Surveillance of Corrected QT Interval-Prolonging Medications upon Admission throughout Hospitalization in a Tertiary Care Geriatric Ward

Long QT syndrome (LQTS) is a multifactorial ECG-measurable disorder of corrected QT (QTc) interval that can lead to the development of torsade de pointes (TdP), a form of malignant cardiac arrhythmia [1]. Not every prolongation of the QTc interval leads to the occurrence of TdP, but a QTc interval over 500 ms increases the risk of this potentially fatal heart rhythm disorder by 2–3 times [2]. On the other hand, in every patient with TdP, a prolongation of the QTc interval was found, which is why it represents a diagnostic criterion for TdP [3].

Older patients may be more susceptible to developing a prolonged QTc interval due to prolonged cardiac repolarization and reduced RR variance observed with advanced age [4]. Additionally, comorbidities and polypharmacy further increase the risk of QTc prolongation in this population [5, 6]. Moreover, acquired QTc interval prolongation is most commonly caused by the use of specific medications, such as antiarrhythmics, antipsychotics, antibiotics, antidepressants, or diuretics affecting electrolytes, which are frequently prescribed in hospitalized older adults [3, 7, 8]. There is evidence that the administration of more than one QTc interval-prolonging medication may further increase the risk of developing QTc interval prolongation and TdP [9].

Due to the indispensable clinical consequences of QTc prolongation, there has been a strong interest in developing models for the prediction of QTc prolongation based on patients’ modifiable and non-modifiable risk factors [10, 11]. However, most studies have been conducted in adults less than 65 years of age; therefore, due to combined risk factors the impact of QTc interval-prolonging medications may differ in the population aged 65 or above [10‒12]. Although a high prevalence of prescription of QTc interval-prolonging medications is expected, both at hospital admission and discharge, potential TdP induced by QT-prolonging medications might be a neglected issue during the routine patient assessments. The primary objectives of this cross-sectional study were to estimate the prevalence and characteristics of LQTS and the use of QTc interval-prolonging medications at admission and during hospitalization in older patients admitted to the geriatric unit. In addition, the study also aimed to identify clinical and medication-associated risk factors of LQTS and to evaluate the burden of co-prescribed medications with a potential risk of QTc interval prolongation in this specific population.

This prospective, observational study was conducted in the Clinical Department of Geriatrics, Zvezdara University Medical Center, Serbia.

The study included older patients (≥65 years) who were consecutively admitted to the Clinical Department of Geriatrics from April to October 2023. Written informed consent was obtained from all patients prior to their participation in the study. Patients with acute myocardial infarction or angina pectoris, bigeminus, patients with pacemakers, bundle branch block on ECG, wide QRS complex (>120 ms), or unidentifiable T waves were excluded. Patients with extreme atrial fibrillation (slow, <50 ms or rapid, >120 ms) were also excluded from the study. All patients aged ≥65 years who were prescribed at least one QTc interval-prolonging medicine on admission were enrolled in the study. The characteristics associated with LQTS were also recorded, these included patient age, sex, serum electrolytes (potassium, calcium, magnesium), serum creatinine concentrations, and medical conditions, including hypertension, hypothyroidism, type 2 diabetes mellitus, heart failure, and atrial fibrillation. Hypokalemia was defined as a serum potassium concentration <3.4 mmol/L, hypocalcemia as a calcium concentration <2.2 mmol/L, hypomagnesemia as a magnesium concentration <0.73 mmol/L, and elevated serum creatinine as a creatinine concentration above 106 μmol/L. The biochemical parameters were recorded on the same day as the ECG. After admission, a clinical pharmacist collected detailed information from patients or family members/caregivers about pharmacological treatment 30 days prior to hospitalization. Details of prescriptions during hospitalization were obtained from medical records.

Within 24 h of the patient’s admission, a specialist in internal medicine conducted a 12-lead ECG using an electrocardiograph equipped with automatic QT interval measurement. LQTS was defined as a QTc interval of >450 ms in men and >470 ms in women using the Bazett formula (QTc = QT/√RR). Patients were followed for 7 days and 12-lead ECGs were performed after 7 days of hospitalization. Biochemical parameters were all obtained on the same day of the ECG. The list of QTc interval-prolonging medications was obtained from CredibleMeds^® (medications with a known risk of TdP – list 1, possible risk – list 2, and conditional risk – list 3).

Continuous variables were described as median and interquartile range (IQR). Categorical variables were expressed as numbers and percentages. Patients were divided into two subgroups according to the presence of LQTS, and a case-control analysis was performed. The Mann-Whitney, two-sided χ², Fisher’s exact (when n < 5), Wilcoxon sign-rank, McNemar’s, Kruskal-Wallis tests, and binary logistic regression analysis were used. All calculations and statistical analyses were performed using the SPSS 20.0 software package. Statistical significance was set at p < 0.05.

The study included 128 patients, predominantly female (58.6%) and with a median age of 81 years [IQR 75–85]. QTc interval-prolonging medications, 35 (27.3%) patients in total had LQTS, while 19 (14.8%) patients had both QTc interval-prolonging medications and LQTS. Patients with LQTS on admission had more frequent QTc interval-prolonging medications in their therapy, list 1 medications, and atrial fibrillation. The difference in proportions was statistically significant (p < 0.05). Altered electrolyte concentrations were not significantly more prevalent in patients with LQTS. Detailed demographic, clinical, and therapy characteristics on admission are presented in Table 1.

After admission, patients were provided with further diagnostic and therapeutic services on the ward. Additionally, for the purpose of this research, a new examination of the QTc interval was performed together with laboratory parameters after 7 days of hospitalization. The prevalence of LQTS significantly increased during the hospital stay, from 35 (27.3%) at admission to 54 (42.2%) after 7 days of hospitalization (p = 0.001). In the total study sample, the QTc interval increased during hospitalization from 440 ms [413–465 ms] to 460 ms [436–479 ms] on average. The difference was found to be statistically significant (p < 0.001). A significant increase in QTc interval was observed in 91 patients (71.1%) and a significant decrease in 34 patients (26.6%) (both p < 0.001), while the same value was measured in 3 patients at two time points (2.3%). The highest absolute increase in QTc interval value was observed among male patients (Table 2). When applying criteria of absolute change in QTc interval of 20 ms or to a QTc interval >500 ms [13], a meaningful increase was observed in 59 (46.1%) patients. The number of administered medications with the potential to cause LQTS also increased during hospital stay (p < 0.001) from a median (IQR) of 0 (0–1) at baseline to a median of 3 (2–3) after 7 days. At hospital admission, about one-third of patients had at least one medication with QTc interval-prolonging properties (38.3%). In contrast, after 7 days, all patients received at least one QTc interval-prolonging medication (100%). The results are presented in Table 2.

A significant increase in list 1 users was observed during hospitalization (from 18% to 68%, Table 2). Since the results confirmed a statistically significant increase in the QTc interval in users of list 1 medications, we further compared the QTc interval with respect to the number of list 1 medications administered during hospital stay. Statistical differences between groups of patients defined by number of list 1 medicines were assessed by the Kruskal-Wallis test, and post hoc pairwise comparisons were performed to identify specific group differences (Fig. 1). Although the difference between patients treated with one and two list 1 medications was not statistically significant, a trend toward increasing QTc interval prolongation was observed (Fig. 1).

In second part of the study, we investigated the potential predictors of LQTS during hospitalization (Table 3). All variables that showed statistically significant associations with LQTS in univariate analysis (presented crude odds ratio) were then included in the multivariate analysis to control for the other variables. Male sex and the use of amiodarone confirmed their statistical significance in multivariate analysis. Male sex increases the odds of having LQTS by 3.45 times, whereas the odds of LQTS are 4 times higher in patients receiving amiodarone. Other list 1 medications or the number of list 1 medications did not show a statistically significant association with LQTS in this sample size. A more detailed analysis of the patients with newly developed LQTS during hospitalization showed that hypomagnesemia was statistically more common (66.7% of patients; p = 0.018), even when no differences were observed in other patients’ characteristics.

Figure 2 shows the QTc interval in different patient groups in relation to their clinical or laboratory parameters based on Mann-Whitney test. The QTc interval was statistically longer in patients with atrial fibrillation, heart failure, who were taking amiodarone or furosemide, or who had hypokalemia or hypocalcemia.

In a sample of patients aged 65 or above admitted to the geriatric ward, we found the following main findings: (1) a high prevalence of prolonged QTc interval both at admission and during hospitalization; (2) a significant association of prolonged QTc during hospitalization with male sex, atrial fibrillation, amiodarone, and list 1 medication (with a known risk for TdP) use; (3) potential predictors of LQTS during hospitalization included male sex, amiodarone therapy; (4) hypomagnesemia was more common in patients with newly developed LQTS during hospitalization; (5) clinicians are not fully aware of risk factors for QTc prolongation in older patients and patients are not routinely monitored.

The findings of the present study indicate that hospitalized older patients are at significant risk for QTc prolongation. We observed a significant proportion of LQTS (27.3%) and 8.6% of QTc interval >500 ms in a cohort of older patients at the time of admission to the geriatric ward. Our prevalence rates were similar to those described in two previous retrospective studies of geriatric inpatients, reporting 22–27% LQTS on geriatric wards [14, 15]. Furthermore, our study revealed a significant increase in LQTS prevalence during hospitalization, from 27.3% up to 42.2% after 7 days of hospital stay. In our cohort of older patients, LQTS during hospital stay as well as absolute increase in QTc interval were significantly associated with the use of amiodarone and list 1 medications. Older patients often use medications that can prolong QTc interval, classified in the CredibleMeds^® lists [1, 9, 16]. The results of present study suggest that the acquired form of LQTS is mostly medication-related because patients with LQTS on admission had more frequent list 1 medications (with known risk for TdP) in their therapy (Table 1). In addition to the high use of QTc interval-prolonging medications during hospitalization, the tendency to add these medications in different hospital settings has also been reported [9, 17]. Our results in Table 3 indicate no relationship between the number of list 1 medications and the occurrence of QTc interval prolongation, which differs from the work of Tisdale et al. [17]. The observed differences can be attributed to variations in patient populations and methodologies. Tisdale et al. [17] included patients admitted to cardiac units with a high prevalence of acute heart failure and myocardial infarction, whereas we excluded patients with acute myocardial infarction.

Co-prescriptions of two or more QTc interval-prolonging medications are also common in our population aged 65 or above ≥2. The QT-prolonging medications were prescribed to 87.5% of patients. The use of QTc interval-prolonging medication combinations increased the risk of LQRS compared to a single medication. In univariate logistic analysis, the number of list 1 medication was a statistically significant predictor of LQTS: one medication increased the odds of LQTS by 2.3-fold and patients with 2 list 1 medications had an almost 5-fold higher odds of LQTS compared with patients with no list 1 medications during hospitalization. However, when adjusting for other variables (sex, amiodarone therapy, atrial fibrillation), the effect was not statistically significant for this sample size. On the other hand, it could be assumed that patients who were prescribed two or more QTc interval-prolonging medications were previously assessed by a physician for the QTc interval prolongation risk stratification and the benefit outweighed the risk of concomitant use of QTc interval-prolonging medications [18]. Our results could be partly explained by a more vigilant assessment of the hospitalized patient cohort due to the prospective interventional design. In line with the observed associations, we also compared the QTc interval with the number of list 1 medications administered during hospitalization. While there was no statistically significant difference between the patient groups treated with one and two list 1 medications, an upward trend in QTc interval prolongation was noted (Fig. 1). Moreover, clinical studies have highlighted an enhanced TdP risk in patients prescribed multiple QTc interval-prolonging medications [19]. These findings show that physicians, even in a hospital setting, are not fully aware of adverse events such as medication-related QTc prolongation and potentially life-threatening TdP in routine practice. A recent report in the literature highlighted the risk of hypomagnesemia with the use of proton pump inhibitors (PPIs) in patients 65 or above [20]. Interestingly, PPIs were most common QTc interval-prolonging medications in our study. Although being classified as list 3 medications (conditional risk), numerous studies have confirmed the increased risk of hypomagnesemia, and what is also clinically important is that the association was dose-dependent [20, 21]. These findings strongly advocate increased physicians’ awareness of PPI-induced hypomagnesemia and to routinely monitor serum magnesium concentrations in older patients. Moreover, a continuous prospective benefit/risk assessment should be performed in older patients, especially with long-term or higher dose PPI use. The clinical approach is indeed more challenging in older patients, due to more comorbidities, complex disease states, and higher demands for diagnostic interventions. However, as there is strong evidence of an association between adverse cardiovascular outcomes and LQTS, the use of ECGs, as a ubiquitous, available, noninvasive, and cost-effective method is strongly advocated [22].

The results of our study confirmed a statistically significant increase in LQTS incidence in patients with diagnosed atrial fibrillation, both at admission and during hospitalization. Previous studies suggest that the association between QTc interval and atrial fibrillation is not solely due to traditional cardiovascular risk factors but also due to inherent characteristics or remodeling of a patient’s cardiac electrophysiology [23]. Moreover, the multivariate analysis identified male sex and amiodarone therapy as potential predictors of LQTS during hospitalization. Male sex increases the odds of having LQTS by 3.45 times, whereas the odds of LQTS are 4 times higher in patients receiving amiodarone. Females are generally considered at greater risk for medication-induced TdP than males because the mean QTc interval is longer in women from adolescence until the sixth decade. Physiological evidence could partly explain this finding because estradiol promotes QTc lengthening while progesterone and testosterone shorten QTc [24]. However, this difference tends to disappear in older patients and previously published studies reported an increased prevalence of prolonged QTc in older males [14, 15, 25]. Amiodarone is an antiarrhythmic agent commonly used in clinical practice and is approved for the treatment of life-threatening ventricular tachyarrhythmias. Amiodarone is also known to prolong the QTc interval, but it is unlikely to induce TdP without additional risk factors, a characteristic seen with many medications that prolong the QTc interval. Nonetheless, some cases of amiodarone-associated TdP have been reported, especially when other risk factors are present [26].

This study emphasizes the need for close ECG measurement of the QTc interval as part of routine monitoring of medications side effects in geriatric practice. In geriatric medicine, more than in any other medical practice, the use of an interdisciplinary team has produced very successful results. Pharmacists are needed to suggest medication therapy monitoring, alternative medications, changes in dosing regimen or route of administration, and provide appropriate instructions for medication use [27]. Our findings emphasize the importance of vigilant QTc interval monitoring in older patients, particularly given the high prevalence of QTc-prolonging medication use and the frequent presence of patient-related risk factors. The frequency of ECG monitoring should be individualized, taking into account a patient’s specific risk factors, comorbidities, and medication regimen. Regular ECG monitoring typically involves conducting frequent assessments, especially for high-risk patients, during the initiation or adjustment of medications known to affect the QTc interval. For such patients, it is recommended to perform ECG monitoring at baseline, after starting the medication, and periodically thereafter – typically every 3–6 months or more often if clinical conditions warrant. Modifiable risk factors contributing to acquired QTc prolongation should also be evaluated appropriately and in a timely manner, especially when it comes to the coadministration of multiple medications affecting the QTc interval, and also to monitor the correction of electrolyte abnormalities. An important secondary finding in our study was a significant increase in hypokalemia from 12.4% at admission to 35.5% after 7 days. The prevalence of hypokalemia on geriatric wards has varied in other studies: from 9.7% to 21.3% in the total sample to 11.2%–29.4% in patients with LQTS [1, 28]. The rates of hypokalemia in our patients with LQTS during hospitalization (45.1%) were consistent with the rates of chronic kidney disease measured as hypercreatinemia (creatinine serum concentration above 106 μmol/L) (49%). However, of the patients with LQTS and hypokalemia during hospitalization (n = 23), only 16 patients were administered furosemide. Our results suggest the need to further investigate other medication classes that may influence QTc prolongation via an indirect mechanism by affecting electrolyte concentrations. Moreover, our result demonstrated that patients with newly developed LQTS during hospitalization had statistically more common hypomagnesemia. In line with that the management of QTc interval prolonging medications implies that modifiable risk factors should be corrected as soon and as much as possible [29]. Some authors suggested further stratification of patients according to the degree of abnormalities in electrolytes concentrations [10]. Moreover, our findings may indicate a better prediction of LQTS if scores are used that simultaneously assess the overall risk resulting from different patient- or therapy-related characteristics.

We must address the limitations of the present study. This was an observational study in a single center, and findings could limit the generalizability of the results. Patients with atrial fibrillation were included in the study; therefore, the QT measurements may be inaccurate. However, we excluded patients with extremely rapid or slow atrial fibrillation. Some previous studies also included patients with atrial fibrillation. The determination of QTc based on the Bazett formula has been criticized for overcorrection at high heart rates. Nevertheless, this formula is the most accessible and commonly used method for QTc determination [30]. Finally, the possible influence of pharmacokinetic interactions on medication-related QTc prolongation was not assessed in this study.

The prevalence of LQTS in geriatric inpatients was high and QTc interval-prolonging medications with known risk of TdP were commonly used during hospitalization. Male sex and amiodarone use were independently associated with LQTS in our cohort. Proton pump inhibitors were the most frequently prescribed class of medications. Considering the increasing evidence on the LQTS risk from proton pump inhibitors, mainly via hypomagnesemia, information on the benefit/risk ratio of this class of medications in an individual patient is needed. In addition, older patients had a high prevalence of modifiable patients- or therapy-related risk factors (electrolytes disturbances, administration of interacting medications), which are conditional risk factors for TdP. The results of our study suggest that older patients admitted to the hospital are at significant risk for QTc prolongation and are not routinely monitored. The study underscores the importance of close monitoring and awareness of QTc interval prolongation risks in older patients, advocating for routine ECG assessments in patients at risk and vigilant management of modifiable risk factors, especially the electrolytes.

Zvezdara University Medical Center Institutional Review Board, Office for Human Research Protections 1, approved the study (No. IRB00009457, IORG0007877; date: August 11th 2023).

The authors declare they have no conflicts of interest.

This research was funded by the Ministry of Science, Technological Development and Innovation, Republic of Serbia through two grant agreements with University of Belgrade – Faculty of Pharmacy (No 451-03-136/2025-03/200161 and No. 451-03-137/2025-03/200161).

Conceptualization: Ivana Baralić Knežević, Katarina Stefanović, Predrag Erceg, and Katarina M. Vučićević; methodology: Ivana Baralić Knežević, Milena Kovačević, Katarina Stefanović, and Katarina M. Vučićević; formal analysis and writing – original draft preparation: Ivana Baralić Knežević, Milena Kovačević, and Katarina M. Vučićević; investigation: Ivana Baralić Knežević, Jovana Aćimović, Katarina Stefanović, and Gordana Mihajlović; data curation: Jovana Aćimović, Katarina Stefanović, Predrag Erceg, and Gordana Mihajlović; writing – review and editing: Milena Kovačević, Katarina M. Vučićević, Katarina Stefanović, Predrag Erceg, and Gordana Mihajlović; supervision: Predrag Erceg, Gordana Mihajlović, and Katarina M. Vučićević. All authors have read and agreed to the published version of the manuscript.

Ivana Baralić Knežević and Milena Kovačević share first authorship of this article.

The data that support the findings of this study are available from the corresponding author upon reasonable request.