Abstract 15161: Prediction of Decompensated Heart Failure Using an Implanted Cardiac Device Risk Analysis
Archana Ganapathy, Brian Li, Samantha Langley, Rebecca Clark, Anthony Dimarco, Rossella Barbagallo, Jason Dungu, Henry Oluwasefunmi Savage- Physiology (medical)
- Cardiology and Cardiovascular Medicine
Introduction: TriageHF is a Heart Failure Risk Status (HFRS) diagnostic tool available on Medtronic implantable cardiac devices which reports dynamic statuses of risk of Heart Failure related Hospitalization (HFH) within the next 30 day, as high, medium or low. We investigated the accuracy of TriageHF in predicting 30-day HFH for patients attending a tertiary heart failure clinic.
Methods: We retrospectively recorded HFRS between November 2021 and May 2023, from patients implanted with Medtronic Cardiac Resynchronization Therapy (CRT) or Internal Cardioverter Defibrillator (ICD) devices. We then correlated HFRS with HHF within 30 days of these alerts. Two thresholds were tested: T1 where High & Medium HFRS were considered a positive alert and Low HFRS a negative alert and T2 where High HFRS was considered a positive alert and Medium & Low HFRS a negative alert. The sensitivity and specificity of each threshold in predicting 30-day HFH was tested in a contingency table.
Results: Over 18-months, we recorded 135 consecutive alerts from 32 patients. Median age was 72 years, 25% were female and 56% had an ischaemic substrate. 36, 88 and 11 alerts were high, medium, and low risk respectively. 0% of low HFRS and 19% of the high HFRS resulted in a 30-day HHF. 10 high HFRS alerts led to clinical interventions which was associated with a higher rate of HHF, compared to 26 high HFRS alerts where no contact was made (HHF 40% vs 11% p=0.05). The sensitivity and specificity of TriageHF in predicting 30-day HFH was 46.7% and 76.03% respectively for T1 and 100% and 9.1% respectively for T2.
Conclusions: This real-world analysis suggests that TriageHF risk prediction using T2 correctly predicted HHF at 30 days with the caveat of a high false positive rate. The potential exists for viable clinical use, however research into the amalgamation of clinical data may improve it and consequently lead to better utilisation of limited clinical resources.