ABSTRACT Background: The electrocardiogram (ECG) can aid in identification of chronic kidney disease (CKD) patients at high risk for cardiovascular diseases. Cohort studies describe ECG abnormalities in patients on hemodialysis (HD), but we did not find data comparing ECG abnormalities among patients with normal kidney function or peritoneal dialysis (PD) to those on hemodialysis. We hypothesized that ECG conduction abnormalities would be more common, and cardiac conduction interval times longer, among patients on hemodialysis vs. those on peritoneal dialysis and CKD 1 or 2. Methods: Retrospective review of adult inpatients’ charts, comparing those with billing codes for “Hemodialysis” vs. inpatients without those charges, and an outpatient peritoneal dialysis cohort. Patients with CKD 3 or 4 were excluded. Results: One hundred and sixty-seven charts were reviewed. ECG conduction intervals were consistently and statistically longer among hemodialysis patients (n=88) vs. peritoneal dialysis (n=22) and CKD stage 1 and 2 (n=57): PR (175±35 vs 160±44 vs 157±22 msec) (p=0.009), QRS (115±32 vs. 111±31 vs 91±18 msec) (p=0.001), QT (411±71 vs. 403±46 vs 374±55 msec) (p=0.006), QTc (487±49 vs. 464±38 vs 452±52 msec) (p=0.0001). The only significantly different conduction abnormality was prevalence of left bundle branch block: 13.6% among HD patients, 5% in PD, and 2% in CKD 1 and 2 (p=0.03). Conclusion: To our knowledge, this is the first study to report that ECG conduction intervals are significantly longer as one progresses from CKD Stage 1 and 2, to PD, to HD. These and other data support the need for future research to utilize ECG conduction times to identify dialysis patients who could potentially benefit from proactive cardiac evaluations and risk reduction.
Automatic Detection of ECG Abnormalities by Using an Ensemble of Deep Residual Networks with Attention
