Abstract: TH-PO204
Phenotyping of Heart Failure (HF) Risk in CKD Using 12-Lead ECG
Session Information
- Hypertension and CVD: Clinical - I
November 02, 2023 | Location: Exhibit Hall, Pennsylvania Convention Center
Abstract Time: 10:00 AM - 12:00 PM
Category: Hypertension and CVD
- 1602 Hypertension and CVD: Clinical
Authors
- Soomro, Qandeel H., New York University Grossman School of Medicine, New York, New York, United States
- Divers, Jasmin, New York University, New York, New York, United States
- Charytan, David M., New York University Grossman School of Medicine, New York, New York, United States
Background
H in CKD is associated with high rates of mortality and recurrent hospitalizations. Early identification of individuals at increased risk for HF in CKD would facilitate improved risk stratification and modification of treatment strategies. Current tools to predict de novo HF in CKD are lacking. We aimed to analyze associations of standard ECG features and the risk of incident HF in CKD.
Methods
We assembled a dataset of clinical and ECG data on individuals at NYU with at least two eGFR measurements <60 mL/min/1.73m2 ≥90 days apart and an ECGs done at NYU between 2012-2021. The index ECG was the first ECG after an eGFR measurement. We excluded individuals with existing history of HF or HF admission within 30 days of index ECG. HF event was defined as primary inpatient discharge diagnosis identified using validated ICD codes. Univariate associations were analyzed using KM curves, multivariable associations of ECG features and incidence HF event were analyzed using Cox proportional hazards adjusting for relevant covariates.
Results
Among 14,503 individuals included, HF event rate was 6.2% with median time to event of 672 days (340,1625). eGFR stages and sex were similar among those with and without HF. Standard ECG features such as PR, QRS, QT, QTc, P and T axis as well as HRV parameters derived from ECG such as SDNN, RMSSD, PRR50% differed significantly between the two groups on univariate analyses (Figure). The concentrations of potassium, albumin, and calcium were significantly different in those with and without HF. In the best fit model standard ECG but not HRV parameters were independent predictors of HF (Table).
Conclusion
Standard ECG features when combined with demographic and comorbidities data can predict incident HF in CKD.
Cox PH model
HF survival probability in the CKD cohort with and without abnormal HRV
Funding
- Commercial Support – ASN Donal E. Wesson Fellowship grant