Abstract
The relationship between corrected QT (QTc) intervals and short-term clinical outcomes remains insufficiently investigated. This study examined correlations between QTc intervals and risks of arrhythmia or mortality over a 3-month follow-up period. This retrospective study analyzed de-identified electronic medical records from the TriNetX Network. Adult patients with at least 2 electrocardiogram assessments were categorized into 5 QTc interval groups: long QTc (>500 milliseconds [ms]), borderline QTc (460–500 ms), high-normal QTc (420–460 ms), control QTc (370–420 ms), and short QTc (200–370 ms). Primary outcomes were 3-month risks of new-onset atrial fibrillation (AF) or ventricular arrhythmia (VA), with secondary focus on 3-month mortality risk. Among 145,111 patients, a U-shaped pattern was observed in AF risk across QTc interval categories. The hazard ratio for AF risk was 7.384 (95% confidence interval: 5.911–9.224) in the long QTc group and 7.735 (95% confidence interval: 4.237–14.122) in the short QTc group compared to controls. Similar U-shaped correlations were observed between QTc intervals and VA or mortality risks. Sensitivity analyses confirmed the U-shaped association for AF. However, sensitivity analyses showed that the association between short QTc intervals and the risk of VA or mortality was attenuated and no longer statistically significant after excluding patients exposed to QTc-prolonging medications or in pre-pandemic cohorts. Our findings revealed a U-shaped correlation between QTc intervals and the occurrence of AF, VA, and mortality within a 3-month observation period, underscoring the potential of QTc intervals as short-term prognostic predictors.
Keywords: arrhythmia, atrial fibrillation, corrected QT interval, mortality, sudden cardiac death, ventricular arrhythmia
1. Introduction
Corrected QT (QTc) prolongation constitutes a cardiac repolarization disorder that can arise congenitally or be acquired owing to various disease processes and medications.[1] While acquired QTc prolongation frequently resolves after addressing the underlying cause, long-term treatment is generally unnecessary.[2] In cases of congenital QTc prolongation, management approaches typically involve oral β-adrenergic blockers, left cervicothoracic sympathectomy, or the implementation of an automatic implanted cardioverter-defibrillator.[3] This condition has been linked to torsades de pointes, a critical ventricular arrhythmia (VA), and serves as an independent predictor of overall mortality in both sexes.[4,5] Prolonged QTc intervals are frequent in hospitalized patients. In a previous study involving 258 patients admitted to a general medical service, analysis of admission electrocardiograms (ECGs) showed that the prevalence of abnormal QTc intervals displayed variability on the basis of the applied criteria, with rates reaching up to 25.2%.[6] The prevalence of prolonged QTc among patients in the surgical intensive care unit is notable, particularly among postoperative surgical intensive care unit patients, where it may reach up to 67%.[7]
Several studies reported that QTc prolongation has been linked to unfavorable cardiovascular outcomes, including arrhythmias, sudden death, and coronary heart diseases.[8–10] Consistently, in a recent meta-analysis that encompassed 36 studies conducted on the general population, where the follow-up duration ranged from 3 to 30 years, a statistically significant correlation was identified between QTc prolongation and a spectrum of cardiovascular conditions, including overall cardiovascular disease, coronary heart diseases, stroke, sudden cardiac death, and atrial fibrillation (AF).[8] This finding highlights the potential impact of QTc prolongation on long-term outcomes.[8] Focusing on short-term outcomes, a previous study involving 1558 patients in the emergency department showed that among those with QTc prolongation, a 5% mortality rate was observed either during their emergency department stay or hospitalization.[11] A study involving 980 patients with QTc intervals of ≥ 500 milliseconds (ms), compared to an age- and sex-matched cohort with QTc intervals of < 500 ms, showed an elevated 30-day all-cause mortality risk (hazard ratio [HR]: 1.90) while adjusting for relevant variables.[12] These findings have suggested the harmful effects of QT prolongation on short-term outcomes, which remains insufficiently investigated and is a subject that receives limited attention in current research endeavors.
Considering the limited sample sizes in previous studies, a more expansive study could enhance the dependability of assessing the impact of QTc prolongation on short-term outcomes. Therefore, utilizing the TriNetX (Cambridge) platform, our study aimed to comprehensively evaluate the correlation between QTc intervals and the risk of new-onset arrhythmias, with a secondary focus on exploring the association between QTc intervals and overall mortality during a 3-month follow-up period.
2. Methods
2.1. Study design, data source, and ethical statement
This retrospective study analyzed de-identified electronic medical records from the TriNetX Network, which includes data from 122 million individuals in 104 U.S. healthcare organizations. Covering a broad range of healthcare settings from academic institutions to community hospitals, the dataset encompasses diagnoses, procedures, medications, laboratory results, and genetic information. All data provided by TriNetX are fully de-identified in accordance with Health Insurance Portability and Accountability Act standards, and no author had access to personally identifiable information during or after data collection. TriNetX, a federated research network, operates under a waiver of informed consent approved by the Western Institutional Review Board, as it exclusively provides de-identified counts and statistical summaries. The robustness and reliability of the TriNetX database have been demonstrated in multiple peer-reviewed studies employing this resource for real-world evidence.[13–16] This study was conducted in alignment with the principles outlined in the Declaration of Helsinki. Ethical approval for the use of TriNetX in this research was obtained from the Chi Mei Medical Center Institutional Review Board (IRB Serial No. 11206–E01), ensuring compliance with ethical research standards. The data used in this retrospective study were accessed through the TriNetX Research Network on October 25, 2023.
2.2. Patient population
The selected individuals were required to have undergone ECG testing on a minimum of 2 occasions. We specifically included individuals who had either documented visits within 1 month before the testing or had revisited medical facilities within 3 months after the testing to ensure patients’ unrestricted access to medical resources. The clinical settings where these ECGs were performed varied across the TriNetX network and included outpatient clinics, emergency departments, inpatient wards, and critical care units. We excluded the following conditions that occurred within the 3-year period before ECG assessment: a QRS duration of ≥ 120 ms (Logical Observation Identifiers Names And Codes: 8633-0), paroxysmal AF (International Classification of Diseases, 10th Revision [ICD-10] I48.0), persistent AF (ICD-10 I48.1), chronic AF (ICD-10 I48.2), ventricular premature depolarization (ICD-10 I49.3), procedures involving pacemaker or implantable defibrillator (Current Procedural Terminology 1006075), heart surgeries (Systematized Nomenclature of Medicine Clinical Terms 64915003), heart transplant status (ICD-10 Z94.1), ventricular tachycardia (ICD-10 I47.2), and ventricular fibrillation and flutter (ICD-10 I49.0). QT interval correction was performed using either Bazett formula or Fridericia formula. Enrolled patients were subsequently categorized into the following 5 distinct groups on the basis of their QTc interval measurements: long QTc group (500–600 ms), borderline QTc group (460–500 ms), high-normal QTc group (420–460 ms), control group (370–420 ms), and short QTc group (200–370 ms) (Fig. 1). The upper and lower thresholds were set to capture both clinically significant QTc prolongation (≥500 ms) and marked QTc shortening (≤370 ms), which are associated with increased arrhythmic risk and have been used in prior research.[12,17] Intermediate categories (borderline and high-normal) allowed us to evaluate gradations of risk across the QTc spectrum.
Figure 1.
Flow chart for patient selection.
Stringent criteria were implemented in the borderline, high-normal, and control QTc groups. For example, patients were assigned to the borderline QTc group only if serial ECG examinations consistently indicated QTc intervals between 460 and 500 ms. Patients with a single ECG reading below 460 ms or above 500 ms were excluded from the analysis. Similar criteria were applied to the high-normal and control groups. For the long and short QTc groups, the assignment was based on a minimum of 2 ECG examinations showing QTc intervals between 500 and 600 ms or 200 and 370 ms, respectively, irrespective of other ECG findings. This definition for the long and short QTc groups facilitates prompt identification of patients at heightened risk of verse outcomes, obviating the necessity for extended ECG follow-up.
2.3. Data collection
The 5 distinct cohorts, delineated by varying QTc intervals, were subsequently selected and analyzed online using the analysis tool provided by the TriNetX platform. We identified patients who were at least 20 years old at the time of their ECG examination within the Global Collaborative Network. The primary analysis was conducted using data spanning from January 1, 2018 to December 31, 2022. To evaluate the potential impact of the COVID-19 pandemic on our findings, we performed a sensitivity analysis using a separate cohort that applied the same inclusion and exclusion criteria but was limited to the pre-pandemic period (January 1, 2015–December 31, 2019). This approach allowed us to evaluate whether the observed associations between QTc intervals and outcomes remained consistent across different time periods.
2.4. Propensity score matching
In the process of analysis, we initially identified patients who fulfilled the inclusion and exclusion criteria. Owing to the platform’s constraint of allowing only one-by-one comparisons, each group (i.e., short QTc, high-normal QTc, borderline QTc, and long QTc groups) was matched with the control group separately to assess the relative risk associated with each group (Fig. 1). A 1:1 propensity score matching was conducted in the TriNetX platform to facilitate comparison between groups. Propensity scores were calculated using logistic regression, and “greedy nearest neighbor matching” was used with a caliper of 0.1 pooled standard deviations. The cohorts were matched on the basis of several demographic and clinical factors, including age at the index date, sex, race, and various comorbidities, including hypertension, kidney diseases, ischemic heart diseases, diabetes mellitus (DM), overweight and obesity, heart failure, sleep apnea, nonrheumatic valve disorders, multiple valve diseases, rheumatic valve diseases, atrioventricular and left bundle-branch block, congenital malformations of the circulatory system, and hyperthyroidism. Additionally, blood pressure and body mass index (BMI) were considered during the matching process. Covariate imbalance was identified when the absolute value of standardized mean difference exceeded 0.1. Missing data for covariates used in propensity score matching and outcome analyses were handled using a complete-case approach, in accordance with the capabilities of the TriNetX platform. Only patients with available data for all required covariates were included in the matching and subsequent analyses.
2.5. Outcomes
This study mainly aimed to evaluate and compare the 3-month cumulative probabilities of new-onset AF (ICD-10 I48.0–I48.2, I48.91) and VA, specifically ventricular premature depolarization (ICD-10 I49.3), ventricular tachycardia (ICD-10 I47.2), and ventricular fibrillation, and flutter (ICD-10 I49.0), among different cohorts. The secondary aim of this study involved the comparison of the 3-month cumulative probabilities of all-cause mortality among different groups.
2.6. Subgroup analyses based on gender
To investigate potential gender differences in QTc interval associations with outcomes, we conducted separate analyses for male and female patients. We employed gender-specific QTc definitions to better align with common clinical thresholds. For males, we defined short QT as < 350 ms, normal QT (control) as 350 to 450 ms, long QT as 451 to 500 ms, and very long QT as > 500 ms. For females, we used < 360 ms for short QT, 360 to 460 ms for normal QT (control), 461 to 510 ms for long QT, and > 510 ms for very long QT. Propensity score matching was performed separately within each gender group, maintaining the same matching variables as in the primary analysis. HRs were calculated for all outcomes (AF, VA, mortality) within each gender-specific QTc category.
2.7. Statistical analysis
The definition of statistical significance was established as a 2-sided alpha value of 0.05. HRs were estimated for all outcomes, accompanied by their corresponding 95% confidence intervals (95% CIs) by using Cox proportional hazard models. To perform risk analysis and Kaplan–Meier survival analysis, the TriNetX user interface was employed. Differences between the cohorts in the Kaplan–Meier survival study were assessed by employing a log-rank test.
To assess robustness of our findings, we conducted 3 sensitivity analyses. First, we excluded patients exposed to moderate- to high-risk QTc-prolonging medications within 1 month before ECG assessment. Second, we analyzed patients enrolled during the pre-pandemic period (January 1, 2015–December 31, 2019). Third, we excluded patients with abnormal heart rates during ECG examination (tachycardia > 100 beats per minute (bpm) or bradycardia < 60 bpm), which minimized potential QTc measurement bias that can persist despite correction formulas.[18] For each analysis, we reassessed the relationship between QTc intervals and outcomes to determine whether our main findings remained consistent under these restricted conditions.
3. Results
3.1. Patient selection and baseline demographics
We conducted a comprehensive query within the Global Collaborative Network of the TriNetX platform, encompassing 105 healthcare organizations. Initially, a total of 145,111 patients were identified for each group: the long QTc group (n = 11,261), borderline QTc group (n = 11,329), high-normal group (n = 81,317), control group (n = 39,773), and short QTc group (n = 1431). An overview of the baseline characteristics of patients before the matching process is presented in Table 1 and Table S1, Supplemental Digital Content, https://links.lww.com/MD/P908. Among the groups with prolonged QTc intervals (i.e., long, borderline, and high-normal QTc groups), a preponderance of females was observed, whereas the control and short QTc groups exhibited a male predominance. The body mass index ranged between 28.3 and 30.5 kg/m2. A similar distribution of races was noted among groups, with a more than 50% proportion of whites.
Table 1.
Baseline characteristics of patient with different corrected QT (QTc) intervals.
| Short QTc group (n = 1431) | Control group (n = 39,773) | High-normal QTc group (n = 81,317) | Borderline QTc group (n = 11,329) | Long QTc group (n = 11,261) |
|
|---|---|---|---|---|---|
| Demographics, n (%) or mean ± sd | |||||
| Age (yr) | 53.7 ± 18.7 | 53.2 ± 17.0 | 55.6 ± 16.3 | 58.2 ± 15.3 | 58.0 ± 15.3 |
| Male | 942 (65.8) | 23,997 (60.3) | 29,842 (36.7) | 4027 (35.6) | 5045 (44.8) |
| BMI (kg/m2) | 28.3 ± 6.5 | 29.0 ± 6.2 | 30.5 ± 7.2 | 30.4 ± 7.6 | 28.9 ± 7.5 |
| BMI > 30 (kg/m2) | 232 (16.2) | 7510 (18.9) | 19,329 (23.8) | 2820 (24.9) | 2873 (25.5) |
| Race, n (%) | |||||
| White | 725 (50.7) | 24,560 (61.8) | 53,296 (65.5) | 7285 (64.3) | 6944 (61.6) |
| Others* | 440 (30.8) | 8172 (20.6) | 15,786 (19.4) | 2621 (23.1) | 2938 (26.1) |
| Asian | 57 (4.0) | 1481 (3.7) | 2465 (3.0) | 280 (2.5) | 286 (2.5) |
| Comorbidities, n (%) | |||||
| HTN | 623 (43.5) | 13,744 (34.6) | 33,038 (40.6) | 5351 (47.2) | 6476 (57.5) |
| DM | 259 (18.1) | 4484 (11.3) | 13,227 (16.3) | 2443 (21.6) | 3334 (29.6) |
| IHD | 350 (24.5) | 5065 (12.7) | 8963 (11.0) | 1540 (13.6) | 3604 (32.0) |
| Overweight† | 218 (15.2) | 5398 (13.6) | 15,717 (19.3) | 2247 (19.8) | 2279 (20.2) |
| Sleep apnea | 164 (11.5) | 3783 (9.5) | 8830 (10.9) | 1225 (10.8) | 1495 (13.3) |
| Kidney disease‡ | 324 (22.6) | 2732 (6.9) | 7868 (9.7) | 2160 (19.1) | 5051 (44.9) |
| Heart failure | 175 (12.2) | 750 (1.9) | 2883 (3.6) | 1135 (10) | 3034 (26.9) |
| Thyrotoxicosis | 33 (2.3) | 410 (1.0) | 971 (1.2) | 134 (1.2) | 167 (1.5) |
| Hemodynamic data (mm Hg), mean ± sd | |||||
| SBP | 123.3 ± 26.7 | 128.8 ± 19.0 | 125.7 ± 25.4 | 124.1 ± 31.5 | 124.6 ± 29.5 |
| SBP > 160 | 516 (36.7) | 12,949 (32.6) | 28,768 (35.4) | 4480 (39.6) | 6318 (56.1) |
| DBP | 71.4 ± 13.6 | 74.7 ± 11.5 | 74.9 ± 12.5 | 73.8 ± 14.7 | 70.6 ± 17.6 |
| DBP > 100 | 284 (19.8) | 6810 (17.1) | 16,094 (17.8) | 2808 (24.8) | 4673 (41.5) |
Short QTc group: QTc < 370 ms; control group: QTc 370–420 ms; high-normal QTc group: QTc 420–460 ms; borderline QTc group: QTc 460–500 ms; long QTc: QTc > 500 ms.
BMI = body mass index, DBP = diastolic blood pressure, DM = diabetes mellitus, HTN = hypertension, IHD = ischemic heart diseases, SBP = systolic blood pressure.
Black or African American.
Included obesity.
Acute kidney failure and chronic kidney disease.
The top 3 comorbidities in the control group were hypertension (34.6%), ischemic heart diseases (12.7%), and overweight and obesity (13.6%). In the long and short QTc groups, the top 3 comorbidities were hypertension (57.5% and 43.5%, respectively), ischemic heart diseases (32% and 24.5%, respectively), and kidney diseases (44.9% and 22.6%, respectively). Patients in the long and short QTc groups had more prevalence of comorbidities than those in the control group. Clear U-shaped associations were observed between QTc intervals and the incidence of various comorbidities, including DM, ischemic heart diseases, kidney diseases, and heart failure.
3.2. Primary endpoint: impact of QTc intervals on the cumulative incidence of arrhythmias at 3-month follow-up
Due to platform constraints limiting direct comparisons, each QTc group was individually matched with controls using 1:1 propensity score matching (e.g., comparisons I, II, III, and IV) (Fig. 1). Post-matching control groups showed consistent new-onset AF incidence rates (0.77–1.21%). For new-onset AF, a clear U-shaped pattern emerged across QTc categories. The long QTc group demonstrated the highest risk (incidence: 8.59%; HR: 7.384, 95% CI: 5.911–9.224), followed closely by the short QTc group (incidence: 7.02%; HR: 7.735, 95% CI: 4.237–14.122) (Fig. 2). The borderline QTc group showed intermediate risk (incidence: 2.28%; HR: 2.278, 95% CI: 1.796–2.888), while the high-normal QTc group had modestly elevated risk (incidence: 1.11%; HR: 1.442, 95% CI: 1.232–1.688) compared to controls.
Figure 2.
The cumulative incidence (A) and risk (B) of new-onset atrial fibrillation (AF) based on the QTc interval, observed over the 3-month follow-up period. CI = confidence interval, QTc = corrected QT (interval).
Similarly, VA exhibited a U-shaped association with QTc intervals. The long QTc group had the highest VA risk (incidence: 2.98%; HR: 5.277, 95% CI: 3.758–7.411), while the short QTc group also showed significantly elevated risk (incidence: 2.11%; HR: 3.876, 95% CI: 1.688–8.902) (Fig. 3). The borderline QTc group demonstrated moderate risk elevation (incidence: 1.10%; HR: 1.714, 95% CI: 1.255–2.340), whereas the high-normal QTc group showed no significant difference from controls (incidence: 0.68%; HR: 1.149, 95% CI: 0.952–1.387). These findings consistently demonstrate that both QTc prolongation and shortening are associated with substantially increased arrhythmia risk within 3 months, with the most pronounced effects at the extremes of the QTc spectrum.
Figure 3.
The cumulative incidence (A) and risk (B) of ventricular arrhythmia (VA) throughout the 3-month follow-up, categorized according to the QTc interval. CI = confidence interval, QTc = corrected QT (interval).
3.3. Secondary endpoint: impact of QTc intervals on the cumulative incidence and risk of overall mortality at 3-month follow-up
All-cause mortality demonstrated a pronounced U-shaped association with QTc intervals. The long QTc group showed the highest mortality incidence (incidence: 12.39%; HR: 7.405, 95% CI: 6.161–8.901), while the short QTc group, despite lower incidence (6.13%), exhibited the greatest mortality risk (HR: 10.031, 95% CI: 4.850–20.749) (Fig. 4). The borderline QTc group had intermediate mortality (incidence: 5.18%; HR: 3.865, 95% CI: 3.191–4.681), and the high-normal QTc group showed modest elevation (incidence: 2.00%; HR: 2.235, 95% CI: 1.951–2.560) compared to matched controls (incidence: 0.64–1.79%). This U-shaped pattern indicates that both QTc extremes confer substantial mortality risk within 3 months. All abnormal QTc categories demonstrated significantly increased mortality risk compared to controls, emphasizing the prognostic importance of QTc interval assessment.
Figure 4.
The cumulative incidence (A) and risk (B) of overall mortality based on the QTc interval. CI = confidence interval, QTc = corrected QT (interval).
3.4. Sensitivity analysis
To validate our findings, we performed 3 sensitivity analyses. First, after excluding patients exposed to moderate- to high-risk QTc-prolonging medications (Table 2), the associations between QTc intervals and all outcomes remained significant. Notably, the long QTc group showed the strongest associations with all 3 outcomes, while the short QTc group demonstrated significant associations only with new-onset AF. Second, we analyzed data from the pre-COVID-19 period (2015–2019) to eliminate potential pandemic-related confounding effects (Table 3). The associations persisted, with prolonged QTc intervals consistently predicting adverse outcomes. The short QTc group showed significant association with new-onset AF but not with VA or mortality during this period. Third, to minimize heart rate-related confounding, we excluded patients with tachycardia or bradycardia (Table 4). This analysis revealed even stronger associations across all QTc groups. Remarkably, the short QTc group, which previously showed limited associations, demonstrated significant relationships with all 3 outcomes after controlling for heart rate extremes. These sensitivity analyses consistently support our primary findings that abnormal QTc intervals, particularly prolongation, independently predict adverse cardiovascular outcomes and mortality.
Table 2.
Associations between QTc intervals and outcomes at 3-month follow-up after excluding patients who had been exposed to moderate- to high-risk medications known to prolong QTc intervals.
| Outcomes | HR (95% CI) | P-value* |
|---|---|---|
| Long QTc group (500–600 ms) vs control group (370–420 ms) | ||
| New-onset AF | 6.775 (4.363–10.52) | <.001 |
| New-onset VAs | 5.036 (2.831–8.959) | <.001 |
| Mortality | 10.789 (5.441–21.395) | <.001 |
| Borderline QTc group (460–500 ms) vs control group (370–420 ms) | ||
| New-onset AF | 2.720 (1.887–3.922) | <.001 |
| New-onset VAs | 2.508 (1.557–4.041) | <.001 |
| Mortality | 4.321 (2.647–7.053) | <.001 |
| High-normal QTc group (420–460 ms) vs control group (370–420 ms) | ||
| New-onset AF | 1.201 (0.969–1.511) | .0929 |
| New-onset VAs | 1.301 (1.035–1.634) | .0235 |
| Mortality | 2.117 (1.658–2.963) | <.001 |
| Short QTc group (200–370 ms) vs control group (370–420 ms) | ||
| New-onset AF | 2.183 (1.672–7.090) | <.001 |
| New-onset VAs | 2.125 (0.738–6.116) | .1524 |
| Mortality | 2.183 (0.672–7.09) | .1826 |
AF = atrial fibrillation, CI = confidence interval, HR = hazard ratio, VAs = ventricular arrhythmias, QTc = corrected QT (interval).
Log-rank test.
Table 3.
Associations between QTc intervals and outcomes at 3-month follow-up before COVID-19 pandemic (2015–2019).
| Outcomes | HR (95% CI) | P-value* |
|---|---|---|
| Long QTc group (500–600 ms) vs control group (370–420 ms) | ||
| New-onset AF | 2.667 (2.198–3.236) | <.001 |
| New-onset VAs | 1.979 (1.405–2.787) | <.001 |
| Mortality | 1.908 (1.646–2.213) | <.001 |
| Borderline QTc group (460–500 ms) vs control group (370–420 ms) | ||
| New-onset AF | 1.921 (1.637–2.253) | <.001 |
| New-onset VAs | 1.460 (1.112–1.917) | .006 |
| Mortality | 2.300 (2.053–2.578) | <.001 |
| High-normal QTc group (420–460 ms) vs control group (370–420 ms) | ||
| New-onset AF | 1.265 (1.102–1.453) | <.001 |
| VAs | 1.003 (0.814–1.236) | .977 |
| Mortality | 1.580 (1.428–1.749) | <.001 |
| Short QTc group (200–370 ms) vs control group (370–420 ms) | ||
| New-onset AF | 2.515 (1.476–4.285) | <.001 |
| New-onset VAs | 1.327 (0.297–5.930) | .710 |
| Mortality | 1.337 (0.787–2.269) | .281 |
AF = atrial fibrillation, CI = confidence interval, HR = hazard ratio, VAs = ventricular arrythmias, QTc = corrected QT (interval).
Log-rank test.
Table 4.
Associations between QTc intervals and outcomes at 3-month follow-up after excluding patients with tachycardia (heart rate > 100/min) or bradycardia (heart rate < 60/min).
| Outcomes | HR (95% CI) | P-value* |
|---|---|---|
| Long QTc group (500–600 ms) vs control group (370–420 ms) | ||
| New-onset AF | 6.910 (5.598–8.529) | <.001 |
| New-onset VAs | 6.235 (4.336–8.966) | <.001 |
| Mortality | 9.790 (7.837–12.230) | <.001 |
| Borderline QTc group (460–500 ms) vs control group (370–420 ms) | ||
| New-onset AF | 2.438 (1.968–3.022) | <.001 |
| New-onset VAs | 2.837 (2.031–3.962) | <.001 |
| Mortality | 4.690 (3.759–5.851) | <.001 |
| High-normal QTc group (420–460 ms) vs control group (370–420 ms) | ||
| New-onset AF | 1.280 (1.118–1.466) | <.001 |
| New-onset VAs | 1.563 (1.290–1.893) | <.001 |
| Mortality | 2.156 (1.847–2.517) | <.001 |
| Short QTc group (200–370 ms) vs control group (370–420 ms) | ||
| New-onset AF | 6.132 (4.216–8.918) | <.001 |
| New-onset VAs | 3.643 (1.808–7.342) | <.001 |
| Mortality | 7.710 (5.16–11.52) | <.001 |
AF = atrial fibrillation, CI = confidence interval, HR = hazard ratio, VAs = ventricular arrythmias, QTc = corrected QT (interval).
Log-rank test.
3.5. Subgroup analyses
Gender-specific analyses confirmed a consistent U-shaped association between QTc intervals and adverse outcomes at 3-month follow-up for both females and males (Tables 5 and 6). In females, very long QTc intervals (>510 ms) conferred the highest risks for VA, AF, and mortality compared to controls (360–460 ms), while moderately prolonged (460–510 ms) and short (<360 ms) QTc intervals were also associated with significantly increased risks. Males showed a similar pattern, with very long QTc intervals (>500 ms) linked to the highest adverse event rates, particularly for VA, and both intermediate long (450–500 ms) and short (<350 ms) QTc intervals also associated with elevated risk. These findings reinforce the U-shaped relationship across both sexes, indicating that both QTc prolongation and shortening are important risk markers, with the greatest risk observed in patients with very long QTc intervals.
Table 5.
Subgroup analysis of associations between QTc intervals and outcomes in females at 3-month follow-up.
| Outcomes | HR (95% CI) | P-value* |
|---|---|---|
| Very long QTc group (>510 ms) vs control group (360–460 ms) | ||
| New-onset AF | 6.579 (5.102–8.485) | <.001 |
| New-onset VAs | 7.726 (4.721–12.643) | <.001 |
| Mortality | 4.769 (3.919–5.803) | <.001 |
| Long QTc group (460–510 ms) vs control group (360–460 ms) | ||
| New-onset AF | 2.145 (1.740–2.643) | <.001 |
| New-onset VAs | 3.031 (2.098–4.378) | <.001 |
| Mortality | 2.564 (2.143–3.067) | <.001 |
| Short QTc group (<360 ms) vs control group (360–460 ms) | ||
| New-onset AF | 2.099 (1.753–2.514) | <.001 |
| New-onset VAs | 1.950 (1.452–2.620) | <.001 |
| Mortality | 2.031 (1.725–2.391) | <.001 |
AF = atrial fibrillation, CI = confidence interval, HR = hazard ratio, VA = ventricular arrythmias, QTc = corrected QT (interval).
Log-rank test.
Table 6.
Subgroup analysis of associations between QTc intervals and outcomes in males at 3-month follow-up.
| Outcomes | HR (95% CI) | P-value* |
|---|---|---|
| Very long QTc group (>500 ms) vs control group (350–450 ms) | ||
| New-onset AF | 7.458 (6.208–8.959) | <.001 |
| New-onset VAs | 9.085 (6.051–13.641) | <.001 |
| Mortality | 6.919 (5.844–8.191) | <.001 |
| Long QTc group (450–500 ms) vs control group (350–450 ms) | ||
| New-onset AF | 2.701 (2.301–3.171) | <.001 |
| New-onset VAs | 3.034 (2.289–4.021) | <.001 |
| Mortality | 3.978 (3.396–4.660) | <.001 |
| Short QTc group (<350 ms) vs control group (350–450 ms) | ||
| New-onset AF | 1.645 (1.417–1.908) | <.001 |
| New-onset VAs | 2.536 (1.968–3.269) | <.001 |
| Mortality | 2.564 (2.212–2.973) | <.001 |
AF = atrial fibrillation, CI = confidence interval, HR = hazard ratio, VAs = ventricular arrythmias, QTc = corrected QT (interval).
Log-rank test.
4. Discussion
Before matching, our results showed a distinct U-shaped association between QTc intervals and the incidence of various comorbidities, including DM, ischemic heart diseases, kidney diseases, and heart failure. After matching with age, sex, race, and comorbidities, analysis on 145,111 patients further showed a U-shaped pattern in the risk of new-onset AF, VA, and mortality across different QTc interval categories in a 3-month follow-up. Sensitivity analysis confirmed the U-shaped association between QTc intervals and AF risk. However, a short QTc interval was not related to an increased VA/mortality risk following sensitivity analysis, indicating that the U-shaped association was not evident. These findings suggest that the elevated risks observed in the primary analysis may be influenced by other factors, highlighting the importance of interpreting QTc-associated risks within relevant subgroups.
AF is the most prevalent cardiac arrhythmia, with its frequency steadily increasing over time.[19–21] In our study, we noted that individuals with prolonged QTc intervals were more likely to experience new-onset AF, which aligns with the results of a previous study.[22,23] This alignment is supported by a previous meta-analysis involving 309,676 participants, which showed that those with prolonged QTc intervals had a significantly higher vulnerability to AF, as indicated by an HR of 1.16.[24] Unlike previous studies that mainly focused on the long-term effects of prolonged QTc intervals,[22,23] our study specifically investigated the short-term impact of long QT intervals on new-onset AF. Substantial and compelling evidence underscores that individuals with AF encounter an increased susceptibility to stroke, congestive heart failure, and dementia.[25–28] These findings emphasize the noteworthy public health implications linked to AF. Our findings highlighted the significance for healthcare professionals to promptly monitor and implement effective strategies for improving patient outcomes when dealing with prolonged QTc intervals. Notably, our findings provided new insights by revealing a link between short QTc intervals and an increased short-term AF risk. Moreover, the noticeable U-shaped association reminds healthcare professionals to be vigilant about specific subgroups in the population, which could be inadvertently overlooked owing to the relatively low occurrence of short QTc intervals.
Our results suggested a U-shaped association of QTc intervals with VA. Furthermore, we quantified the VA risk in direct relation to specific QTc interval lengths. Current evidence suggests that QTc prolongation is associated with an increased risk of malignant ventricular dysrhythmias.[29] Despite the established association, the exact estimated VA risk corresponding to specific QTc interval lengths remains unknown.[29,30] For example, while a 650-ms QTc interval is presumably more indicative of malignant ventricular dysrhythmias than a 420-ms QTc interval, a precise quantitative risk value for VA is lacking.[29] This deficiency in quantification impedes clinical decision-making, thereby leaving clinicians to only express an increased likelihood of elevated risk for malignant ventricular dysrhythmias beyond the baseline.[29] Despite a recent attempt through a meta-analysis to address this matter,[29] the authors have highlighted the inadequacy of the existing data in the current literature to define a precise range of QTc interval values that can anticipate the vulnerability to ventricular dysrhythmias in a quantifiable manner. Our findings have now contributed to addressing this knowledge gap by offering updated insights.
In the current study, the association of all-cause mortality with QTc intervals also showed a distinctive U-shaped pattern. An interesting finding was that patients in the short QTc group (HR, 10.031) had a relatively higher mortality risk than those in the long QTc group (HR, 7.405). Our findings aligned with those of a previous study, which reported that both shortened and prolonged QT interval durations were linked to increased long-term mortality risk within the general population.[31] Previous studies demonstrated that QTc prolongation was associated with mortality risk in a long-term follow-up.[32–34] However, this relationship is inconsistent in the current literature. For example, involving 10,804 participants over an average follow-up of 141.9 ± 28.3 months, a previous study showed a progressive increase in cardiovascular and stroke mortality with QTc prolongation.[34] However, a meta-analysis of 7 prospective cohort studies involving 36,031 participants yielded inconclusive evidence regarding heightened risks of total or cardiovascular mortality, as well as sudden death, associated with QTc interval prolongation.[35] In contrast to that meta-analysis,[35] another meta-analysis that included 23 observational studies reported associations between prolonged QT intervals and heightened risks of total, cardiovascular, coronary, and sudden cardiac death.[36] The presence of inconsistent findings may be attributed to the U-shaped association between QT intervals and mortality risk, potentially introducing a bias.
We investigated additional variables that could potentially impact the outcomes of our study. Recognizing the dual etiology of long QT syndrome (congenital and acquired) we made efforts to exclude patients who had been exposed to pharmacological agents with known potential to extend QTc intervals. Our sensitivity analysis demonstrated the enduring consistency of the U-shaped correlation between QTc interval and AF risk, indicating the robustness of the evidence. Nonetheless, of note, the U-shaped association between QTc intervals and occurrences of VA or mortality might not be as discernible upon the exclusion of patients who were exposed to QTc interval-prolonging medications. Intriguingly, this association had also disappeared before the COVID-19 pandemic. The underlying reasons for these observations remain enigmatic, necessitating further investigation. However, these results emphasize the significance of vigilant monitoring for VA and mortality risk in cases of shortened QTc intervals, particularly among individuals exposed to QTc-prolonging medications or during the COVID-19 pandemic.
Our study population was limited to individuals with at least 2 ECGs and prespecified clinic/hospital visits, which may introduce selection bias. This criterion likely favors patients with frequent healthcare interactions, potentially excluding those who undergo ECGs infrequently or in acute settings. Consequently, our findings may not be fully generalizable to individuals receiving a single, incidental ECG outside routine medical care. Furthermore, our cohort may include patients with underlying conditions requiring repeated ECG monitoring, making them distinct from the general population. However, our study is highly relevant to populations requiring ongoing cardiac monitoring, where repeated ECGs are standard practice. Future research should explore whether these results apply to broader populations, including those receiving only a single or first-time ECG, to improve external validity.
Our study examined QTc interval–outcome associations using categories that differ from conventional clinical thresholds. This approach allowed systematic examination of risk patterns rather than validating specific cutoffs. The U-shaped relationship between QTc duration and outcomes remained consistent across multiple sensitivity analyses, suggesting this finding is robust regardless of category boundaries. While our nomenclature differs from standard definitions, the elevated risks at both QTc extremes align with cardiac repolarization physiology and provide valuable insights for risk stratification.
Compared to the study by Zhang et al,[31] which identified a U-shaped relationship between QTc intervals and long-term mortality, our study offers several noteworthy contributions. First, we offer updated information by incorporating data during the COVID-19 pandemic period. Second, long-term outcomes are often influenced by several factors that can change over extended periods. In the short-term scenario of a 3-month follow-up, the influence of confounding variables and external factors may be relatively minimized, thereby allowing for a clearer interpretation of the results. Third, we have demonstrated that the U-shaped associations between QTc intervals and VA or mortality risk can become evident in patients exposed to QTc interval-prolonging medications and during the COVID-19 pandemic. This crucial information has the potential to serve as supplementary guidance for clinicians in their decision-making process when managing patients with short QTc intervals.
Despite the recognized challenges of QTc measurement (including correction biases and influences from medications, chronic conditions, and acute illnesses) our large-scale cohort with propensity score matching provides robust evidence supporting its clinical relevance. While QTc-prolonging medications and COVID-19 pandemic influenced associations with VA and mortality, these findings underscore the importance of considering external factors rather than diminishing QTc’s predictive value. Given its widespread availability, QTc remains a practical marker for identifying high-risk patients who may benefit from closer monitoring. Though QTc should not be viewed in isolation, our results reinforce its utility within comprehensive risk assessment frameworks.
While our study demonstrates a U-shaped association between QTc intervals and outcomes after propensity matching, underlying pathophysiological conditions likely drive both repolarization abnormalities and clinical endpoints, rather than the QTc interval itself being causative. QTc interval should be interpreted as a marker that integrates multiple physiological disruptions rather than an independent therapeutic target. Clinicians should view QTc measurements as 1 component within comprehensive risk assessment frameworks that prioritize underlying conditions, comorbidities, and clinical context when making monitoring and treatment decisions. The prognostic value of QTc likely reflects its role as an accessible biomarker of cardiac vulnerability rather than suggesting that intervention should target the interval measurement itself.
The current study had several limitations. First, the predominance of White patients in the sample may limit the generalizability of the findings. Second, the retrospective design introduced potential biases associated with uneven parameter registrations, unrecorded confounding variables, and missing data. However, of note, the observed trend in QTc intervals and outcomes were consistent with the results of a previous study, indicating potential minimization of bias risks in the study’s design. Third, the dynamic nature of QTc intervals introduced the possibility of misclassification bias when assigning patients to each group in this study. Although rigorous criteria were employed for the borderline QTc, high-normal QTc, and control groups to mitigate misclassification bias, a degree of risk persisted, particularly in the long and short QTc groups. Conversely, the non-strict criteria applied to these 2 high-risk groups facilitated informed decision-making for clinicians without necessitating prolonged ECG follow-up. Fourth, the TriNetX platform does not allow us to distinguish which correction formula (Bazett or Fridericia) was used for individual ECGs in the database, preventing direct comparison between formulas. To address this limitation, we conducted a sensitivity analysis excluding patients with abnormal heart rates (tachycardia > 100 bpm or bradycardia < 60 bpm), where correction formulas are most prone to systematic bias. Fifth, while our sensitivity analysis excluded patients taking QT-prolonging drugs, we could not account for all medications potentially affecting QTc intervals, such as antihypertensives altering heart rate or diuretics affecting electrolyte balance, which represents an important limitation of our findings.
In conclusion, our findings unveiled a U-shaped correlation between QTc intervals and the occurrence of AF, VA, and mortality within a 3-month observation period, underscoring the viability of QTc intervals as a short-term prognostic predictor. To explore the potential enhancement of patient outcomes through the integration of QTc intervals into clinical practice, future studies are warranted.
Author contributions
Conceptualization: Ti-Chuan Chiu, Chun-Ning Ho, James Cheng-Chung Wei, Yu-Yu Li.
Data curation: Ti-Chuan Chiu, Chun-Ning Ho, James Cheng-Chung Wei, Kuo-Chuan Hung, Yu-Yu Li.
Formal analysis: Ti-Chuan Chiu, Chun-Ning Ho, Kuo-Chuan Hung, Yu-Yu Li.
Funding acquisition: Jing-Yang Huang, Ping-Heng Tan.
Investigation: James Cheng-Chung Wei, Jing-Yang Huang.
Methodology: James Cheng-Chung Wei, Kuo-Chuan Hung, Ping-Heng Tan.
Resources: Kuo-Chuan Hung.
Software: Chun-Ning Ho, Ping-Heng Tan.
Validation: Jing-Yang Huang.
Visualization: Jing-Yang Huang.
Writing – original draft: Ti-Chuan Chiu, Chun-Ning Ho, James Cheng-Chung Wei, Kuo-Chuan Hung, Yu-Yu Li.
Writing – review & editing: Ti-Chuan Chiu, James Cheng-Chung Wei, Kuo-Chuan Hung, Yu-Yu Li.
Supplementary Material
Abbreviations:
- AF
- atrial fibrillation
- CI
- confidence interval
- DM
- diabetes mellitus
- ECG
- electrocardiogram
- HR
- hazard ratio
- ICD-10
- International Classification of Diseases, 10th Revision
- ms
- milliseconds
- QTc
- corrected QT (interval)
- VA
- ventricular arrhythmias
This study was approved by the Institutional Review Board of Chi Mei Medical Center (IRB Serial No. 11206-E01).
The authors have no funding and conflicts of interest to disclose.
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.
Supplemental Digital Content is available for this article.
How to cite this article: Chiu T-C, Ho C-N, Wei JC-C, Hung K-C, Huang J-Y, Tan P-H, Li Y-Y. Impact of prolonged and short QT intervals on immediate risks of newly diagnosed arrhythmias and mortality: A retrospective study. Medicine 2025;104:37(e44449).
C-NH, T-CC, and JC-CW contributed equally to this work.
Contributor Information
Ti-Chuan Chiu, Email: cdjcircling@gmail.com.
Chun-Ning Ho, Email: chunning.ho@gmail.com.
James Cheng-Chung Wei, Email: jccwei@gmail.com.
Kuo-Chuan Hung, Email: ed102605@gmail.com.
Jing-Yang Huang, Email: hjc1395@gmail.com.
Ping-Heng Tan, Email: tanphphd@yahoo.com.tw.
References
- [1].Ponce-Balbuena D, Deschênes I. Long QT syndrome – bench to bedside. Heart Rhythm O2. 2021;2:89–106. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [2].Wilde AAM, Amin AS, Postema PG. Diagnosis, management and therapeutic strategies for congenital long QT syndrome. Heart. 2022;108:332–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [3].Galić E, Bešlić P, Kilić P, et al. Congenital long QT syndrome: a systematic review. Acta Clin Croat. 2021;60:739–48. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [4].Tan MS, Heise CW, Gallo T, et al. Relationship between a risk score for QT interval prolongation and mortality across rural and urban inpatient facilities. J Electrocardiol. 2023;77:4–9. [DOI] [PubMed] [Google Scholar]
- [5].Weeke PE, Kellemann JS, Jespersen CB, et al. Long-term proarrhythmic pharmacotherapy among patients with congenital long QT syndrome and risk of arrhythmia and mortality. Eur Heart J. 2019;40:3110–7. [DOI] [PubMed] [Google Scholar]
- [6].Golzari H, Dawson NV, Speroff T, Thomas C. Prolonged QTc intervals on admission electrocardiograms: prevalence and correspondence with admission electrolyte abnormalities. Conn Med. 2007;71:389–97. [PubMed] [Google Scholar]
- [7].Pham JC, Banks MC, Narotsky DL, Dorman T, Winters BD. The prevalence of long QT interval in post-operative intensive care unit patients. J Clin Monit Comput. 2016;30:437–43. [DOI] [PubMed] [Google Scholar]
- [8].Welten S, Elders PJM, Remmelzwaal S, et al. Prolongation of the heart rate-corrected QT interval is associated with cardiovascular diseases: Systematic review and meta-analysis. Arch Cardiovasc Dis. 2023;116:69–78. [DOI] [PubMed] [Google Scholar]
- [9].Patel SI, Ackerman MJ, Shamoun FE, et al. QT prolongation and sudden cardiac death risk in hypertrophic cardiomyopathy. Acta Cardiol. 2019;74:53–8. [DOI] [PubMed] [Google Scholar]
- [10].Ahmadizar F, Soroush N, Ikram MA, Kors JA, Kavousi M, Stricker BH. QTc-interval prolongation and increased risk of sudden cardiac death associated with hydroxychloroquine. Eur J Prev Cardiol. 2022;28:1875–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [11].Seftchick MW, Adler PH, Hsieh M, et al. The prevalence and factors associated with QTc prolongation among emergency department patients. Ann Emerg Med. 2009;54:763–8. [DOI] [PubMed] [Google Scholar]
- [12].Gibbs C, Thalamus J, Kristoffersen DT, et al. QT prolongation predicts short-term mortality independent of comorbidity. Europace. 2019;21:1254–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [13].Chen IW, Chang LC, Wu JY, et al. Association between preoperative COVID-19 infection and postoperative outcomes in patients with obstructive sleep apnea undergoing metabolic surgery: a retrospective analysis. Obes Surg. 2025;35:2218–26. [DOI] [PubMed] [Google Scholar]
- [14].Hung KC, Yu TS, Hung IY, Wu JY, Yew M, Chen IW. Impact of vitamin D deficiency on postoperative outcomes in patients with chronic kidney disease undergoing surgery: a retrospective study. Sci Rep. 2025;15:9757. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [15].Hung KC, Chang LC, Chang YJ, et al. Vitamin D deficiency and diabetic retinopathy risk in patients with newly diagnosed type 2 diabetes mellitus: a retrospective analysis. Front Nutr. 2025;12:1614287. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [16].Hung KC, Weng HL, Lai YC, et al. Association between preoperative anemia and postoperative acute kidney injury in patients undergoing metabolic and bariatric surgery: a multi-institute study. Obes Surg. 2025;35:1827–37. [DOI] [PubMed] [Google Scholar]
- [17].Dewi IP, Dharmadjati BB. Short QT syndrome: the current evidences of diagnosis and management. J Arrhythm. 2020;36:962–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [18].Andršová I, Hnatkova K, Šišáková M, et al. Influence of heart rate correction formulas on QTc interval stability. Sci Rep. 2021;11:14269. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [19].Brundel B, Ai X, Hills MT, Kuipers MF, Lip GYH, de Groot NMS. Atrial fibrillation. Nat Rev Dis Primers. 2022;8:21. [DOI] [PubMed] [Google Scholar]
- [20].Lippi G, Sanchis-Gomar F, Cervellin G. Global epidemiology of atrial fibrillation: An increasing epidemic and public health challenge. Int J Stroke. 2021;16:217–21. [DOI] [PubMed] [Google Scholar]
- [21].Westerman S, Wenger N. Gender differences in atrial fibrillation: a review of epidemiology, management, and outcomes. Curr Cardiol Rev. 2019;15:136–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [22].Mandyam MC, Soliman EZ, Alonso A, et al. The QT interval and risk of incident atrial fibrillation. Heart Rhythm. 2013;10:1562–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [23].Patel N, O’Neal WT, Whalen SP, Soliman EZ. The association of QT interval components with atrial fibrillation. Ann Noninvasive Electrocardiol. 2018;23:e12467. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [24].Zhang N, Gong M, Tse G, et al. Prolonged corrected QT interval in predicting atrial fibrillation: a systematic review and meta-analysis. Pacing Clin Electrophysiol. 2018;41:321–7. [DOI] [PubMed] [Google Scholar]
- [25].Bansal N, Zelnick LR, An J, et al. Incident atrial fibrillation and risk of dementia in a diverse, community-based population. J Am Heart Assoc. 2023;12:e028290. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [26].Kotalczyk A, Mazurek M, Kalarus Z, Potpara TS, Lip GY. Stroke prevention strategies in high-risk patients with atrial fibrillation. Nat Rev Cardiol. 2021;18:276–90. [DOI] [PubMed] [Google Scholar]
- [27].Karnik AA, Gopal DM, Ko D, Benjamin EJ, Helm RH. Epidemiology of atrial fibrillation and heart failure: a growing and important problem. Cardiol Clin. 2019;37:119–29. [DOI] [PubMed] [Google Scholar]
- [28].Rivard L, Friberg L, Conen D, et al. Atrial fibrillation and dementia: a report from the AF-SCREEN International Collaboration. Circulation. 2022;145:392–409. [DOI] [PubMed] [Google Scholar]
- [29].Robison LB, Brady WJ, Robison RA, Charlton N. QT interval prolongation and the risk of malignant ventricular dysrhythmia and/or cardiac arrest: systematic search and narrative review of risk related to the magnitude of QT interval length. Am J Emerg Med. 2021;49:40–7. [DOI] [PubMed] [Google Scholar]
- [30].Shah SR, Park K, Alweis R. Long QT syndrome: a comprehensive review of the literature and current evidence. Curr Probl Cardiol. 2019;44:92–106. [DOI] [PubMed] [Google Scholar]
- [31].Zhang Y, Post WS, Dalal D, Blasco-Colmenares E, Tomaselli GF, Guallar E. QT-interval duration and mortality rate: results from the Third National Health and Nutrition Examination Survey. Arch Intern Med. 2011;171:1727–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [32].Yap J, Jin AZ, Nyunt SZ, Ng TP, Richards AM, Lam CS. Longitudinal community-based study of QT interval and mortality in Southeast Asians. PLoS One. 2016;11:e0154901. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [33].Fagher K, Nilsson A, Löndahl M. Heart rate‐corrected QT interval prolongation as a prognostic marker for 3‐year survival in people with Type 2 diabetes undergoing above‐ankle amputation. Diabet Med. 2015;32:679–85. [DOI] [PubMed] [Google Scholar]
- [34].Ishikawa J, Ishikawa S, Kario K. Relationships between the QTc interval and cardiovascular, stroke, or sudden cardiac mortality in the general Japanese population. J Cardiol. 2015;65:237–42. [DOI] [PubMed] [Google Scholar]
- [35].Montanez A, Ruskin JN, Hebert PR, Lamas GA, Hennekens CH. Prolonged QTc interval and risks of total and cardiovascular mortality and sudden death in the general population: a review and qualitative overview of the prospective cohort studies. Arch Intern Med. 2004;164:943–8. [DOI] [PubMed] [Google Scholar]
- [36].Zhang Y, Post WS, Blasco-Colmenares E, Dalal D, Tomaselli GF, Guallar E. Electrocardiographic QT interval and mortality: a meta-analysis. Epidemiology. 2011;22:660–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.