Timing and Location of Sudden Cardiac Arrests in Hemodialysis Patients

Since its widespread adoption as the primary modality for KRT more than 50 years ago, hemodialysis has been delivered as an intermittent three-times-a-week therapy for patients with ESKD. This pattern was born out of practical considerations for scheduling and convenience for dialysis organizations and patients. The potential for adverse events tied to the nonphysiologic nature of thrice weekly intermittent dialysis have been recognized for decades,1 but it was seminal observations by Bleyer et al.2 in 1999 that first reported the heightened risks of sudden cardiac arrest (SCA) associated with hemodialysis treatment days. Since that report, multiple studies have confirmed the timing of SCA events in intermittent hemodialysis patients with greater granularity. SCA events cluster around the timing of dialysis treatment, particularly on the first dialysis day of the week after the 2-day dialysis-free weekend.3 Recognition from registries and clinical trials that sudden death is the leading cause of death in this population has further increased awareness of SCA risks and its relationship with dialysis treatment. Despite the sizable body of evidence tying SCA events to the timing of hemodialysis treatments, a shortcoming of previous investigations is that they primarily relied on administrative data or adverse event reporting data from dialysis clinics, neither of which were purpose built to catalog SCA events. Thus, these studies could be prone to misclassification. Furthermore, comprehensive data on the exact location and circumstances of SCA events are difficult to determine using administrative data alone. Cardiac monitoring devices have been used by a few prospective studies to collect detailed information on arrhythmia events in relation to dialysis treatments,4 and these studies confirm the timing of significant arrhythmias to the peridialytic and intradialytic time frames, but given the relatively small number of subjects monitored for short time periods, few SCA were captured. Thus, it is of interest to confirm the timing and circumstances of SCA using a study platform specifically designed to study SCAs. In addition, while prior studies using dialysis clinic-reported event data have uncovered much about dialysis-related risk factors (for example, predialysis laboratory data and dialysate composition5), less is known about the timing and characteristics of SCA events occurring outside of dialysis facilities. The present study in this issue of Kidney360 aims to fill in these knowledge gaps.6 Using data from the Oregon Sudden Unexpected Death Study, a population-based prospective study collecting SCA events attended by emergency medical services in the Portland metropolitan area that used a rigorously adjudicated SCA definition to capture events, the authors identified SCA victims maintained with chronic hemodialysis. Access to records from emergency medical services and emergency departments of participating local hospitals allowed for further reporting on the circumstances of SCA and outcomes. The timing of SCA in relation to the timing of dialysis treatment was determined according to review of medical records from emergency medical services and emergency departments; unfortunately, records from dialysis facilities were not available in this study. Details on resuscitation management, presenting arrhythmias, laboratory data, and outcomes were extracted from hospital records. In summary, during the 18-year study period (2002–2020), 9765 total SCAs were captured, from which 210 individuals were identified as chronic hemodialysis patients. The weekly hemodialysis schedule was able to be established for 176 of these participants. SCA occurred within dialysis clinics in 23% of all cases, with 82% of these occurring during dialysis treatment and 18% occurring within an hour after treatment in clinic waiting rooms. The proportion of events occurring during dialysis or shortly afterward was three-fold higher than would be expected by chance alone, confirming previous observations identifying the high-risk association with dialysis treatment delivery. Furthermore, more SCA events occurred on the first hemodialysis treatment day of the week (Monday or Tuesday) than any other day of the week. Among other participants who experienced SCA distant from dialysis treatment, most of these occurred at home (62%) although a sizable proportion (29%) had SCA in another health care facility (i.e., nursing facility). The study found that patients who had an intradialytic/immediate postdialytic SCA had higher rates of receiving bystander cardiopulmonary resuscitation (CPR; 72% versus 38%) and had better resuscitation outcomes compared with non-intradialytic cardiac arrest (30% versus 5% survival to hospital discharge). This is not surprising because nearly all of these events occurred in dialysis clinics where trained medical personnel were present. However, it is concerning that 27% of these patients did not receive bystander CPR; this confirms previous findings reporting similar deficiencies in CPR performance within dialysis clinics7 and underscores the need to improve monitoring and readiness to provide CPR among dialysis clinic staff. The considerable number of SCA occurring after dialysis in clinic waiting rooms suggests a need for continued vigilance and monitoring of patients, even after patients have completed treatment. Finally, the study compared risk factors between patients who had an intradialytic/immediate postdialytic SCA with other patients. There were no significant differences except that patients with intradialytic/immediate postdialytic cardiac arrest had higher serum sodium and bicarbonate levels and lower potassium levels at emergency department presentation compared with patients arresting at other times. It is hard to make too much of this finding as specific risk factors for an intradialytic cardiac arrest, given the nonuniform availability of lab data between the two comparison groups and the fact that these laboratory differences would be expected when comparing lab values between any postdialysis and predialysis treatment group. Nevertheless, it underscores the findings of other studies suggesting that rapid intradialytic electrolyte changes, particularly intradialytic hypokalemia provoked by large serum-dialysate potassium gradients, may be a potent and modifiable risk factors for cardiac arrest.8 In sum, this study firmly reinforces the large body of evidence we already had highlighting the striking association between hemodialysis treatment and the risk of SCA. As long as thrice-weekly treatment remains the standard of care, we should expect that hemodialysis treatments will provoke cardiac arrest and that they will occur in our dialysis clinics. Every dialysis clinic should be prepared to respond to SCA with immediate and effective CPR and consider ways to monitor patients even after dialysis treatment is completed. Dialysis prescribers should pay close attention to minimizing risks based on our current understanding of dialytic risk factors until further interventions are developed that will minimize the unphysiological nature of thrice-weekly hemodialysis.