Abstract: SA-PO897
Performance of Stroke Risk Scores in Dialysis Patients
Session Information
- Dialysis: Cardiovascular, BP, Volume
October 27, 2018 | Location: Exhibit Hall, San Diego Convention Center
Abstract Time: 10:00 AM - 12:00 PM
Category: Dialysis
- 701 Dialysis: Hemodialysis and Frequent Dialysis
Authors
- Ocak, Gurbey, University Medical Center Utrecht, Utrecht, Netherlands
- Rookmaaker, Maarten B., University Medical Center Utrecht, Utrecht, Netherlands
- Blankestijn, Peter J., University Medical Center Utrecht, Utrecht, Netherlands
- Verhaar, Marianne C., University Medical Center Utrecht, Utrecht, Netherlands
- Dekker, Friedo W., Leiden University Medical Center, Leiden, Netherlands
- Van diepen, Merel, Leiden University Medical Center, Leiden, Netherlands
Background
Dialysis patients have an increased ischemic stroke risk. Stroke risk scores, including the CHADS2, ATRIA and CHA2DS2VASC, have been developed to identify patients with an increased stroke risk in patients with atrial fibrillation allowing for personalization of vitamin K antagonist prescription. In the original articles, CHADS2 (C-statistic 0.82), ATRIA (C-statistic 0.73) and CHA2DS2VASC (C-statistic 0.67) had reasonable predictive abilities. However, the predictive performances of these stroke risk scores have not been validated in dialysis patients. Therefore, the aim of this study was to validate existing stroke risk scores in dialysis patients.
Methods
A total of 755 incident dialysis patients from the NECOSAD study were prospectively followed for validated ischemic stroke within five years of dialysis. Hazard ratios with 95% confidence intervals (CIs) were calculated using Cox proportional hazards analyses for high and intermediate risk scores as compared with low risk scores for the CHADS2 (low risk= 0, intermediate risk=1-2 and high risk= ≥3), ATRIA (low risk= 0-5, intermediate risk=6 and high risk=≥7) and the CHA2DS2VASC (low risk= 0, intermediate risk=1 and high risk= ≥2). Furthermore, we evaluated the discriminative performance of these bleeding risk groups by calculating Harrell’s C-statistics.
Results
During a median follow-up of 2.0 years (interquartile range 0.9-3.7), 58 (7.7% of all patients) first ischemic stroke events occurred. Of the 755 patients, 14% were classified as high risk by the CHADS2, 28% by the ATRIA and 72% by the CHA2DS2VASC. A high risk score as compared with a low risk score was associated with a 8.2-fold (95% CI 1.1-63.1) increased ischemic stroke risk for the CHADS2 and a 2.2-fold (95% CI 1.3-3.9) increased ischemic stroke risk for the ATRIA. A high CHA2DS2VASC risk score as compared with an intermediate risk score was associated with a 8.7-fold (2.1-35.7) increased ischemic stroke (no stroke events in low risk group). The C-statistics were 0.59 for the CHADS2, 0.61 for the ATRIA and 0.62 for the CHA2DS2VASC.
Conclusion
In this prospective cohort with validated data on ischemic stroke, we showed that the CHADS2, ATRIA and CHA2DS2VASC had poor predictive abilities. Therefore, these stroke risk scores may not be useful for guiding individual decision-making in dialysis patients.