapine, 75 mg of quetiapine, 60 mg of ziprasidone, 7.5 mg of aripiprazole (5), 2 mg of haloperidol, and 10 mg of molindone (Haase 1983; Woods 2003). Analyses of variance, chi-square test, and Fischer's Exact test determined the significance of univariate differences, with alpha set at <0.05. Variables with differences significant at p < 0.1 were further analyzed with logistic regression. Data were processed with JMP 5.0.1., 1989–2003, SAS Institute, Inc, Cary, NC.

Results

Discussion

The QTc prolongation was infrequent (1.97%) and mild (range: 441–481 msec) in a cohort of 811 children and adolescents who had EKGs recorded at the time of admission for hospital treatment of psychiatric disorders. Based on retrospective chart review, a nested case–control comparison of youth with prolonged QTc with gender- and age-matched patients with QTc < 420 msec indicated that obesity and hypokalemia were significantly more common among patients with QTc prolongation. In contrast, treatment with antipsychotic drugs, exposure to individual atypical antipsychotics, chlorpromazine equivalent doses of antipsychotic treatment, and exposure to other psychotropic medications were not associated with QTc prolongation. The results must be interpreted cautiously, because our study is limited by its cross-sectional design and absence of plsma levels of antipsychotic drugs.

The low prevalence of QTc prolongation in our pediatric cohort stands in stark contrast to the much higher prevalence of QT interval disturbance in older age groups (Lubart et al. 2009; Vieweg et al. 2009). Depending on sample selection and type correction for heart rate, the QTc was found to be to prolonged in 27%–75% of hospitalized patients older than 60 years (Lubart et al. 2009). The explanation for the much lower rate of QTc prolongation in pediatric samples is undoubtedly the prevalence of heart disease, as 43% of the elderly with prolonged QTc had ischemic heart disease (Lubart et al. 2009), a condition rare in children and adolescents.

The higher frequency of hypokalemia among patients with prolonged QTc is not surprising, as this electrolyte disturbance delays myocardial repolarization (Haddad and Anderson 2002). Hypokalemia must be suspected in all pediatric patients with a frequent history of inadequate food intake, vomiting, diarrhea, or use of diuretics or laxatives. The association between obesity and QTc prolongation identified in our study confirms work demonstrating increased heterogeneity of ventricular repolarization in obese nonhypertensive children (Guven et al. 2010; Nigro et al. 2010) and it is particularly worrisome, given the marked weight gain associated with atypical antipsychotic medications in children and adolescents (Correll et al. 2009; Varley and McClellan 2009).

Does a 2% prevalence of baseline QTc prolongation justify mandatory electrocardiographic screening of children and adolescents before the initiation of treatment with antipsychotics? We could not assess the future arrhythmic events in our cohort, but addressing a similar question about the use of psychostimulants, the American Academy of Pediatrics has found that “it has not been shown that screening EKG's before starting stimulants have an appropriate balance of benefit, risk and cost effectiveness for general use in identifying risk factors for sudden death” (Perrin et al. 2008). If confirmed in larger studies, our findings would suggest restricting screening ECG's in youth requiring antipsychotic treatment to those considered to be at higher risk on account of heart disease, obesity, and hypokalemia.