Mortality prediction using a modified Seattle Heart Failure Model may improve patient selection for ventricular tachycardia ablation

Kairav P. Vakil, Henri Roukoz, Roderick Tung, Wayne C. Levy, Inder S. Anand, Kalyanam Shivkumar, Thomas S. Rector, Marmar Vaseghi, Venkatakrishna Tholakanahalli

Research output: Contribution to journalArticle

10 Scopus citations

Abstract

Background Catheter ablation is frequently used as a palliative option to reduce shock burden in patients with ventricular tachycardia (VT). A risk prediction tool that accurately predicts short-term survival could improve patient selection for VT ablation. Objective The objective of the study is to assess utility of the Seattle Heart Failure Model (SHFM) to predict 6-month mortality in patients undergoing VT ablation. Methods Data on patients who underwent VT ablation at 2 tertiary institutions were retrospectively compiled. The SHFM score at the time of ablation, including 2 added VT variables, was used to predict 6-month mortality. The predicted number of deaths was compared to the observed number to assess model calibration. Model discrimination of those who died within 6 months was assessed by both K- and C-statistics. Results Mean age of the 243 patients was 63 ± 12 years; 89% were male. Mean SHFM score for the cohort was 1.3 ± 1.3. The Kaplan-Meier probability of death within 6 months was 14% (34 patients). The number of deaths estimated by the SHFM at 6 months was 31 (13%) giving a predicted to observed ratio of 0.91 (95% CI 0.64-1.30). The K-statistic for 6-month mortality predictions was 0.77 (95% CI 0.73-0.81), whereas the C-statistic was 0.84 (95% CI 0.78-0.92). Patients with an SHFM score 4.0 had an estimated positive predictive value of 80% (95% CI 28%-99%) for dying within 6 months of VT ablation. Conclusion The SHFM was well calibrated to a sample of patients who underwent VT ablation and provided good discrimination of short-term deaths. This model could be useful as a prognostic tool to improve patient selection for VT ablation.

Original languageEnglish (US)
Pages (from-to)1099-1104
Number of pages6
JournalAmerican Heart Journal
Volume170
Issue number6
DOIs
StatePublished - Dec 1 2015

    Fingerprint

Cite this