Abstract
BACKGROUND AND OBJECTIVES: Data collected by the US Renal Data System (USRDS) identify sudden cardiac death (SCD) as the leading cause of death among hemodialysis patients. However, evidence suggests that clinical events captured on the USRDS death notification form may be inaccurate. A new method for classifying SCD was recently developed to enhance the accuracy of SCD classification. This study examined the performance characteristics of this refined definition using a cohort of hemodialysis patients who experienced a witnessed SCD as the reference standard. DESIGN, SETTING, PARTICIPANTS, & MEASUREMENTS: This is a retrospective cohort study of 363 patients who experienced a witnessed SCD in US Gambro (DaVita) outpatient dialysis clinics. Sensitivity of SCD defined by death notification forms and SCD defined using additional administrative sources was compared. Clinical data recorded near time of death were also examined. RESULTS: Existing USRDS death notification forms reported 70.8% of witnessed SCD as "cardiac arrest/cause unknown" or "arrhythmia." The refined definition significantly improved identification to 83.8% of witnessed SCD events (P<0.001). Verified SCD cases that were not identified by either definition were more likely to be reported on the death notification form as death due to myocardial infarction, hyperkalemia, sepsis, malignancy, or unknown cause. CONCLUSIONS: Compared with the death notification form alone, the refined SCD definition significantly improves the sensitivity of reporting of witnessed SCD occurring within outpatient hemodialysis clinics. More accurate reporting of cardiac events by clinicians and refinements to existing death notification forms may further improve recognition and understanding of SCD.