Drug-related bradycardia precipitating hospital admission in older adults: an ongoing problem

Abstract

Background

Drug-related bradycardia (DRB) is a common clinical conundrum and can result in multiple hospital admissions as a result of the increased prescription of rate-limiting medications that can predispose to presyncopal or syncopal episodes.

Aim

To evaluate the incidence of DRB in elderly hospital inpatients.

Methods

We conducted a retrospective analysis of all patients admitted to our acute medical unit between November 2018 and February 2019 and identified patients over the age of 70 with more than one diurnal bradycardic episode during their admission. We extracted patient demographics, presenting complaint, admission 12-lead ECG and medications from the hospital electronic database.

Results

We screened 2312 adults and identified 100 patients over the age of 70 years with two or more episodes of diurnal bradycardia during their hospital admission. This constituted 4.32% of total admissions. Beta blockers were the most commonly prescribed rate-limiting medication (n=54, 87.1%), of which bisoprolol was the most frequently prescribed (n=41) and sinus bradycardia was the most commonly identified rhythm disturbance in our cohort of patients (n=41, 41%). Syncope was the most common presenting symptom and occurred in 23 patients, 14 (60.9%) of which were diagnosed with a DRB. Atrial fibrillation was more common in those with DRB compared with those with bradycardia not caused by medications (35.5% vs 10.5%, p=0.006), and atrial fibrillation was a significant predictor of DRB (OR=10.2, 95% CI 3.3 to 31.6, p<0.001).

Conclusion

Bradycardia is a significant cause of hospital admissions in older adults and can be avoided with pharmacovigilance. Caution should be exercised when initiating or changing the dose of rate-limiting agents in these patients; while those with atrial fibrillation should undergo regular review of their heart rate followed by appropriate medication dose adjustments.

Introduction

Drug-related bradycardia (DRB) is a common clinical problem and can result in multiple hospital admissions.¹ Patients may present with presyncope, syncope or injuries sustained from a syncopal episode. Common culprits of DRB include beta-adrenergic antagonists (beta blockers), non-dihydropyridine (DHP) calcium channel blockers and digoxin. These medications act to supress conduction through the atrioventricular (AV) node and therefore are commonly used to treat atrial tachyarrhythmias. Beta blockers also have an established role in the treatment of heart failure, ischaemic heart disease and hypertension.^1–4

With an ageing and increasingly comorbid population, the prevalence of these conditions is increasing, and hence, the prescription of these medications is also rising. With the increasing use of rate-limiting drugs, there may also be an increased incidence in hospital admissions for DRB.^{5 6}

The primary aim of our study was to evaluate the incidence of DRB in older hospital inpatients. We further evaluated the effects of having an admission with DRB on morbidity and mortality.

Methods

We conducted a retrospective analysis of all patients admitted from the emergency department to the acute medical unit and the elderly care unit at our institution between November 2018 and February 2019. During this period, we screened 2312 hospital admissions for bradycardia. For the purpose of our study, we defined bradycardia as a heart rate (HR) of less than 60 beats/min on two or more occasions during the daytime (nocturnal bradycardic episodes were not deemed significant). We also excluded patients based on age; therefore, only patients over the age of 70 years were included in our analysis.

The case notes from the hospital database were used to obtain the presenting complaint, admission ECG, medication chart and cardiac background of these patients. We used this information to establish whether the bradycardia identified was caused by prescribed medications and whether it was the reason for admission to the hospital.

Patients who had the following rhythms were included in our study: sinus bradycardia, second-degree AV block, third-degree AV block and slow atrial fibrillation. We documented if the patients in our study were taking rate-limiting medication. Patients were excluded from the study if there was a clear concomitant cause of their bradycardia, unrelated to their medications such as pacemaker failure, hypothyroidism, infection or hypothermia.

Patients were followed up for 12 months to assess the impact of DRB on all-cause mortality, hospital admission rates and pacemaker implantation rates.

Statistical analysis

Results are expressed as mean±SD, absolute numbers (N) and frequencies (%). Comparisons were made between those deemed to have DRB and those not deemed to have DRB. Chi-squared tests and Fisher’s exact tests were used to compare nominal variables and the students T-test was used for the comparison of continuous variables. The correlation between DRB and patient outcomes was determined by binary logistic regression analysis and the results were expressed as an OR and 95% CI. Results with p values<0.05 were regarded as significant.

Results

During the 4-month study period, we identified 100 patients over the age of 70 years who were admitted with two or more episodes of bradycardia. This translated to 4.32% of total admissions.

The patients’ baseline characteristics and regular medications are summarised in table 1. DRB was significantly more common than bradycardia not related to medications, with nearly two-thirds of patients identified subsequently being diagnosed with DRB (62% vs 38%, p<0.001). Causality assessment via Naranjo algorithm demonstrated a definite causal association between the administration of rate-limiting medications and bradycardia in these patients (score of 9).⁷

Table 1.

Summary of the baseline characteristics, reason for admission, cardiac rhythm on admission and medications in our cohort of patients

	Drug-related bradycardia (N=62)	Bradycardia not caused by rate-limiting medications (N=38)	P value
Baseline characteristics
Age (years)	83.45±6.15	83.79±6.25	0.803
Atrial fibrillation	22 (35.5%)	4 (10.5%)	0.006
Hypertension	17 (27.4%)	8 (21.1%)	0.482
Ischaemic heart disease	11 (17.7%)	10 (26.3%)	0.308
Heart failure	5 (8.1%)	3 (7.9%)	0.972
Previous stroke	2 (3.2%)	2 (5.3%)	0.605
Diabetes mellitus	7 (11.3%)	4 (10.5%)	0.902
Dementia	7 (11.3%)	5 (13.2%)	0.778
Pulmonary disease	12 (19.4%)	9 (23.7%)	0.610
Reason for admission
Syncope	15 (24.2%)	8 (21.1%)	0.730
Presyncope	3 (4.8%)	5 (13.2%)	0.135
Cardiac rhythm on admission
Sinus bradycardia	28 (45.2%)	13 (34.2%)	0.280
Second-degree atrioventricular block	1 (1.6%)	1 (2.6%)	0.729
Complete heart block	1 (1.6%)	1 (2.6%)	0.729
Atrial fibrillation with a slow ventricular response (HR<60 beats/min)	12 (19.4%)	5 (13.2%)	0.426
Atrial fibrillation with a normal ventricular response (HR=60–100 beats/min)	5 (8.1%)	3 (7.9%)	0.972
Medications
Beta blockers	54 (87.1%)
Atenolol	6
Bisoprolol	41
Metoprolol	1
Nebivolol	2
Propranolol	1
Sotalol	3
Calcium channel blockers	4 (6.5%)
Diltiazem	2
Verapamil	2
Amiodarone	1
Digoxin	3

HR, heart rate.

Atrial fibrillation was significantly more common in those with DRB, compared with those with bradycardia not caused by medications (35.5% vs 10.5%, p=0.006), and atrial fibrillation was a significant predictor of DRB (OR=10.2, 95% CI 3.3 to 31.6, p<0.001).

The most common presenting symptom in our cohort of patients was syncope (23 patients), of which, 14 (60.9%) were diagnosed with a DRB, while of the 8 patients presenting with presyncope, 3 (37.5%) were diagnosed with a DRB.

The patient’s diurnal bradycardic episodes ranged from an HR of 45–60 beats/min on two or more occasions during their hospital admission. ECGs were analysed on the day of admission, and 70% of the patients had bradycardia on their initial ECG (table 1).

The following rhythms were recorded on the admission ECG: sinus bradycardia (42 patients), sinus bradycardia with junctional escape beats (2 patients), second-degree AV block (2 patients), complete heart block (3 patients), atrial fibrillation with slow ventricular response (21 patients) and atrial fibrillation with normal ventricular response (5 patients). The admission ECG did not always capture the bradycardia subsequently documented in observation charts; 21 patients had admission ECGS with rates more than 60 beats/min. In these patients, the rhythms recorded were normal sinus rhythm, bifasicular block and trifasicular block, and four patients did not have an admission ECG on the database.

Each patient was followed up for 12 months after his or her acute admission to hospital. During the follow-up period, 30 patients had died, 18 (60.0%) of which had been diagnosed with DRB. Five patients underwent implantation of a permanent dual-chamber pacemaker, four for treatment of complete heart block (one of which occurred post-transcatheter aortic valve implantation) and one for symptomatic sick sinus syndrome, three (60.0%) of which had been diagnosed with DRB. Nineteen patients were readmitted during the follow-up period, 14 (73.7%) of which had been diagnosed with DRB. However, an admission with DRB was not a significant predictor of all-cause mortality (OR=0.89, 95% CI 0.36 to 2.13, p=0.787), pacemaker implantation (OR=0.92, 95% CI 0.146 to 5.74, p=0.924) or readmission to hospital (OR=1.92, 95% CI 0.63 to 5.86, p=0.249).

Forty-four patients in the DRB cohort were alive at 12 months. Of these patients, only six (13.6%) had their rate-limiting medications reduced or stopped, and of these patients, one was readmitted with pneumonia. However, of the 38 patients who did not have any adjustments made to their rate-limiting medications, 13 (34.2%) were readmitted to the hospital, 4 of which were admitted with syncopal episodes.

Discussion

Our study demonstrates that DRB is a common comorbidity in the elderly population and can result in hospital admissions secondary to syncopal or presyncopal episodes. Within our study population, 23 of the patients had presented to the hospital with a syncopal episode, of which 14 (60.9%) were subsequently diagnosed with a DRB.

Over half of the patients in our study were diagnosed with a DRB. Beta blockers were the most commonly prescribed rate-limiting medication, of which bisoprolol was the most frequently prescribed and sinus bradycardia was the most commonly identified rhythm disturbance in this cohort, as demonstrated in figure 1.

Figure 1.

Flowchart of patients admitted with bradycardia. NDHP, non-dyhydropyridine.

Atrial fibrillation was more common in those with DRB and was a predictor of DRB in our study population. Patients initially diagnosed with atrial fibrillation often have a rapid ventricular response and require rate-limiting medications such as beta-blockers and non-DHP calcium channel blocker drugs to mediate their HR. As atrial fibrillation progresses, there is electrical remodelling and changes in the expression of ion channels within the cardiomyocytes, along with upregulation of inflammatory cytokines, resulting in myocardial fibrosis. This, in turn, slows conduction through the cardiomyocytes and leads to a delayed ventricular response, also known as ‘slow atrial fibrillation’. This process occurs over many years and can often go unnoticed. Without regular medication reviews, these elderly patients may continue taking rate-limiting drugs despite a gradual reduction in their HR.^{5 8}

In a rapidly ageing global population, it is likely we will see an increase in hospital admissions caused and complicated by bradyarrhythmias. In the elderly comorbid patient, it may initially be difficult to isolate the underlying aetiology, which can often be multifactorial. Common pathologies resulting in bradycardia and chronotropic incompetence include sinus node dysfunction (SND, also referred to as sick sinus syndrome⁹), AV block and an iatrogenic sequela of treatment with various cardiovascular medications. In cases of DRB, there is further uncertainty as to whether the extrinsic pharmacology is the primary cause or simply uncovering an existing underlying cardiac pathology. Review of current literature suggests that although cessation of the suspected culprit agent is certainly a first-line management strategy, conduction dysfunction, particularly at the AV node, may persist or recur.¹⁰

It is widely accepted that DRB can occur in response to several common cardiovascular medication classes, particularly beta blockers, non-DHP calcium channel blockers and digoxin.¹¹ Beta blockers are widely used to improve outcomes in ischaemic heart disease, atrial arrhythmias, hypertension and chronic heart failure,¹² all of which are prevalent in the geriatric population. These conditions are also often implicated in the pathogenesis of intrinsic SND and/or AV block,⁵ and therefore, in a frail and physiologically vulnerable patient group, it is necessary to evaluate and balance the prognostic benefits of negatively chronotropic agents.

It can be difficult to ascertain whether prescribed medications are the ‘true cause’ of the underlying bradycardia. Zeltser et al analysed patients treated with beta blockers and non-DHP calcium channel antagonists and suggested that these medications were implicated in only 15% of bradycardias, and in most cases, the suspected drugs were found to be ‘bystanders’.¹⁰ A more plausible explanation is that DRB may ‘unmask’ underlying sinus and/or AV node disease, and this is supported by many studies demonstrating that therapeutic doses of these agents do not cause symptomatic bradycardia in structurally normal hearts. This concept also raises a dosing dilemma since many elderly comorbid patients are underweight and sarcopenic. Due to standard dosing regimens, these patients may not be receiving the correct drug dose based on their weight and size. Therefore, we recommend caution when initiating such medications and careful up-titration to therapeutic dosages in older patients.

Limitations

Our study was conducted within a single centre with a relatively small cohort size, and therefore, these results cannot be generalised to other centres. Furthermore, the ECGs that were analysed were recorded on admission and not necessarily at the time of the bradycardic episode; therefore, the incidence of bradyarrhythmias may have been underestimated. Continuous cardiac monitoring would provide a more reliable assessment of the HR and cardiac conduction abnormalities, and a larger prospective study of elderly inpatients admitted with syncope/presyncope that used continuous cardiac monitoring would be of interest.

Conclusion

This study has demonstrated that bradycardia remains a significant cause for admission to the hospital in older patients who are inherently vulnerable to DRB. These admissions can potentially be avoided with careful prescribing practices and regular review of rate-limiting medications. We identified that atrial fibrillation was a predictor of DRB, and the most common cause of DRB was beta blockers. DRB can result in syncopal episodes, injuries and subsequent prolonged hospital inpatient episodes. In order to minimise adverse events, caution should be exercised when initiating or changing the dose of rate-limiting medications in older adults, while those with atrial fibrillation should undergo regular review of their HR followed by appropriate medication dose adjustments.

What this paper adds.

What is already known on this subject

• Drug-related bradycardia (DRB) is a common clinical problem which can result in multiple hospital admissions for older adults.

• Common culprits are beta blockers and non-dihydropyridine calcium channel antagonists.

• Because of a high prevalence of cardiovascular comorbidities within an increasingly ageing population, negatively chronotropic agents are frequently prescribed, and we set out to determine the incidence of DRB in the current era.

What this study adds

• A background of atrial fibrillation is a predictor of DRB.

• DRB may ‘unmask’ chronotropic incompetence from underlying sinus and/or atrioventricular node dysfunction.

• Older adults may continue taking rate-limiting drugs despite a gradual reduction in their heart rate.

Data availability statement

Data are available upon reasonable request.

Ethics statements

Patient consent for publication

Not required.

Ethics approval

This survey and retrospective analysis were approved by the trust audit department with reference CB426. As a registry report using clinically collected, non-identifiable data, this work does not fall under the remit of NHS Research Ethics Committees. All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards.