Abstract
Background
QT interval measures cardiac repolarization, and prolongation is associated with adverse cardiovascular outcomes and death. The exponential Bazett correction formula overestimates QT interval during tachycardia.
Objective
We evaluated four methods of QT correction in individuals with sinus tachycardia for the development of coronary artery disease, heart failure and mortality.
Methods
The Penn Atrial Fibrillation Free Study (PAFF) is a large cohort of patients without AF. This study examined 6,723 PAFF patients without a history of HF and with baseline sinus rate ≥100 beats per minute. Medical records were queried for index clinical parameters, incident cardiovascular events, and all-cause mortality. QT was corrected by Bazett (QT/RR0.5); Fridericia (QT/RR0.33); Framingham (QT + 0.154*(1000-RR)); and Hodges (QT+105*(1/RR-1)).
Results
Among 6,723 patients with median follow-up of 4.5 [1.9, 6.4] years, annualized cardiovascular event rate was 2.3% and annualized mortality rate was 2.2%. QT prolongation was diagnosed in 39% of the cohort by Bazett, 6.2% by Fridericia, 3.7% by Framingham, and 8.7% by Hodges. Only Hodges’ formula was an independent risk marker for death across the range of QT values [HR highest tertile: 1.26, 95% CI (1.03–1.55)].
Conclusions
Although all correction formulas demonstrated an association between QTc values and cardiovascular events, only the Hodges formula identified one-third of individuals with tachycardia that are at higher risk for all-cause mortality. Further, Bazett’s correction overestimates the number of patients with a prolonged QT and was not associated with mortality. Future work may validate these findings and result in changes to automated algorithms for QT interval assessment.
Keywords: tachycardia, QT Interval, mortality, cardiovascular disease
Introduction
Prolongation of the QT interval reflects abnormalities in cardiac repolarization and is a risk marker for ventricular arrhythmias, other adverse cardiovascular events including heart failure (HF) and coronary artery disease (CAD), and death in various populations.1–9 The repolarization time reflects the electrophysiologic properties of ion channels that are responsible for maintaining equilibrium in myocardial cells. In addition, repolarization is also a function of the underlying heart rate and decreases with a progressive rise in the ventricular rate. This response ensures that the myocardium remains completely excitable and avoids zones of heterogeneous conduction that could result in the induction of re-entrant arrhythmias. Assessment of the QT interval, therefore, requires correction for the heart rate in order to enable comparisons with standard, reference values.
Multiple correction formulas have been empirically developed in an attempt to provide a standard index for the duration of myocardial repolarization.10–16 The Bazett method, which multiplies the measured QT interval by the inverse square root of the RR interval,10 is most commonly used to derive the corrected QT interval in clinical practice. Since the Bazett formula utilizes an exponential term, it is believed to overestimate the QT interval during states of tachycardia.17–20 We sought to compare the prognostic significance of the various correction methods in states of tachycardia and hypothesized those formulas utilizing a linear term such as Hodges or Framingham will provide a more accurate assessment of cardiac repolarization and be a better marker of adverse risk than the traditional Bazett formula.18 Further, resting sinus tachycardia results from a multitude of disease processes and is independently associated with both heart failure and all-cause mortality.21–23 In a large cohort of patients with resting sinus tachycardia, we compared four different QT-correcting formulas and their association with incident cardiovascular events and all-cause mortality.
Methods
The Penn Atrial Fibrillation Free (PAFF) study is a large, multi-hospital cohort of patients from the University of Pennsylvania Health System (UPHS) that were free of atrial fibrillation (AF) or atrial flutter at the index visit. The index visit was defined as having both a clinical encounter with a practitioner within the University of Pennsylvania Health System and a standard twelve-lead electrocardiogram (ECG). Of the 91,274 patients that had an index electrocardiogram (ECG) between January 1, 2004 and December 31, 2009, 67,265 were eligible for the study (Figure 1). To ensure adequate longitudinal data, subjects with fewer than 30 days of clinical follow-up within the UPHS were excluded. We also excluded individuals with baseline AF, atrial flutter, or a history of heart failure given the known alterations in repolarization currents that occur among patients with cardiac dysfunction.24–26 Our final analysis included the 6,723 subjects with sinus tachycardia on index ECG. This study was approved by the Institutional Review Board of the University of Pennsylvania.
Figure 1.
Design of the Penn Atrial Fibrillation Free (PAFF) Cohort.
Measurement of clinical parameters
Information about patient demographics, cardiovascular risk factors, laboratory values, and medications were extracted from electronic medical records and included a comprehensive set of clinical encounters. Clinical risk factors and diagnoses were assessed utilizing the International Classification of Diseases, version 9 (ICD-9) codes that were generated from emergency room visits, inpatient hospitalizations, outpatient encounters, and telephone encounters. These sources also provided data related to laboratory investigations and medications. In particular, we accessed Sunrise Clinical Manager (Eclipsys Corporation, Atlanta, GA), Epic (Verona, WI), Cerner (North Kansas City, MO), EmTrac (University of Pennsylvania, Philadelphia, PA), and Medview (University of Pennsylvania, Philadelphia, PA) (see Supplementary Materials). Race was self-reported by subjects. ICD-9 codes were assessed for the following conditions: heart failure; hypertension; coronary artery disease; valvular heart disease; diabetes; and chronic kidney disease (see Supplementary Materials for full listing of all ICD-9 codes). CAD was defined as either an ICD-9 code consistent with either acute myocardial infarction or chronic coronary heart disease or having evidence of prior myocardial infarction on the index ECG. Concordance between automated queries and manual chart review was confirmed in a subset of randomly chosen subjects (see Supplementary Materials).
ECG measures and assessment of the corrected QT interval
The MUSE Cardiology Information Manager (GE Healthcare, Little Chalfont, Buckinghamshire, United Kingdom) was queried to identify selected ECG data. ECGs are performed primarily using the MAC 5500 HD system, which utilizes GE Marquette 12SL ECG analysis programs, versions 231–239 (see Supplementary Materials). Left ventricular hypertrophy, left bundle branch block, or right bundle branch block was diagnosed by the over-reading cardiologist on index ECG. RR interval and QT interval measurements were performed by automated software (MUSE, GE Healthcare, Little Chalfont, Buckinghamshire, United Kingdom). In particular, the QT interval was measured as a global interval from the earliest depolarization in any lead to the latest offset of repolarization, which was defined as the T wave crossing back to baseline.27
Four methods for QT correction were used: Bazett (QTcBaz = QT/RR0.5)10; Fridericia (QTcFrid = QT/RR0.33)11, 16; Framingham (QTcFram = QT+0.154*(1000-RR))14; and Hodges (QTcHodg = QT+105*(1/RR-1))15, 28. RR interval was in seconds for all formulas except for Framingham, where it was in milliseconds. QT prolongation was defined as a corrected QT interval that was ≥460 ms for women or ≥450 ms for men.29
Ascertainment of Incident Outcomes
Patients were followed over the course of clinical care from the date of index visit to December 31, 2013. The outcomes of interest were ascertained via ICD-9 codes (see Supplementary Materials) and included incident cardiovascular disease, which is defined as the development of either heart failure or CAD, and all-cause mortality.
Statistical analysis
We compared baseline characteristics among patients with QT prolongation versus a normal QT interval according to the Bazett’s formula. Categorical variables are presented as frequencies and percentages, and continuous variables as mean ± standard deviation if normally distributed and median [interquartile range] if skewed. We calculated the corrected QT interval using the four different formula: Bazett, Fridericia, Framingham and Hodges. Using each correction method, we then estimated associations between the QT interval and incident cardiovascular disease and all-cause mortality using Cox proportional hazards models. Since we had already excluded individuals with a history of heart failure as part of our study design, the analysis assessing associations between the QT interval and incident cardiovascular disease required the additional exclusion of patients with any history of CAD. For the all-cause mortality analysis, we did not exclude individuals with CAD at study baseline.
Initially, we assessed associations by QT tertiles allowing the lowest tertile to serve as the reference category. Then we assessed associations by the presence of prolonged QT defined as ≥450 ms in men and ≥ 460 ms in women. Potential confounding variables in the multivariable model included age, gender, race, diabetes, hypertension. For the all-cause mortality endpoint, we added CAD as an additional confounding variable in the multivariable analysis. Finally, we assessed the risk of incident cardiovascular disease and all-cause mortality among individuals who were classified as having a prolonged, Bazett-corrected QT interval and normal QTc according to one of the other formulas. All P values are 2-tailed, with P <0.05 indicating statistical significance. Analyses were performed using SPSS version 20.0 (IBM Corporation, Armonk, New York, USA).
Results
In the PAFF study of 6,723 individuals with sinus tachycardia, women comprised 55% of the population. In addition, nearly half of all participants in our study were Black (Table 1). Compared to patients with a normal QTc interval, which was calculated using the traditional Bazett criteria, patients with QTc prolongation were more likely to be older and men. They were also more likely to have a history of hypertension, coronary artery disease, diabetes, and chronic kidney disease.
Table 1.
Baseline characteristics of Individuals with Sinus Tachycardia from the PAFF cohort
| Parameter | Total population | Prolonged QTc* | Normal QTc |
|---|---|---|---|
| No. of participants, n (%) | 6,723 (100) | 2,593 (39) | 4,130 (61) |
| Age (years), median [IQR] | 46 [31, 59] | 50 [37, 62] | 43 [28, 57] |
| Female, n (%) | 3,769 (55) | 1,257 (48) | 2,512 (61) |
| Black, n (%) | 3,211 (47) | 1,154 (45) | 2,057 (50) |
| Hypertension, n (%) | 1,838 (27) | 983 (40) | 855 (21) |
| Coronary artery disease, n (%) | 1,837 (27) | 827 (32) | 1010 (24) |
| Diabetes, n (%) | 792 (12) | 390 (15) | 402 (9.7) |
| Valvular heart disease, n (%) | 126 (2) | 51 (2) | 75 (2) |
| Chronic kidney disease, n (%) | 256 (4) | 127 (5) | 129 (3) |
| Hemoglobin (mg/dl)±SD | 12.2±2.5 | 12.3±2.5 | 12.1±2.5 |
| ECG metrics | |||
| Left ventricular hypertrophy, n (%) | 453 (7) | 200 (8) | 253 (6) |
| Left bundle branch block, n (%) | 87 (1) | 56 (2) | 31 (1) |
| Right bundle branch block, n (%) | 447 (7) | 302 (12) | 145 (4) |
| Heart rate (beats/min) ±SD | 114±13 | 113±12 | 114±14 |
| QT (ms) ±SD | 329±33 | 353±37 | 312±24 |
| QTc Bazett correction (ms) ±SD | 450±36 | 483±31 | 430±20 |
The QTc was corrected using the Bazett’s method, and a prolonged QTc was defined as ≥450ms (males) or ≥460ms (females).
IQR, interquartile range; SD, standard deviation
Calculation of the corrected QT interval with either one of the four equations resulted in a normal distribution; however, the mean QTc and subsequent number of individuals diagnosed with a prolonged QT interval varied by each formula (Figures 2A and 2B). The mean ± standard deviation QTc was 450±36 ms by Bazett, 405±33 ms by Fridericia, 400±28 ms by Framingham, and 423±27 ms by Hodges. A diagnosis of QT prolongation, defined as ≥450 ms in men and ≥460 ms in women, varied according to the criteria used: 2,593 (39%) patients had QT prolongation using Bazett’s, 419 (6.2%) using Fridericia’s, 246 (3.7%) using Framingham’s, and 583 (8.7%) using Hodges’ criteria.
Figure 2.
Distribution of the Corrected QT Interval in Men (Figure 2A) and Women (Figure 2B). The vertical line at 450 ms in men and 460 ms in women demarcates the cutoff for long QT.
After a median follow-up of 4.5 [interquartile range (IQR) 1.9, 6.4] years, there were 495 incident cardiovascular events (annualized rate 2.3%) and 629 deaths (annualized mortality rate 2.4%). Among the 4,891 patients in our cohort without any baseline history of coronary heart disease or heart failure, there was an increase in the rate of cardiovascular events across QTc tertiles of all four correction formulas (Figure 3A). After multivariable adjustment, the highest QTc tertile remained an independent risk marker for incident cardiovascular events using all four equations (Table 2A). The corrected QT interval calculated using the Hodges formula provided the largest gradient in risk. Specifically, the highest tertile was independently associated with a 65% higher incidence of cardiovascular disease.
Figure 3.
Corrected QT Intervals and Cardiovascular Event Rates (Figure 3A) and All-Cause Mortality Rates (Figure 3B). Figure 3A: The rate (% per year) of incident cardiovascular events is higher in tertile 3 compared to tertile 1 for all QT correction methods (P<0.001). Figure 3B: Mortality rate is higher in tertile 3 compared to tertile 1 for the Hodges correction formula only (P=0.001; all other P>0.05).
Table 2A.
Association between Corrected QT Interval and Incident Cardiovascular Disease
| Cardiovascular event | QTc interval | # events/# at risk | Unadjusted HR (95% CI) | P value | Adjusted HR*(95% CI) | P value |
|---|---|---|---|---|---|---|
| Bazett | ||||||
| Tertile 1 | <435 ms | 142/1,696 | 1.00 (ref) | 1.00 (ref) | ||
| Tertile 2 | 435–460 ms | 154/1,677 | 1.18 (0.94–1.48) | 0.2 | 1.13 (0.90–1.42) | 0.3 |
| Tertile 3 | >460 ms | 200/1,518 | 1.78 (1.43–2.21) | <0.001 | 1.60 (1.28–1.99) | <0.001 |
| Fridericia | ||||||
| Tertile 1 | <391 ms | 139/1,677 | 1.00 (ref) | 1.00 (ref) | ||
| Tertile 2 | 391–415 ms | 157/1,690 | 1.17 (0.93–1.47) | 0.2 | 1.13 (0.90–1.42) | 0.3 |
| Tertile 3 | >415 ms | 199/1,524 | 1.70 (1.36–2.11) | <0.001 | 1.52 (1.22–1.90) | <0.001 |
| Framingham | ||||||
| Tertile 1 | <389 ms | 138/1,672 | 1.00 (ref) | 1.00 (ref) | ||
| Tertile 2 | 389–409 ms | 158/1,692 | 1.17 (0.93–1.47) | 0.2 | 1.15 (0.91–1.44) | 0.2 |
| Tertile 3 | >409 ms | 200/1,530 | 1.65 (1.33–2.06) | <0.001 | 1.49 (1.20–1.87) | <0.001 |
| Hodges | ||||||
| Tertile 1 | <411 ms | 143/1,721 | 1.00 (ref) | 1.00 (ref) | ||
| Tertile 2 | 411–429 ms | 153/1,653 | 1.25 (0.99–1.57) | 0.06 | 1.24 (0.99–1.57) | 0.06 |
| Tertile 3 | >429 ms | 198/1,507 | 1.82 (1.46–2.26) | <0.001 | 1.65 (1.33–2.06) | <0.001 |
Adjusted for age, gender, race, diabetes and hypertension.
In the full cohort of 6,723 patients, only the Hodges correction demonstrated an increase in mortality rates across tertiles (Figure 3B). The other three correcting formulas including Bazett, Framingham, and Friderica demonstrated V-shaped mortality curves across the range of corrected QT intervals. Further, only the Hodges correction method resulted in a QTc that was an independent risk marker for all-cause mortality (Table 2B). The highest tertile was comprised of over 2,200 patients that had a 26% higher risk of death compared to patients in the lowest tertile. None of the other three methods identified a higher risk population in either unadjusted or adjusted analyses.
Table 2B.
Association between Corrected QT Interval and Death
| Cardiovascular event | QTc interval | # events/# at risk | Unadjusted HR (95% CI) | P value | Adjusted HR*(95% CI) | P value |
|---|---|---|---|---|---|---|
| Bazett | ||||||
| Tertile 1 | <435 ms | 211/2,239 | 1.00 (ref) | 1.00 (ref) | ||
| Tertile 2 | 435–460 ms | 180/2,243 | 0.88 (0.71–1.08) | 0.2 | 0.85 (0.69–1.04) | 0.1 |
| Tertile 3 | >460 ms | 238/2,236 | 1.17 (0.96–1.43) | 0.1 | 1.07 (0.88–1.31) | 0.5 |
| Fridericia | ||||||
| Tertile 1 | <391 ms | 223/2,241 | 1.00 (ref) | 1.00 (ref) | ||
| Tertile 2 | 391–415 ms | 175/2,238 | 0.78 (0.63–0.96) | 0.02 | 0.74 (0.60–0.91) | 0.004 |
| Tertile 3 | >415 ms | 231/2,240 | 1.07 (0.88–1.30) | 0.5 | 0.98 (0.81, 1.19) | 0.8 |
| Framingham | ||||||
| Tertile 1 | <389 ms | 228/2,242 | 1.00 (ref) | 1.00 (ref) | ||
| Tertile 2 | 389–409 ms | 177/2,251 | 0.76 (0.62–0.94) | 0.01 | 0.72 (0.59–0.89) | <0.01 |
| Tertile 3 | >409 ms | 224/2,230 | 1.03 (0.85–1.25) | 0.7 | 0.95 (0.81–1.16) | 0.6 |
| Hodges | ||||||
| Tertile 1 | <411 ms | 177/2,236 | 1.00 (ref) | 1.00 (ref) | ||
| Tertile 2 | 411–429 ms | 209/2,233 | 1.20 (0.97–1.48) | 0.09 | 1.14 (0.93–1.41) | 0.2 |
| Tertile 3 | >429 ms | 242/2,238 | 1.38 (1.13–1.70) | 0.002 | 1.26 (1.03–1.55) | 0.03 |
Adjusted for age, gender, race, diabetes, hypertension and coronary artery disease.
When assessing risk in the minority of patients that met criteria for prolonged QTc interval (defined as ≥450 ms in men or ≥460 ms in women), we detected independent associations between prolonged QTc interval and cardiovascular events and mortality using all formulas (Tables 3A and 3B). The only exception was that a Bazett-corrected prolonged QTc was not an independent marker of mortality risk compared to patients with a normal QTc. Further, compared to the Fridericia, Hodges, and Framingham formulas, the Bazett’s correction method identified more than 2,000 additional tachycardic patients (30% of the cohort) with prolonged QTc. The majority of individuals with a prolonged, Bazett-corrected QT interval had a normal QTc when assessed by one of the other formulas (Figure 4). Individuals who had a prolonged QTc using Bazett’s and a normal QTc by either Fridericia’s or Framingham’s correction method continued to have a greater than 50% incidence of cardiovascular disease when compared to those with a normal QTc according to both Bazett’s and either Fridericia’s or Framingham’s. The subgroup of individuals with a prolonged Bazett QTc and normal Hodges QTc did not have a significantly increased risk of cardiovascular events when compared to the normal Bazett and Hodges QTc. Additional classification of the prolonged, Bazett-corrected QT interval patients to a normal QTc according to any of the 3 other formulas did not result in an increased risk of death.
Table 3A.
Association between Prolonged QTc* and Incident Cardiovascular Disease
| Cardiovascular event | # events/# at risk | Unadjusted HR (95% CI) | P value | Adjusted† HR (95% CI) | P value |
|---|---|---|---|---|---|
| Bazett | |||||
| Normal QT | 267/3,120 | 1.00 (ref) | 1.00 (ref) | ||
| Prolonged QTc | 229/1,774 | 1.73 (1.45–2.07) | <0.001 | 1.38 (1.15–1.66) | <0.001 |
| Fridericia | |||||
| Normal QT | 437/4,635 | 1.00 (ref) | 1.00 (ref) | ||
| Prolonged QTc | 59/259 | 2.40 (1.82–3.18) | <0.001 | 1.78 (1.34–2.36) | <0.001 |
| Framingham | |||||
| Normal QT | 455/4,743 | 1.00 (ref) | 1.00 (ref) | ||
| Prolonged QTc | 41/151 | 2.74 (1.97–3.80) | <0.001 | 1.95 (1.40–2.72) | <0.001 |
| Hodges | |||||
| Normal QT | 418/4,524 | 1.00 (ref) | 1.00 (ref) | ||
| Prolonged QTc | 78/370 | 2.51 (1.96–3.21) | <0.001 | 1.86 (1.45–2.40) | <0.001 |
The QTc was corrected using the Bazett’s method, and a prolonged QTc was defined as ≥450ms (males) or ≥460ms (females).
Adjusted for age, gender, race, diabetes and hypertension.
Table 3B.
Association between Prolonged QTc* and Death
| All-cause mortality | # events/# at risk | Unadjusted HR (95% CI) | P value | Adjusted† HR (95% Cl) | P value |
|---|---|---|---|---|---|
| Bazett | |||||
| Normal QT | 354/4,130 | 1.00 (ref) | 1.00 (ref) | ||
| Prolonged QTc | 275/2,593 | 1.29 (1.09–1.53) | 0.003 | 0.94 (0.79–1.11) | 0.5 |
| Fridericia | |||||
| Normal QT | 563/6,304 | 1.00 (ref) | 1.00 (ref) | ||
| Prolonged QTc | 66/419 | 1.99 (1.52–2.62) | <0.001 | 1.59 (1.20–2.09) | <0.001 |
| Framingham | |||||
| Normal QT | 587/6,477 | 1.00 (ref) | 1.00 (ref) | ||
| Prolonged QTc | 42/246 | 2.20 (1.60–3.05) | <0.001 | 1.70 (1.22–2.37) | <0.001 |
| Hodges | |||||
| Normal QT | 541/6,140 | 1.00 (ref) | 1.00 (ref) | ||
| Prolonged QTc | 88/583 | 1.84 (1.45–2.35) | <0.001 | 1.50 (1.17–1.91) | <0.001 |
The QTc was corrected using the Bazett’s method, and a prolonged QTc was defined as ≥450ms (males) or ≥460ms (females).
Adjusted for age, gender, race, diabetes, hypertension, and coronary artery disease.
Figure 4.
Cardiovascular event and mortality risk among individuals with a QTc that was prolonged by the Bazett criteria only. The hazard ratio (HR) assesses the risk of death among individuals with prolonged QT interval by the Bazett correction method and a normal QT interval using the Hodges, Fridericia, or Framingham formulas compared to individuals with a normal QT interval by Bazett and either one of the other formulas. For the outcome of incident cardiovascular disease, the multivariable model adjusts for age, gender, race, diabetes, and hypertension. For the outcome of death, the multivariable model adjusts for these same variables plus coronary artery disease.
Discussion
Our findings demonstrate that the Bazett method over-diagnoses prolonged QTc by identifying nearly 40% of sinus tachycardia patients. Most of these individuals have prolonged QT interval by the Bazett’s correction method only and did not have an increased risk of death compared to those with a normal QT interval. In addition, the Hodges formula displayed the steepest gradient in cardiovascular risk and was the only method that demonstrated an increase in all-cause mortality across the range of QTc values. It identified nearly one-third of the cohort patients that had a 26% higher and independent risk of death than those in the lowest tertile. Compared to the Hodges’ correction method, the Bazett’s formula demonstrated a similar increase in cardiovascular risk. However, none of the other formulas including Bazett, Fridericia or Framingham demonstrated a gradient in mortality risk across the entire range of QT intervals. The Fridericia and Framingham correction formulas identified only 5% of our cohort with prolonged QT interval and higher risk of death compared to the respective group of patients with normal QTc. As a result, these formulas have limited ability to stratify risk across the range of QTc.
Computerized ECG software, which utilizes Bazett’s formula, often overestimates the corrected QT interval in patients with sinus tachycardia.17, 18 Approximately 2,000 patients in our study had a prolonged QT based on the Bazett formula only. These individuals did not have a higher risk of death compared to those with a normal QT and suggests that Bazett’s has poor specificity in identifying a group of sinus tachycardia patients at higher risk for death. In addition, the Bazett formula did not provide any additional information on cardiovascular risk when used in conjunction with the Hodges’ correction method. The Bazett correction for tachycardia was formulated based on empiric observations of the duration of systole and diastole at various heart rates in animal models and humans.10 However, the original investigation included just a limited number of subjects, and only three out of 42 male subjects and one out of 21 female subjects had resting tachycardia. Its hyperbolic association with heart rate makes the Bazett formula ill-suited for accurate measurement of the QT during tachycardia.29–33 Our study is one of the largest studies in patients with sinus tachycardia to demonstrate that the Bazett correction results in an overestimation of cardiac repolarization abnormalities that are not associated with an increased risk of death.
The corrected QT interval is clinically relevant only when it is a proven risk marker for adverse events. The various QT correction formulas differ in their ability to identify high-risk subgroups. Of all four formulas assessed, only the Hodges’ method demonstrated an increase in risk of incident cardiovascular disease and death across the range of QTc values. Both Fridericia and Framingham methods revealed V-shaped mortality curves suggesting a protective effect of the QTc in the second tertile. This finding for all-cause death, however, is discordant with the higher risk of incident cardiovascular disease in this same tertile. Our findings also suggest that the Hodges’ method was the only one that could simply stratify the approximately 2600 patients with a Bazett’s prolonged QTc into those at increased cardiovascular or mortality risk. As such, the QTc-Hodges should be assessed routinely in patients with sinus tachycardia. In patients with a longer QTc, avoidance of QT prolonging medicines or safeguarding against drug interactions that may increase the QT interval may be potential interventions that reduce the risk for arrhythmic complications. 34 Future work should also determine whether a wider spectrum of Hodges-corrected QTc values, and not just the traditional long QT cutoffs, could be incorporated into clinical risk scores.
Several limitations of our study should be considered. We used automated software for measurement of QT. Though measurements were over-read by cardiologists, it is unclear how many of the automated QT measurements were modified after personal inspection. However, all measurements were made similarly across the large study population, which diminishes the influence of errors in measurement. Also, long QT is associated with arrhythmic sudden cardiac death, and our primary endpoint was all-cause mortality. Our study design precluded a thorough evaluation and adjudication of the modes of death in the UPHS patient population. However, the strong correlations with incident cardiovascular disease suggest that the QTc may also be a marker for other adverse pathways that result in non-arrhythmic complications. In addition, our study design utilized administrative data collection within a large healthcare network. As a result, we may not have collected information on all the potential confounders.
Conclusions
All 4 formulas demonstrated an increase in the risk of incident cardiovascular disease including both CAD and CHF in our cohort. The Hodges formula identified the steepest gradient across the range of QTc values and identified up to one third of the cohort at risk for mortality. Further validation studies will be important to assess the routine use of the Hodges formula by automated computer algorithms especially among individuals with sinus tachycardia.
Supplementary Material
Clinical Perspectives.
Prolongation of the QT interval, which is a measure of cardiac repolarization, is associated with adverse cardiovascular events and death. The repolarization period is a function of the underlying heart rate and decreases as the ventricular rate rises. Several methods have been proposed for correcting the QT interval. However, limited studies have compared the corrected QT interval according to the different methods in patients with sinus tachycardia. In a large cohort of individuals with sinus tachycardia, we evaluated the corrected QT intervals from four different formulas (Bazett, Fridericia, Framingham, and Hodges) as a risk marker for incident cardiovascular events and death. The linear Hodges formula provided the best correlation with cardiovascular outcomes and all-cause mortality. In addition, the traditional Bazett-correction formula overestimated the QTc and identified nearly half of all sinus tachycardia patients with a prolonged QT interval. Automated ECG reading algorithms that incorporate the Hodges formula during sinus tachycardia could offer a more accurate measurement of QTc. This method may help to identify patients at higher risk for cardiovascular events or death.
Acknowledgments
This work was supported in part by the F. Harlan Batrus Research Fund, the Murray and Susan Bloom Research Fund, and grant K23DK089118 from the National Institutes of Health (RD). The sponsors had no role in the study design, actual research, or manuscript preparation. The authors would like to thank the Penn Data Store for assistance in assembling the information used in this study. PJP and RD had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Abbreviations
- AF
Atrial Fibrillation
- PAFF
Penn Atrial Fibrillation Free
- HF
Heart Failure
- CAD
Coronary Artery Disease
- UPHS
University of Pennsylvania Health System
- ECG
Electrocardiogram
- ICD-9
International Classification of Diseases, version 9
Footnotes
Conflicts of interest: None.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Arsenos P, Gatzoulis KA, Dilaveris P, Gialernios T, Sideris S, Lazaros G, Archontakis S, Tsiachris D, Kartsagoulis E, Stefanadis C. The rate-corrected QT interval calculated from 24-hour Holter recordings may serve as a significant arrhythmia risk stratifier in heart failure patients. International journal of cardiology. 2011 Mar 3;147:321–323. doi: 10.1016/j.ijcard.2010.12.076. [DOI] [PubMed] [Google Scholar]
- 2.Balasubramaniyam N, Palaniswamy C, Aronow WS, Khera S, Balasubramanian G, Harikrishnan P, Doshi JV, Nabors C, Peterson SJ, Sule S. Association of corrected QT interval with long-term mortality in patients with syncope. Archives of medical science : AMS. 2013 Dec 30;9:1049–1054. doi: 10.5114/aoms.2013.39383. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Buckingham TA, Bhutto ZR, Telfer EA, Zbilut J. Differences in corrected QT intervals at minimal and maximal heart rate may identify patients at risk for torsades de pointes during treatment with antiarrhythmic drugs. Journal of cardiovascular electrophysiology. 1994 May;5:408–411. doi: 10.1111/j.1540-8167.1994.tb01179.x. [DOI] [PubMed] [Google Scholar]
- 4.Chiladakis J, Kalogeropoulos A, Zagkli F, Koutsogiannis N, Chouchoulis K, Alexopoulos D. Predicting torsade de pointes in acquired long QT syndrome: optimal identification of critical QT interval prolongation. Cardiology. 2012;122:3–11. doi: 10.1159/000338345. [DOI] [PubMed] [Google Scholar]
- 5.Jons C, Moss AJ, Goldenberg I, Liu J, McNitt S, Zareba W, Qi M, Robinson JL. Risk of fatal arrhythmic events in long QT syndrome patients after syncope. Journal of the American College of Cardiology. 2010 Feb 23;55:783–788. doi: 10.1016/j.jacc.2009.11.042. [DOI] [PubMed] [Google Scholar]
- 6.Straus SM, Kors JA, De Bruin ML, van der Hooft CS, Hofman A, Heeringa J, Deckers JW, Kingma JH, Sturkenboom MC, Stricker BH, Witteman JC. Prolonged QTc interval and risk of sudden cardiac death in a population of older adults. Journal of the American College of Cardiology. 2006 Jan 17;47:362–367. doi: 10.1016/j.jacc.2005.08.067. [DOI] [PubMed] [Google Scholar]
- 7.Vrtovec B, Delgado R, Zewail A, Thomas CD, Richartz BM, Radovancevic B. Prolonged QTc interval and high B-type natriuretic peptide levels together predict mortality in patients with advanced heart failure. Circulation. 2003 Apr 8;107:1764–1769. doi: 10.1161/01.CIR.0000057980.84624.95. [DOI] [PubMed] [Google Scholar]
- 8.Vrtovec B, Knezevic I, Poglajen G, Sebestjen M, Okrajsek R, Haddad F. Relation of B-type natriuretic peptide level in heart failure to sudden cardiac death in patients with and without QT interval prolongation. The American journal of cardiology. 2013 Mar 15;111:886–890. doi: 10.1016/j.amjcard.2012.11.041. [DOI] [PubMed] [Google Scholar]
- 9.Dekker JM, Crow RS, Hannan PJ, Schouten EG, Folsom AR, Study A. Heart rate-corrected QT interval prolongation predicts risk of coronary heart disease in black and white middle-aged men and women: the ARIC study. Journal of the American College of Cardiology. 2004 Feb 18;43:565–571. doi: 10.1016/j.jacc.2003.09.040. [DOI] [PubMed] [Google Scholar]
- 10.Bazett HC. An Analysis of the Time-Relations of the Electrocardiogram. Heart. 1920;VII:353–370. [Google Scholar]
- 11.Fridericia LS. Die Systolendauer im Elektrokardiogramm bei normalen Menschen und bei Herzkranken. Acta Medica Scandinavica. 1920;53:469–486. [Google Scholar]
- 12.Hodges M, Salerno DM, Erlein D. Bazett QT Correction Reviewed - Evidence that a Linear Correction for Heart Rate is Better. Journal of the American College of Cardiology. 1983 Jan;:1. [Google Scholar]
- 13.Kovacs SJ., Jr The duration of the QT interval as a function of heart rate: a derivation based on physical principles and a comparison to measured values. Am Heart J. 1985 Oct;110:872–878. doi: 10.1016/0002-8703(85)90472-7. [DOI] [PubMed] [Google Scholar]
- 14.Sagie A, Larson MG, Goldberg RJ, Bengtson JR, Levy D. An improved method for adjusting the QT interval for heart rate (the Framingham Heart Study) The American journal of cardiology. 1992 Sep 15;70:797–801. doi: 10.1016/0002-9149(92)90562-d. [DOI] [PubMed] [Google Scholar]
- 15.Hodges M. Rate Correction of the QT Interval. Cardiac Electrophysiology Review 1997/09/01. 1997;1:360–363. [Google Scholar]
- 16.Fridericia LS. The duration of systole in an electrocardiogram in normal humans and in patients with heart disease. 1920. Ann Noninvasive Electrocardiol. 2003 Oct;8:343–351. doi: 10.1046/j.1542-474X.2003.08413.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Chiladakis J, Kalogeropoulos A, Arvanitis P, Koutsogiannis N, Zagli F, Alexopoulos D. Preferred QT correction formula for the assessment of drug-induced QT interval prolongation. Journal of cardiovascular electrophysiology. 2010 Aug 1;21:905–913. doi: 10.1111/j.1540-8167.2010.01738.x. [DOI] [PubMed] [Google Scholar]
- 18.Luo S, Michler K, Johnston P, Macfarlane PW. A comparison of commonly used QT correction formulae: the effect of heart rate on the QTc of normal ECGs. Journal of electrocardiology. 2004;37(Suppl):81–90. doi: 10.1016/j.jelectrocard.2004.08.030. [DOI] [PubMed] [Google Scholar]
- 19.Zabel M, Franz MR, Klingenheben T, Mansion B, Schultheiss HP, Hohnloser SH. Rate-dependence of QT dispersion and the QT interval: comparison of atrial pacing and exercise testing. Journal of the American College of Cardiology. 2000 Nov 1;36:1654–1658. doi: 10.1016/s0735-1097(00)00921-9. [DOI] [PubMed] [Google Scholar]
- 20.Noseworthy PA, Peloso GM, Hwang SJ, Larson MG, Levy D, O’Donnell CJ, Newton-Cheh C. QT interval and long-term mortality risk in the Framingham Heart Study. Annals of noninvasive electrocardiology : the official journal of the International Society for Holter and Noninvasive Electrocardiology, Inc. 2012 Oct;17:340–348. doi: 10.1111/j.1542-474X.2012.00535.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Jouven X, Empana JP, Schwartz PJ, Desnos M, Courbon D, Ducimetiere P. Heart-rate profile during exercise as a predictor of sudden death. N Engl J Med. 2005 May 12;352:1951–1958. doi: 10.1056/NEJMoa043012. [DOI] [PubMed] [Google Scholar]
- 22.Diaz A, Bourassa MG, Guertin MC, Tardif JC. Long-term prognostic value of resting heart rate in patients with suspected or proven coronary artery disease. Eur Heart J. 2005 May;26:967–974. doi: 10.1093/eurheartj/ehi190. [DOI] [PubMed] [Google Scholar]
- 23.Opdahl A, Ambale Venkatesh B, Fernandes VR, Wu CO, Nasir K, Choi EY, Almeida AL, Rosen B, Carvalho B, Edvardsen T, Bluemke DA, Lima JA. Resting heart rate as predictor for left ventricular dysfunction and heart failure: MESA (Multi-Ethnic Study of Atherosclerosis) J Am Coll Cardiol. 2014 Apr 1;63:1182–1189. doi: 10.1016/j.jacc.2013.11.027. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Breidthardt T, Christ M, Matti M, Schrafl D, Laule K, Noveanu M, Boldanova T, Klima T, Hochholzer W, Perruchoud AP, Mueller C. QRS and QTc interval prolongation in the prediction of long-term mortality of patients with acute destabilised heart failure. Heart. 2007 Sep;93:1093–1097. doi: 10.1136/hrt.2006.102319. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Watanabe E, Yasui K, Kamiya K, Yamaguchi T, Sakuma I, Honjo H, Ozaki Y, Morimoto S, Hishida H, Kodama I. Upregulation of KCNE1 induces QT interval prolongation in patients with chronic heart failure. Circulation journal : official journal of the Japanese Circulation Society. 2007 Apr;71:471–478. doi: 10.1253/circj.71.471. [DOI] [PubMed] [Google Scholar]
- 26.Domenighetti AA, Boixel C, Cefai D, Abriel H, Pedrazzini T. Chronic angiotensin II stimulation in the heart produces an acquired long QT syndrome associated with IK1 potassium current downregulation. Journal of molecular and cellular cardiology. 2007 Jan;42:63–70. doi: 10.1016/j.yjmcc.2006.09.019. [DOI] [PubMed] [Google Scholar]
- 27.Healthcare G. Marquette 12SL ECG Analysis Program: Physician’s Guide. General Electric Company; 2010. Revision A ed. [Google Scholar]
- 28.Hodges M, Salerno D, Erlien D. BAZETT QT CORRECTION REVIEWED-EVIDENCE THAT A LINEAR QT CORRECTION FOR HEART-RATE IS BETTER. Paper presented at: Journal of the American College of Cardiology; 1983. [Google Scholar]
- 29.Rautaharju PM, Surawicz B, Gettes LS, et al. AHA/ACCF/HRS recommendations for the standardization and interpretation of the electrocardiogram: part IV: the ST segment, T and U waves, and the QT interval: a scientific statement from the American Heart Association Electrocardiography and Arrhythmias Committee, Council on Clinical Cardiology; the American College of Cardiology Foundation; and the Heart Rhythm Society. Endorsed by the International Society for Computerized Electrocardiology Journal of the American College of Cardiology. 2009 Mar 17;53:982–991. doi: 10.1016/j.jacc.2008.12.014. [DOI] [PubMed] [Google Scholar]
- 30.Ahnve S, Vallin H. Influence of heart rate and inhibition of autonomic tone on the QT interval. Circulation. 1982 Mar;65:435–439. doi: 10.1161/01.cir.65.3.435. [DOI] [PubMed] [Google Scholar]
- 31.Milne JR, Camm AJ, Ward DE, Spurrell RA. Effect of intravenous propranolol on QT interval. A new method of assessment. British heart journal. 1980 Jan;43:1–6. doi: 10.1136/hrt.43.1.1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Chiladakis J, Kalogeropoulos A, Arvanitis P, Koutsogiannis N, Zagli F, Alexopoulos D. Heart rate-dependence of QTc intervals assessed by different correction methods in patients with normal or prolonged repolarization. Pacing and clinical electrophysiology : PACE. 2010 May;33:553–560. doi: 10.1111/j.1540-8159.2009.02657.x. [DOI] [PubMed] [Google Scholar]
- 33.Karjalainen J, Viitasalo M, Manttari M, Manninen V. Relation between QT intervals and heart rates from 40 to 120 beats/min in rest electrocardiograms of men and a simple method to adjust QT interval values. Journal of the American College of Cardiology. 1994 Jun;23:1547–1553. doi: 10.1016/0735-1097(94)90654-8. [DOI] [PubMed] [Google Scholar]
- 34.Sauer AJ, Moss AJ, McNitt S, et al. Long QT syndrome in adults. Journal of the American College of Cardiology. 2007 Jan 23;49:329–337. doi: 10.1016/j.jacc.2006.08.057. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.