Implantable Cardiac Rhythm and Hemodynamic Monitors

Clinical Utility

The first ILR was a modified dual-chamber pacemaker generator with sensing electrodes on the lower aspect of the can and on a modified leadless header. It was implanted in 24 patients with recurrent syncope that remained unexplained despite extensive investigation. In this initial group, the device was able to establish a symptom-to-rhythm correlation in 21 patients (88%).

In a subsequent study, Krahn et al evaluated the benefit of prolonged rhythm monitoring with the first-generation Reveal ILR (Medtronic, Minneapolis, MN) in 85 patients with recurrent unexplained syncope after a negative initial evaluation that included tilt-table testing in 49% and electrophysiology study in 43%. Eligible patients had experienced at least two syncopal events within the preceding 12 months, with a mean of 5.1 ± 5.5 events, and 71% of patients had had symptoms for more than 2 years. Enrolled patients had a mean age of 59 ± 18 years, 52% were male, and concomitant cardiovascular disease was present in 62%. During a mean follow-up of 10.5 ± 4 months, symptoms recurred in 58 patients (68%) at an average of 2.3 ± 2.6 months after device insertion. An arrhythmia was diagnosed in 42% of the 50 patients in whom the rhythm was documented. Failure to activate ECG storage occurred in eight patients. There were no adverse events associated with recurrent syncope and no episodes of sudden death. Pocket infection occurred in three patients, two of whom had not received antibiotic prophylaxis at implantation. Within the population studied, the strategy of prolonged monitoring with an ILR was demonstrated to be safe and effective.

In further series, researchers found similar findings in patients with unexplained recurrent syncope with prior negative investigations.

In the Randomized Assessment of Syncope Trial (RAST), 60 patients (mean age of 66 years) were randomly assigned to conventional testing with Holter monitors, tilt testing, and electrophysiologic studies or prolonged monitoring with ILR for 1 year. A diagnosis was reached in 47% of ILR patients compared with 20% in conventional testing. A crossover period was offered to patients in whom a diagnosis was not found. This produced a diagnosis in a further 8 of 21 patients who went on to have an ILR placed. Although potentially biased by the exclusion of patients with a history typical of neurally mediated syncope and the long history of syncope in enrolled patients (on average >6 years), the study supports early implementation of ILR monitoring while also indicating tilt testing and electrophysiologic study appear to have limited value in this population.

A cost analysis done in the RAST study concluded that the ILR group had a higher up-front cost but a lower cost per diagnosis and ultimately a lower cost overall. In the prolonged monitoring first followed by crossover to conventional testing group, the cost was $2937 per patient, or $5875 per diagnosis; in the conventional testing first followed by crossover to prolonged monitoring group, the cost was $3683 per patient ( P = 0.013), or $7891 per diagnosis ( P = 0.02).

The Eastbourne Syncope Assessment Study (EasyAs) researchers followed 201 elderly patients (mean age, 74 years) with unexplained syncope with negative tilt testing who were randomly assigned to either ILR or conventional testing. ILR established an ECG diagnosis in 33% of patients compared with 4% of patients with conventional testing. The investigators reported a £809 saving per patient in the ILR group because of fewer hospital days and fewer investigations, although this did not include the cost of the device or its implantation, which was estimated to be £1350 at the time.

A study done with the Place of Reveal In the Care Pathway and Treatment of patients with Unexplained REcurrent syncope (PICTURE) registry reported outcomes in 570 patients with ILR implants. Recurrent syncope occurred in 38% of patients after a mean follow-up of 10 ± 6 months. Data from the ILR directly contributed to a diagnosis in 75%. Interestingly, patients had undergone a median of 13 other prior diagnostic tests, suggesting that early ILR implant could potentially eliminate some unnecessary testing and save costs.

Unfortunately, these studies did not take wider health costs into consideration, such as lost income to patients, families, and businesses associated with delays in diagnosis. Overall, the economic case for ILR use is suggestive but far from definitive. Further studies are needed that incorporate more detailed health economic analyses, with the use of ILRs readily assessable through cost–utility analysis.

The safety and efficacy of early ILR use was assessed by Brignole et al in 392 patients with recurrent, suspected, neurally mediated syncope. Patients with significant ECG or cardiac abnormalities were excluded, as were patients with orthostatic hypotension and carotid hypersensitivity. Although there was no control group, the risks of delaying therapy until diagnosis were low. Syncope occurred in 106 patients, of whom 53 received therapy; almost all required pacemaker insertion. Seven patients (2%) experienced major trauma from syncope, and 1% had ILR pocket infections.

When an ILR is implanted in patients with recurrent unexplained syncope, the most common arrhythmic finding is transient bradycardia. Krahn et al report a study of 206 patients in which 69% experienced recurrent syncope within 6 months; bradyarrhythmia was noted in 17%. Sinus rhythm was documented in 42% and tachyarrhythmia in 6%, and the device failed to activate in 4%. Age was the only independent variable that predicted the need for pacing and was associated with a 3% increase in risk per advancing year of age. However, no age group could be identified in which the likelihood of requiring pacing exceeded 30%, questioning the role of empiric pacing in this patient population.

The observation that bradyarrhythmic mechanisms dominate in patients with bundle branch block, recurrent syncope and an ejection fraction ≥35% raises the issue of empiric pacemaker implantation as an alternate strategy to ILR in this setting. The Syncope: Pacing or Recording in the Later Years (SPRITELY) study is currently being conducted to address this question. The investigators plan to enroll 120 patients with bifascicular block, preserved left ventricular function, and at least one syncopal spell in the preceding year with randomization to pacemaker or ILR implantation in an open-label, parallel-group study. The primary outcome measure will be a composite of major adverse study-related events in a 2-year observation period, where the events will be death, syncope, symptomatic bradycardia, asymptomatic diagnostic bradycardia, and acute and chronic device complications.

The ILR has been evaluated in patients at significant risk of developing malignant arrhythmia, who are currently asymptomatic, and in whom specific therapeutic intervention is not currently indicated. The Cardiac Arrhythmias and Risk Stratification After Acute Myocardial Infarction (CARISMA) study enrolled 297 patients, mean age 64.0 ± 11.0 years, left ventricular ejection fraction 31 ± 7%. During follow-up every 3 months for an average of 1.9 years, predefined bradyar­rhythmias and tachyarrhythmias were recorded in 137 patients, 86% of whom were asymptomatic. The implantable cardiac monitor documented a 28% incidence of new-onset AF, a 13% incidence of nonsustained ventricular tachycardia (≥16 beats), a 10% incidence of high-degree atrioventricular block (≤30 beats per minute [bpm] lasting ≥8 seconds), a 5% incidence of sinus arrest (≥5 seconds), a 3% incidence of sustained ventricular tachycardia, and a 3% incidence of ventricular fibrillation. Cox regression analysis with time-dependent covariates revealed that high-degree atrioventricular block was the most powerful predictor of cardiac death (hazard ratio, 6.75; 95% confidence interval, 2.55 to 17.84; P < 0.001). Although primary prevention implantable defibrillator therapy is now indicated in many of the patients enrolled in the CARISMA trial, the observations suggest a potential role in patients with structural heart disease of lesser severity where clinical guidance is uncertain.

The ILR has also been used for evaluation of generalized tonic-clonic seizures. The specific myopotential pattern observed during generalized seizures because of diffuse muscle contractions is readily seen by the ILR and appears different from patterns seen with either convulsive syncope or nongeneralized seizures. In a study including 14 patients with documented refractory epilepsy, an ILR was implanted as part of a protocol to evaluate cardiac rhythm abnormalities in patients at risk for sudden death. Tonic-clonic seizure episodes were detected by the ILR in six patients (43%). However, the use of ILR for diagnosis of generalized seizure episodes bears some limitations, as rapid frequency of myopotential artifacts may exceed the nonprogrammable bandpass filter of the device (32 Hz), preventing any automatic detection. The inability to automatically capture these high-frequency myopotentials highlights that their absence upon ILR interrogation cannot formally exclude seizures, except when manually activated.

The ILR has been proposed for AF monitoring. Montenero et al reported the results of a pilot study of nine candidates for ablation with frequent and highly symptomatic episodes of drug-refractory AF. Six patients received implants of an ILR and were monitored 1 month before and 6 months after the procedure. Overall, 178 events were recorded; 30% were patient-activated and 70% were automatically activated. Of note, one third of recordings were inappropriate, largely due to premature atrial or ventricular beats. In subsequent studies, researchers have encountered problems with inappropriate detection, although it is hoped that algorithm refinements may reduce the false-positive rate (see later).

In the Discerning the Incidence of Symptomatic and Asymptomatic Episodes of Atrial Fibrillation Before and After Catheter Ablation (DISCERN AF) study, 50 patients underwent ILR implant 3 months before AF ablation. Total AF burden was reduced by 86% from a mean of 2.0 ± 0.5 hours/day per patient before to 0.3 ± 0.2 hours/day per patient after ablation ( P < 0.001). Fifty-six percent of all episodes were asymptomatic. The ratio of asymptomatic to symptomatic AF episodes increased after ablation from 1.1 to 3.7 ( P = 0.002). Symptomatic assessment underestimated the number of patients with AF recurrence postablation, with 12% having only asymptomatic episodes, reducing the percentage of patients free of AF from 58% to 46%.

The Cryptogenic Stroke and Underlying Atrial Fibrillation (CRYSTAL AF) trial investigators randomized 441 patients with cryptogenic stroke (i.e., with no prior history of AF and well investigated with no other cause of stroke found) to ILR insertion or control. The latter consisted of 12-lead ECGs at the time of scheduled follow-up visits or if the patient was symptomatic. AF was detected in 8.9% of patients in the ILR group and 1.4% in the control group ( P < 0.001) at 6 months. Infection or discomfort led to the extraction of 2.4% of ILRs.

Although this suggests an important role for ILR implantation in cryptogenic stroke, the study design has several limitations. Among these is that the control intervention of conventional follow-up can be questioned. Although pragmatic and typical of usual care at the time, this intervention consisted of routine clinical follow-up augmented by further Holter monitoring in a minority of subjects. As such, implantable cardiac monitoring was not compared with the best noninvasive strategy. Illustrating this point, the researchers in the Event Monitor Belt for Recording Atrial fibrillation after a Cerebral ischemic Event (EMBRACE) study, published at the same time as the CRYSTAL AF trial, reported a 16.1% incidence of AF episodes of 30 seconds or longer in patients randomized to an additional 30-day event-triggered ILR compared with 3.2% in those randomized to further Holter monitoring.

Additionally, the significance of brief episodes of AF detected with an implantable cardiac monitor is unknown, CRYSTAL AF was not powered to address clinical outcomes, and cost-effectiveness was not assessed. The continuing divergence in AF incidence between the implantable cardiac monitor and conventional follow-up limbs at 3 years is also of interest. Although several explanations for this result are possible, given the progressive nature of AF, which should lead eventually to catch-up in the control limb, this observation raises the possibility that not all AF events detected by ILR are necessarily indicative of a progressive AF substrate. Finally, although a strategy of selective anticoagulation of patients with ILR-detected AF is rational and may subsequently be demonstrated to improve clinical outcome, the possibility that routine anticoagulation may be equally or more effective in subsets of the cryptogenic stroke population requires investigation.

The ILR has a firmly established role in the investigation of recurrent unexplained syncope and has been demonstrated to be cost-effective in that setting. Currently, its application for other indications is largely investigational; however, by providing long-term continuous rhythm monitoring, it has the power to advance understanding of arrhythmias, to inform clinical trials, and to improve patient outcomes.

Medtronic Reveal

The Reveal family of ILRs record a single-lead subcutaneous ECG from a pair of electrodes on the device surface. It is implanted in the left pectoral region with the patient under local anesthetic using a standard sterile technique. The proposed site can be mapped before implant by placing the device on the skin and interrogating it to ensure an adequate vector will be obtained; this is not generally required, however. A short incision is made and carried down to the pectoral fascia, a subcutaneous pocket is fashioned into which the device is inserted and sutured in place, and the incision is then closed. Gain and sensitivity are adjusted according to manufacturer recommendations. Prophylactic antibiotic therapy is recommended to prevent pocket infection.

The original ILR device (Reveal model 9525) was released in 1998 and had a minimum battery life of 14 months. It measured 61 × 19 × 8 mm and weighed 17 g. The recorded signal was maintained in a circular buffer capable of storing 21 minutes of uncompressed signal or 42 minutes of compressed data, triggered by use of a handheld patient activator. The second-generation ILR, released in 2000 (Reveal Plus model 9526) added the capacity for automatic detection and capture of arrhythmic events (asystole, bradycardia, and tachycardia) according to physician-programmed settings. It is nominally set to record heart rates less than 30 bpm, more than 160 bpm, or pauses longer than 3 seconds. Multiple 2-minute autoactivated episodes could be stored together with either one 14-minute patient-activated event or three 10-minute patient-activated events.

The Reveal DX and Reveal XT ( Fig. 25-1 ) provide significant enhancements, including longevity of up to 3 years, autodetection of ventricular tachyarrhythmia and fast ventricular tachyarrhythmias, magnetic resonance imaging conditional status, and the capability of remote monitoring via the Medtronic CareLink Network. Total memory capacity is increased to 49.5 minutes, allowing for 22.5 minutes of patient-activated episodes and 27 minutes of automatically activated episodes. The Reveal XT has the added feature of an AF detection algorithm allowing automatic capture of atrial tachyarrhythmia and AF episodes. It also provides information regarding average heart rate, heart rate variability, activity throughout the day, and rhythm trending data, including up to 14 months of daily atrial tachycardia/AF burden and ventricular rate during atrial tachycardia/AF.

The Reveal system includes the Patient Assistant, a handheld, battery-operated device that communicates via telemetry to the Reveal. It is used to activate the recording function of the implanted cardiac monitor, storing an event in the device's memory. The Reveal XT version of the Patient Assistant has a query function that signals patients to call their physician or clinic when programmed arrhythmia detection or device criteria are met. The Medtronic 2090 Programmer allows in-office device programming and data retrieval, and remote transmission of stored data is possible for patients enrolled in the CareLink Network.

The value of the autoactivation feature in the Reveal Plus model was compromised by frequent, inappropriate detections. In a series of 50 patients, 682 autoactivations were collected over an 18-month period; 83% were inappropriate, due to undersensing in 76% and oversensing in 24%. In addition to their potential to overwrite valid detections and waste time spent on review, inappropriate detections have the potential to result in false-positive diagnoses, as described in the report of Chrysostomakis et al, who found that sudden reduction in R-wave amplitude led to undersensing and the recording of an event mimicking ventricular standstill ( Fig. 25-2 ). An improved sensing and detection scheme was implemented in the Reveal DX and XT models. In an analysis of 2613 previously recorded, automatically detected Reveal Plus episodes from 533 patients in which 71.9% of episodes were inappropriate, the new algorithm reduced inappropriate detections by 85.2%. The number of patients with inappropriate detections was reduced by two thirds, with only a 1.7% decrease in appropriate detections.

The performance of the Reveal XT ILR for the detection of AF has been evaluated in the XPECT trial. In patients with known paroxysmal AF ( n = 247) and an implanted Reveal XT, 46 hours of subcutaneous ECG, ILR markers, and two surface ECG leads were recorded with a special Holter monitor. The ILR automatic arrhythmia classification was compared with the core laboratory classification of the surface ECG. Of the 206 analyzable Holter recordings collected, 76 (37%) contained at least 1 episode of core laboratory-classified AF. The sensitivity, specificity, positive predictive value, and negative predictive value for identifying patients with any AF were 96.1%, 85.4%, 79.3%, and 97.4%, respectively. Per-episode sensitivity for AF detection increased from 88.2% for AF of at least 2 minutes to 92.9% for episodes of at least 20 minutes. False-positive AF detections occurred in 14.6% of 130 patients, being triggered by frequent atrial or ventricular premature beats, myopotentials, irregular sinus rhythm, or R-wave undersensing. A substantially higher rate of false-positive AF detection was observed during continuous monitoring. In a series of 64 patients with paroxysmal AF, 72% of patients had at least one false-positive AF detection during a median monitoring period of 49 days (interquartile range 23-99 days). Due to the high rate of AF misdiagnosis, a software update was introduced in June 2008. The noise rejection threshold was reduced from 60 seconds to 5 seconds, resulting in the rejection of episodes containing at least 5 seconds of noise in any 2-minute window. Incorrect AF detections were reduced from 72% to 44% of patients ( P = 0.001). Compared with serial 7-day Holter monitoring, the ILR had a tendency to detect more patients with AF recurrences (31% vs. 24%; P = 0.125). However, limited specificity remained postupgrade, potentially undermining the utility of the AF detection algorithm in clinical practice.

Medtronic Reveal LINQ

The Reveal LINQ is significantly smaller than its predecessors (see Fig. 25-1 ). It has a lithium carbon monofluoride battery with expected longevity of 3 years, dimensions of 44.8 × 7.2 × 4.0 mm, volume of 1.2 cm ³ , and weight of 2.5 g. Despite its small size, its memory capacity has been increased to a total of 59 minutes, allowing 30 minutes of patient-activated episodes and 27 minutes of automatically detected episodes. The extended memory allows the storage of two to four patient-activated events. Atrial tachycardia and AF detection capability is standard, and enhancements to the AF detection algorithm have been made that are predicted to reduce false-positive AF detection by up to 46% while maintaining sensitivity for AF detection. The clinical performance of the Reveal LINQ remains to be reported. The Reveal LINQ is magnetic resonance imaging conditional and connects wirelessly to the wireless connectivity to Medtronic CareLink system.

St. Jude Medical Confirm

The Confirm (St. Jude Medical, St. Paul, MN) is similar in shape to, but smaller than, the Reveal DX or XT devices with dimensions of 56.3 × 18 × 8 mm, volume of 6.5 mL, and weight of 12 g (see Fig. 25-1 ). It has a battery life of 3 years and ECG storage of up to 48 minutes. A maximum of 4 minutes of ECG before manual activation can be recorded, compared with 14 minutes with Reveal LINQ. Given prior reports of a delay between bradycardia onset and manual activation of more than 4 minutes in 24% of patients subsequently receiving an ILR indicated pacemaker, this could be a limitation of the Confirm device.

Biotronik Biomonitor

The Biomonitor (Biotronik, Berlin, Germany) is shaped like a pacemaker, has dimensions of 53.3 × 42.7 × 7.1 mm, a volume of 12.5 cm ³ , and a weight of 26 g. The header contains two electrodes with a third on the bottom of the housing (see Fig. 25-1 ). Signals from the three sensing vectors are filtered and summed to provide a single ECG output with a signal-to-noise ratio 1.8- to 3-fold higher and an average increase in R-wave of 60%. Subsequent beat-to-beat R-wave analysis classifies each signal as authentic or noise ( Fig. 25-3 ). Potential advantages of the multivector input include no requirement for vector mapping at device implantation and enhanced specificity of arrhythmia detection. The Biomonitor can store a total of 35.8 minutes of subcutaneous ECG. Automatic detection of asystole, bradycardia, AF, and ventricular high-rate events can be programmed with up to 40 seconds of ECG per episode. Patient-triggered events of up to 7.5 minutes can be stored by application of the Biotronik M-50 magnet over the device for 1 to 2 seconds. Recorded episodes can be automatically transmitted via the Biotronik home monitoring system. With daily transmission, projected longevity is up to 5.1 years; with weekly transmission, longevity may be extended to 6.4 years.

Sleuth AT

The Sleuth implantable ECG monitor was released in 2007. The Sleuth ECG Monitoring System included a high-definition ILR, a Personal Diagnostic Manager, a Base Station, and a Monitoring Center staffed 24/7. An improved device, the Sleuth AT, was approved in 2009. It was able to communicate wirelessly with a personal device carried in a pocket or purse or with a home base station that could automatically contact a patient's physician if an abnormal rhythm was detected. Despite encouraging trial data, production of the Sleuth AT stopped in December 2009 due to its producer, Transoma Medical, being unable to secure further development funding.

Monitoring Features of Implantable Pacemakers and Defibrillators

Pacemakers and ICDs are capable of performing as continuous monitoring devices. Modern dual-chamber pacemakers and ICDs have built-in algorithms to detect supraventricular arrhythmias. To avoid tracking the tachycardia, the device will mode-switch from DDD to VVI. This continues until the atrial arrhythmia terminates, at which time the device returns to dual-chamber tracking mode.

The device can provide a detailed arrhythmia log that contains the number, duration, and dates of arrhythmia episodes, as well as their maximal atrial and ventricular rates. Many can record intracardiac electrograms from arrhythmic episodes. In addition, devices may describe arrhythmia trends, generally presented as a histogram, which show the proportion of the day spent in atrial arrhythmia for the previous 6 months. When appropriately programmed, implanted devices will identify AF with very high sensitivity and specificity.

Monitoring ventricular tachyarrhythmia and its response to device therapy is a critical component of ICD function, but lies outside the scope of this chapter.

The development of AF is associated with increased morbidity and mortality in patients with heart failure. Hemodynamics may deteriorate as a result of the resulting rapid ventricular rate, loss of atrial kick, or loss of ventricular synchrony. There is an association with an increase in ventricular arrhythmias and inappropriate ICD therapy. These patients have an increased risk of stroke. AF detection is therefore important, and it stands to reason that many patients with implanted devices would benefit from prompt introduction of anticoagulation and rate or rhythm control medication. Furthermore, it is possible that AF burden, perhaps combined with conventional risk factor scoring, might be an important factor in predicting potential stroke risk. Implantable cardiac monitors have furthered knowledge of AF and thromboembolic risk.

Implantable Cardiac Devices and Atrial Fibrillation

AF may be completely asymptomatic and can therefore go unnoticed. It is not uncommon for so-called silent AF to be detected on routine ECG. Although approximately 15% of strokes are attributed to AF, 25% to 33% of strokes are labeled cryptogenic. It has long been suspected that some of these cryptogenic strokes may be due to undetected, asymptomatic AF. With the advent of ILRs, it has become clear that a significant proportion of patients with cryptogenic stroke will be found to have silent AF if a clinician looks long and hard enough, even if a prior Holter monitor reading has been negative.

The incidence of AF detection in the pacemaker patient population is high, often around 40%, and it is reasonable to expect a significant proportion will be silent (see Case Study 25-1 ). In a study of 110 patients with permanent pacemakers undergoing attempted rhythm control treatment of symptomatic AF, the device detected episodes longer than 48-hour duration in 50 patients, 38% of whom were asymptomatic.

Case Study 25-1

Atrial Fibrillation Detected by a Permanent Pacemaker

Patient

• Age: 83 years

• Gender: Male

• Occupation: Retired

• Working diagnosis: asymptomatic atrial fibrillation

History

An 83-year-old man presented to an outpatient clinic for a routine device check. His pacemaker (Ensura generator, CapSure Sense atrial lead 4574, CapSureFix 5076 right ventricular lead; Medtronic, Minneapolis, MN) had been implanted 18 months before for symptomatic complete atrioventricular block. Following this procedure, he was able to return to his quiet, predominantly sedentary life. He experienced no cardiac symptoms but was somewhat limited by chronic ankle pain and had started to use a low walking frame. Despite this, he had recently enjoyed an overseas vacation.

His pacing check revealed satisfactory lead impedance, threshold, and sensitivity; however, there had been 56 mode-switch episodes, and one had lasted more than 48 hours. The stored electrograms were consistent with atrial fibrillation.

He had a past medical history of treated hypertension. Two years earlier, he had experienced a left paramedian pontine infarct that had left no significant sequelae. No dysrhythmia was documented at that time, despite use of a 24-hour Holter monitor.

Current Symptoms

There were no new symptoms associated with the development of atrial fibrillation.

Current Medications

• Aspirin 100 mg daily

• Atorvastatin 40 mg daily

• Bendrofluazide 2.5 mg daily

• Amlodipine 2.5 mg daily

• Ibuprofen 200 mg daily

Physical Examination

• Blood pressure, heart rate: 160/100 mm Hg, 80 beats/min

• Height/weight: 173 cm/61 kg

• Neck veins: Normal

• Lungs/chest: Normal

• Heart: Normal

• Abdomen: Normal

• Extremities: Normal

Laboratory Data

• Hemoglobin: 14.7 g/dL

• Hematocrit: 42%

• Platelet count: 240 × 1000 µL

• Sodium: 135 mmol/L

• Potassium: 3.8 mmol/L

• Creatinine: 98 µmol/L

• Blood urea nitrogen: 6.2 mmol/L

• Thyroid-stimulating hormone: 0.78 mIU/L (supports euthyroidism)

Chest Radiograph

A posteroanterior erect chest radiograph revealed the pacing leads in appropriate locations ( Fig. E25-1 ).

Electrocardiogram

A 12-lead electrocardiogram confirmed a return to normal sinus rhythm with paced QRS complexes ( Fig. E25-2 ).

Echocardiogram

A transthoracic echocardiogram obtained 6 weeks after pacemaker implant found a mildly dilated left ventricle with moderate systolic impairment. The atria were of normal size, and the valves were nearly normal.

Focused Clinical Questions and Discussion Points

Question: How often is atrial fibrillation detected like this?

Discussion: The incidence of atrial fibrillation detection in the pacemaker population is approximately 40%. ¹ A significant proportion of atrial fibrillation is silent, as evidenced by electrocardiograms in the general population. ² Furthermore, patients with symptomatic atrial fibrillation may be unaware of a significant number of their episodes. ³

Question: What is this patient's risk of thromboembolic events?

Discussion: The main risk factors for stroke are a past history of stroke and increasing age. A past history of hypertension, heart failure, or diabetes is also associated with increased risk. The CHADS ₂ and CHA ₂ DS ₂ -VASc scores are widely used and well-validated tools to assess an individual's stroke risk, although the latter is more powerful at identifying truly low-risk patients ( Table E25-1 ).

TABLE E25-1

CHA ₂ DS ₂ -VASc Score

CHA 2 DS 2 -VASc Categories	Score
C ongestive heart failure	1
H ypertension	1
A ge ≥75 y	2
D iabetes mellitus	1
S troke/TIA/thromboembolism	2
V ascular disease (myocardial infarction, peripheral vascular disease, or aortic plaque)	1
A ge 65-74 y	1
S ex c ategory (i.e., female)	1
Maximum score	9

CHA ₂ DS ₂ -VASc score of 0: recommend no antithrombotic therapy; CHA ₂ DS ₂ -VASc score of 1: recommend antithrombotic therapy with oral anticoagulant or antiplatelet therapy, but preferably antithrombotic therapy; CHA ₂ DS ₂ -VASc score ≥2: recommend oral anticoagulation.

TIA, Transient ischemic attack.

This man's CHA ₂ DS ₂ -VASc score is 6. He scores points for age, prior stroke, hypertension, and objective evidence of cardiac dysfunction on his echocardiogram. This places him in the high-risk group, and his estimated annual risk of stroke if off aspirin is 9.8% (6.9-13.5%). ^{4 , 5} Anticoagulant therapy is recommended.

Question: Isn't an elderly patient like this at high risk of bleeding complications?

Discussion: Many of the risk factors for stroke are also risk factors for serious bleeding while on anticoagulants. Calculating an individual's bleeding risk is a little more difficult, because there are fewer validated risk assessment tools. Of the number of available tools, the HAS-BLED score (hypertension, abnormal renal and liver function, stroke, bleeding, labile international normalized ratio [INR], elderly, drugs or alcohol) appears to have demonstrated the highest accuracy ⁶ ( Table E25-2 ).

TABLE E25-2

HAS-BLED Score

HAS-BLED Categories	Score
H ypertension (uncontrolled)	1
A bnormal renal and liver function (1 point each)	1 or 2
S troke	1
B leeding tendency or predisposition	1
L abile INRs (if on warfarin)	1
E lderly (age >65 y)	1
D rugs or alcohol (1 point each)	1 or 2
Maximum score	9

Hypertension (uncontrolled) is defined as a systolic blood pressure >160 mm Hg. Abnormal renal function is defined as the presence of chronic dialysis, renal transplantation, or a serum creatinine level ≥200 mmol/L. Abnormal liver function is defined as chronic hepatic disease (i.e., evidence of cirrhosis or biochemical evidence of significant hepatic derangement; bilirubin two times the upper limit of normal in association with aspartate aminotransferase/alanine aminotransferase/alkaline phosphatase three times the upper limit of normal). Labile INR is defined as having an INR within the therapeutic range less than 60% of the time. Drugs are concomitant antiplatelet or nonsteroidal antiinflammatory drugs. Alcohol is excess alcohol.

INR, International normalized ratio.

Assuming this man is prescribed a novel anticoagulant rather than warfarin and the aspirin is discontinued, his HAS-BLED score is 4. He scores points for uncontrolled hypertension, past history of stroke, age, and use of nonsteroidal anti-inflammatory drugs. This places him in the high-risk group, and his estimated annual risk of major bleeding is 8.7% ⁶ or 9.5%. ⁷ A weakness of this assessment is the wide confidence interval: in his case, from 4.9% to 19.6%.

Major bleeding is defined as intracranial bleeding, bleeding requiring hospitalization, a hemoglobin decrease of more than 2 g/dL, or the need for transfusion secondary to bleeding. In general, physicians are most concerned about the risk of fatal or intracranial bleeding, which constitutes a small proportion of major bleeding.

Unlike the CHA ₂ DS ₂ -VASc score, a number of the identified risk factors are modifiable. For instance, hypertension can be treated, INRs may be more closely monitored, nonsteroidal anti-inflammatory drugs may be withdrawn, and patients may reduce their alcohol intake if excessive.

Final Diagnosis

• Asymptomatic paroxysmal atrial fibrillation

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here