Evaluation of sudden cardiac arrest and sudden cardiac death in patients on dialysis

OVERVIEW — Patients on dialysis are at extraordinarily high risk for death [1-3]. In 2019, the estimated mortality rate of United States patients on dialysis was 173 deaths per 1000 patient-years, a figure that subsequently increased in 2020 in association with the coronavirus disease 2019 (COVID-19) pandemic [3].

Cardiovascular disease is the major cause of death, accounting for approximately 40 percent of all-cause mortality in patients on dialysis [3]. In the United States Renal Data System (USRDS) database, the single, largest, specific cause of death is attributed to arrhythmic mechanisms or sudden cardiac arrest (SCA) [3]. (See "Patient survival and maintenance dialysis".)

The epidemiology, clinical manifestations, and evaluation of SCA and sudden cardiac death (SCD) in the dialysis population are provided in this topic review. Detailed discussions of treatment and prevention of SCA and SCD are presented separately. (See "Overview of sudden cardiac arrest and sudden cardiac death".)

DEFINITION AND EPIDEMIOLOGY — In the general population, the term "sudden cardiac death" (SCD) is commonly used to describe SCA in the setting of heart disease (although some have structurally normal hearts) with cessation of cardiac function whether or not resuscitation or spontaneous reversion occurs.

Previously, the term "sudden cardiac death" has been used even if a patient was successfully resuscitated. Such cases have been referred to as "aborted SCD" or "resuscitated SCD," and patients who experienced such events were said to be "sudden death survivors." Clearer and more rational definitions of SCA and SCD were proposed in 2017 by the American College of Cardiology/American Heart Association/Heart Rhythm Society (ACC/AHA/HRS) [4] (see "Overview of sudden cardiac arrest and sudden cardiac death"):

●SCA is the sudden cessation of cardiac activity such that the victim becomes unresponsive, with either persisting gasping respirations or absence of any respiratory movements, and no signs of circulation as manifest by the absence of a perceptible pulse. An arrest is presumed to be of cardiac etiology unless it is known or likely to have been caused by trauma, drowning, respiratory failure or asphyxia, electrocution, drug overdose, or any other noncardiac cause.

●SCD refers to sudden and unexpected death occurring within an hour of the onset of symptoms or occurring in patients found dead within 24 hours of being asymptomatic and presumably due to a cardiac arrhythmia or hemodynamic catastrophe.

Except where noted, we will use the terms "SCA" and "SCD" as defined in the 2017 ACC/AHA/HRS document. However, most of the available epidemiologic data were published prior to this standardization and are therefore based upon different standards. In addition, alternative definitions of sudden death, sudden cardiac death, and aborted cardiac arrest have been proposed for the dialysis population, given the number of unique competing risks in individuals with end-stage kidney disease.

In the United States Renal Data System (USRDS) database, the cause of death attributed to arrhythmic mechanisms is noted in the Centers for Medicare and Medicaid Services (CMS) death notification form 2746 by either "cardiac arrest/cause unknown" or arrhythmia. Based upon this definition, arrhythmias may therefore be responsible for [3]:

●Eighty percent of all cardiovascular deaths or 33 percent of all-cause mortality in patients on peritoneal dialysis

●Eighty-three percent of all cardiovascular deaths or 33 percent of all-cause mortality in patients on hemodialysis

However, death attributed to arrhythmias that is based entirely upon the CMS death notification form has obvious limitations and may be inaccurate. Thus, in addition to this definition, the USRDS Cardiovascular Special Studies Center (CVSSC) used a more complex method in the 2006 USRDS annual data report, which incorporated cause of death in the context of death location (eg, a patient succumbing to myocardial infarction on an ambulance run would be identified as sustaining SCD) and excluded patients with deaths occurring in the setting of sepsis, malignancy, hyperkalemia, and, importantly, withdrawal from dialysis. Using this method, it has been estimated that 29.7 percent of deaths in patients on dialysis are related to SCD [5,6]. The USRDS CVSSC method, however, has not been applied to contemporary data.

We think that the overall best estimate is that SCD is responsible for 29 percent of all-cause mortality in patients on dialysis. Similar findings on the relative contribution (22 to 26 percent) of sudden death to all-cause mortality in patients on dialysis have been reported in multiple studies including HEMO, 4D trial, the Choices for Healthy Outcomes in Caring for End-Stage Renal Disease (CHOICE) cohort, and the Dialysis Outcomes and Practice Patterns Study (DOPPS) [7-10] and the Evaluation of Cinacalcet HCL Therapy to Lower Cardiovascular Events (EVOLVE) trial [11]. EVOLVE, which enrolled 3883 patients on hemodialysis and provided formal adjudication of cause-specific mortality, was the largest randomized, clinical trial ever performed in patients on hemodialysis. In EVOLVE, 25 percent of all deaths were adjudicated as "sudden deaths."

One noteworthy trend is the "disconnect" between the rates of SCD and cardiovascular death among patients on dialysis (figure 1). Cardiovascular death rates have declined with little change in SCD rates. For this reason, the attributable percentage of SCD to cardiovascular and all-cause mortality has increased over time. In 2020, 33 percent of all-cause mortality in patients on dialysis was attributed to arrhythmic mechanisms [3].

The rate of SCD in patients on dialysis is significantly higher than that observed in the general population and is roughly comparable to that of individuals not on dialysis who have a history of an adverse cardiovascular event. (See "Overview of sudden cardiac arrest and sudden cardiac death".)

There is also an enhanced risk of SCD in the first hemodialysis session of the week. Compared with the average risk of SCD, there is a 50 percent increased frequency of SCD on Monday (for patients dialyzing Monday, Wednesday, and Friday) and on Tuesday (for patients having hemodialysis Tuesday, Thursday, and Saturday) [12]. In addition, one study reported a threefold increased risk of sudden death in the 12 hours before the end of the long weekend interval and a 1.7-fold increased risk of SCD in the 12 hours starting with the dialysis procedure following this long interval [13].

The initiation of hemodialysis is a special period of heightened risk for both all-cause mortality and SCD. In 2015, the annualized death rate for United States patients on dialysis in the first 90 days after dialysis initiation was 312 deaths per 1000 patient-years [1]. As shown in the figures (figure 2 and figure 3), the highest rate of sudden death is in the first month after dialysis initiation, with at least 30 percent of all deaths being attributable to SCD.

MECHANISMS — In the general as well as the dialysis population, most SCA events are believed to be due to ventricular arrhythmias; that is, ventricular tachycardia (VT) or ventricular fibrillation (VF). A minority may have been attributed to bradyarrhythmias. (See "Pathophysiology and etiology of sudden cardiac arrest".)

However, several studies using insertable cardiac monitors (sometimes referred to as implantable loop recorders) have suggested that bradyarrhythmia may be a contributing mechanism to SCD [14-18]. For example, in one study of 66 patients on dialysis who had implantable loop recorders, bradycardia and asystole were the dominant events [17].

Although SCA can occur in patients with structurally normal hearts, most patients with SCA have some form of underlying heart disease. A triggering event or condition interacts with the underlying substrate to produce the fatal arrhythmia [19].

Although many triggers have been identified, acute myocardial ischemia is felt to be the most common initiating event in the general population [20]. As will be discussed, the end-stage kidney disease (ESKD) patient has some unique factors that can both alter the underlying substrate as well as trigger ventricular arrhythmic events.

RISK FACTORS AND CAUSES — In the general population, abnormalities of the coronary arteries, myocardium, and cardiac conduction system are the most common underlying causes of the life-threatening arrhythmias that result in SCA. (See "Pathophysiology and etiology of sudden cardiac arrest" and "Overview of screening and diagnosis of heart disease in patients on dialysis" and "Valvular heart disease in patients with end-stage kidney disease".)

Predictors of SCD among patients on hemodialysis were evaluated using data from the HEMO study [21]. Among 1745 enrolled patients on hemodialysis, 808 died over a median follow-up of 2.5 years, 22 percent of which were due to SCD. Age, diabetes, peripheral vascular disease, ischemic heart disease, a low serum creatinine (reflecting decreased muscle mass and poor nutrition), and an elevated alkaline phosphatase predicted a higher risk for SCD. Traditional cardiovascular risk factors such as smoking and cholesterol did not, but risks conferred by these factors may have been incorporated into the overall increased risk associated with ischemic heart disease. This study did not adjust for dialysis-related risk factors (such as potassium dialysate) nor for other known risk factors such as left ventricular (LV) hypertrophy. The major results of the HEMO study are discussed elsewhere. (See "Prescribing and assessing adequate hemodialysis", section on 'Target Kt/V'.)

The following is a brief overview of the more important abnormalities found in the dialysis population that may underlie their increased incidence of SCA:

●Obstructive coronary artery disease is likely an important contributor to SCA. However, data from the United States Renal Data System Cardiovascular Special Studies Center (USRDS CVSSC) have found an unexplained high mortality due to arrhythmic mechanisms after successful coronary revascularization, suggesting that other factors must be significant [22,23]. In the dialysis population, for example, there was a two-year mortality of 48 and 43 percent after nondrug eluting coronary artery stents and coronary artery bypass surgery (CABG) incorporating internal mammary graft use, respectively. The annual mortality attributed to arrhythmic mechanisms was 8.5 and 7 percent after stenting and CABG, respectively, which is higher than that observed in the general population. This implies that reliance solely upon ameliorating myocardial ischemia by coronary revascularization may be an inadequate clinical strategy for the prevention of SCD in patients on dialysis.

●There is a markedly increased incidence of myocardial abnormalities, such as LV hypertrophy (approximately 75 percent of patients on dialysis) and alterations in myocardial ultrastructure and function (including endothelial dysfunction, interstitial fibrosis, decreased perfusion reserve, and diminished ischemia tolerance). (See "Overview of screening and diagnosis of heart disease in patients on dialysis" and "Hypertension in patients on dialysis".)

●Rapid electrolyte shifts during hemodialysis sessions and the presence of hyperkalemia due to kidney failure increase the risk of arrhythmias. This enhanced risk is not surprising given the profound fluid and electrolyte derangements before hemodialysis and the physiologic demands of the hemodialysis session [24]. The nonphysiologic nature of conventional, thrice-weekly hemodialysis sessions may further increase the risk of SCA [12]. This is supported by the observational data previously mentioned showing an increased risk surrounding the first hemodialysis session of the week [25]. (See 'Definition and epidemiology' above.)

●The interplay between the type of kidney replacement therapy and dialysis vintage appears to have an impact upon the risk of SCA, with the relative hazard of cardiac arrest in hemodialysis compared with peritoneal dialysis varying with time after initiation of kidney replacement therapy. The rate of cardiac arrest is approximately 50 percent higher in patients on hemodialysis three months after dialysis initiation, but they are similar at two years. Three years after dialysis initiation, the rate of cardiac arrest is higher in patients on peritoneal dialysis (figure 4). There is a strikingly higher adjusted mortality rate in hemodialysis versus peritoneal dialysis early after initiation of kidney replacement therapy (figure 5).

Risk factors related to dialysis prescription — The following features related to the dialysis prescription or modality may be modifiable risk factors for SCA.

●Lower dialysate potassium and calcium – In a case-control study of 43,200 patients, low-potassium dialysate (<2 mEq/L) was an independent risk factor for SCA [26]. The increased risk associated with low-potassium dialysate was greatest at lower levels of predialysis serum potassium. Low-calcium dialysate concentrations were also linked to SCA in this study. Compared with dialysate potassium ≥3 mEq/L, dialysate potassium concentrations ≤1.5 and ≤2 to 2.5 mEq/L were associated with increased risk of SCD (hazard ratios [HRs] 1.39, 95% CI 1.12-1.74 and 1.17, 95% CI 1.01-1.37, respectively). The magnitude of the association of SCD with dialysis potassium ≤1.5 was greater among patients with serum potassium <5 mEq/L. The Dialysis Outcomes and Practice Patterns Study (DOPPS) including 55,183 patients did not demonstrate a significant difference in clinical outcome related to the use of dialysate potassium concentrations of 2 versus 3 mEq/L [27].

●Shorter treatment times – The effect of hemodialysis session length on cardiovascular mortality independent of conventional markers of dialysis adequacy is unclear. However, some observational studies of large databases suggest that increased session length improves survival [28-30]. (See "Patient survival and maintenance dialysis", section on 'Hemodialysis session length'.)

A discussion of the effects of prolonged dialysis session length on cardiovascular outcomes, including survival, with nocturnal HD is presented separately. (See "Outcomes associated with nocturnal hemodialysis".)

●Higher ultrafiltration rates – Higher ultrafiltration rates have been associated with increased cardiovascular mortality [10,31,32]. This is discussed in detail elsewhere. (See "Patient survival and maintenance dialysis", section on 'Rapid ultrafiltration' and "Prescribing and assessing adequate hemodialysis", section on 'Patient-specific parameters'.)

●Dialysate temperature – Dialysate cooling has been advocated to reduce intradialytic hypotension and myocardial stunning; however, the putative benefit on reduction of subsequent cardiac events (including sudden death) remains to be established. This is discussed in detail elsewhere. (See "Intradialytic hypotension in an otherwise stable patient", section on 'Second-line approach'.)

●Dialysis modality – No dialysis modality (eg, hemodialysis, peritoneal dialysis, or hemodiafiltration) has a proven superiority over another at reducing cardiovascular mortality among patients with end-stage kidney disease. (See "Dialysis modality and patient outcome", section on 'Survival by modality'.)

CLINICAL MANIFESTATIONS — In patients with and without end-stage kidney disease (ESKD), most individuals suffering from SCA become unconscious within seconds to minutes as a result of insufficient cerebral blood flow. There are usually no premonitory symptoms. Symptoms, if present, are nonspecific and include chest discomfort, palpitations, shortness of breath, and weakness. (See "Overview of sudden cardiac arrest and sudden cardiac death".)

EVALUATION

Survivor of sudden cardiac arrest — In the general population, the evaluation of the survivor of SCA includes the following (see "Cardiac evaluation of the survivor of sudden cardiac arrest"):

●Identification and treatment of acute reversible causes

●Evaluation for structural heart disease

●In patients without obvious arrhythmic triggers or cardiac structural abnormalities, an evaluation for primary electrical diseases

●Neurologic and psychologic assessment

●In selected patients with a suspected or confirmed heritable syndrome, evaluation of family members

The evaluation in the patient on dialysis who survives SCA is generally the same as that in the patient without kidney failure. However, close attention should be paid to the presence of myocardial dysfunction and/or ischemia (since they are so common), the possibility of improper medication dosing in the patient with kidney failure, and the circumstances associated with the event, particularly if it occurred during and/or surrounding a hemodialysis session.

As examples:

●Since both myocardial ischemia and/or dysfunction are relatively common in the patient on dialysis, their presence, either alone or in combination, may markedly enhance the risk of SCA. However, retrospective analysis has shown that 71 percent of patients on dialysis who experienced SCD had either normal or only mild to moderate left ventricular (LV) dysfunction, suggesting that other factors may underlie SCA in many patients on dialysis [33]. (See 'Identification of the high-risk patient on dialysis' below.)

●Increased electrical instability resulting in SCA may have been due to fluid shifts, autonomic imbalance/increased sympathetic activity (including sleep apnea), acid/base disturbances, and electrolyte abnormalities [26,34-40]. An increased risk of cardiac arrest may be particularly associated with a low-potassium concentration in the dialysate. (See "Acute complications during hemodialysis".)

●Improper dosing of certain medications may predispose the patient with kidney failure to brady/tachyarrhythmias and/or proarrhythmic effects, thereby causing SCA. (See "Overview of sudden cardiac arrest and sudden cardiac death".)

One study reported that low predialysis serum potassium (<4.3 mEq/L) was associated with an increased mortality hazard in patients on hemodialysis receiving digoxin, suggesting that even low "normal" potassium levels may enhance the proarrhythmic risk of digoxin [41]. In this study, the mortality risk associated with digoxin was attenuated in patients on dialysis with serum potassium >4.6 mEq/L. Digoxin should be used with extreme caution in such patients.

Identification of the high-risk patient on dialysis — The approach to screening and diagnosis of cardiac disease in patients on dialysis is discussed elsewhere. (See "Overview of screening and diagnosis of heart disease in patients on dialysis".)

It is uncertain how best to identify patients on dialysis at highest risk of SCA in whom prophylactic interventions may be beneficial.

In the general population, it is known that reduced LV dysfunction is the strongest predictor of SCA. Clinically, the presence of heart failure also identifies patients who are at high risk of SCA, perhaps by additional arrhythmogenic factors, such as activation of the neurohormonal cascade and electrolyte shifts. In the nondialysis population, most primary prevention implantable cardioverter-defibrillator (ICD) trials have shown significant improvement in survival in the following high-risk groups who received ICDs:

●The earlier primary prevention trials included patients who had decreased systolic function, prior myocardial infarction, nonsustained ventricular tachycardia (VT), and positive electrophysiology (EP) study for induction of VT.

●In later trials, enrollment did not require nonsustained VT or positive EP studies. Instead, enrollment was based upon decreased systolic function and/or heart failure. In addition, nonischemic cardiomyopathy patients were included.

Thus, in the nondialysis population, primary prevention ICD trials have principally demonstrated survival improvement in groups with decreased systolic function who receive ICDs. (See "Primary prevention of sudden cardiac death in patients with cardiomyopathy and heart failure with reduced LVEF".)

However, it is unclear if patients on dialysis with decreased systolic function also receive survival benefits with ICD. No trials have examined this issue in patients with ESKD, and observational studies have not reported a survival benefit of primary prevention ICDs among patients on dialysis [42,43]. Furthermore, higher complication rates of transvenous devices, including infection, lead dislodgements, high defibrillation thresholds, and increased in-hospital mortality have been reported [42,44,45]. Whether subcutaneous ICDs eliminate the issues surrounding transvenous leads has not been determined.

In addition, despite the very high annual mortality from SCD, the patient on dialysis with reduced LV function is not typical of the general dialysis population. Fifteen percent or fewer patients on dialysis have significantly decreased LV function [46-49]. In the largest study to date, among 1254 consecutive patients starting hemodialysis in Japan, 5 percent had LV ejection fractions of less than 40 percent [48].

Thus, other unique factors/circumstances may contribute to the general increased risk of SCA in end-stage kidney disease (ESKD) patients in dialysis. Possibilities include the following:

●There is additional substrate (myocardial) modification, such as interstitial fibrosis due to chronic uremia, microvascular disease, or endothelial dysfunction; increased calcium/phosphate deposition; and significant LV hypertrophy due to hypertension and/or anemia [50-55].

●Increased electrical instability may be present due to fluid shifts, autonomic imbalance/increased sympathetic activity (including sleep apnea), inflammatory state, acid/base disturbances, and/or electrolyte abnormalities [26,34-40,56].

It is likely that the increased risk of ventricular arrhythmias in ESKD patients is due to a combination of these many interacting factors (table 1).

Finally, while an ejection fraction of <35 to 40 percent is considered the major risk factor for SCA in patients not on dialysis, it is possible that more mild LV dysfunction imparts a greater risk of cardiovascular events in the dialysis population, regardless of the etiology and despite optimal management. As an example, one study found that the best predictor of SCD risk in patients on peritoneal dialysis was an LV ejection fraction of ≤48 percent [57]. However, a randomized trial of dual-chamber ICD implantation in 188 patients on dialysis with ejection fraction >35 percent showed no difference in the rate of SCD or all-cause mortality [58].

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Dialysis".)

SUMMARY AND RECOMMENDATIONS

●Epidemiology – Cardiac disease is the major cause of death among patients on dialysis. In the United States Renal Data System (USRDS) database, the single, largest, specific cause of death is attributed to arrhythmic mechanisms or sudden cardiac death (SCD). The overall best estimate is that SCD is responsible for approximately 29 percent of all-cause mortality in patients on dialysis. (See 'Definition and epidemiology' above.)

●Risk factors – Many causes of and risk factors for sudden cardiac arrest (SCA) are shared among patients with and without end-stage kidney disease (ESKD), although patients on dialysis frequently have a relatively increased incidence of abnormalities of the coronary arteries, myocardium, and cardiac conduction system. There are also issues unique to patients on dialysis. Rapid electrolyte shifts during hemodialysis sessions increase the risk of arrhythmias. Variables in the hemodialysis prescription such as the use of low-potassium dialysate, low-calcium dialysate, or large ultrafiltration volumes and/or rates may be modifiable risk factors for SCA. (See 'Risk factors and causes' above.)

●Evaluation of survivors – The evaluation in the patient on dialysis who survives SCA is generally the same as that in the patient without kidney failure. However, close attention should be paid to the presence of myocardial dysfunction and/or ischemia, the possibility of improper medication dosing in the patient with kidney failure, and the circumstances associated with the event, particularly if it occurred during and/or surrounding a hemodialysis session. (See "Cardiac evaluation of the survivor of sudden cardiac arrest" and 'Evaluation' above.)