HeartLogic alerts were associated with medication changes (OR 1.87), and timely interventions during alerts lowered the 30-day risk of heart failure hospitalization (HR 0.32; 95% CI 0.10-0.99; p<0.05).
Cohort (n=1,799)
Blinded to HeartLogic-data
Does timely clinical intervention during an active HeartLogic alert reduce the 30-day risk of HF hospitalization in ICD/CRT patients?
Timely medication adjustments during HeartLogic alert periods significantly reduce the 30-day risk of heart failure hospitalization in ICD/CRT patients.
Effect estimate: HR 0.32 (95% CI 0.10-0.99)
p-value: p=<0.05
Abstract Background Previous studies have linked the HeartLogic index to heart failure (HF) decompensation, but have primarily focused on HF-related hospital admissions as the sole outcome, overlooking the impact of earlier outpatient management or medication changes. In Denmark, general practitioners (GPs) commonly serve as first-line contact and may initiate interventions before hospitalization. We aimed to investigate whether HeartLogic alerts in patients with ICD/CRT are associated with clinical actions during the active alert period, including medication adjustments, and subsequent HF hospitalizations. Methods We retrospectively analyzed 1,799 ICD/CRT patients (blinded to HeartLogic-data) contributing 4,290 patient-years from 2018-2024. Clinicians had no access to daily HeartLogic-index or other telemonitoring parameters during routine care. Interventions were therefore most likely driven by patient-reported symptoms or routine clinical evaluations, rather than HeartLogic-data. Medication changes were identified using prescription registries and estimated by the reverse waiting time distribution method, intervention was flagged if there was an initiation of a new HF medication, discontinuation of HF medication, or a ≥50% dose change relative to the most recent prescription. We then linked HeartLogic alerts with healthcare records (GP consultations, prescription refills) to identify medication changes or clinic contacts. Using logistic regression, we assessed the association between HeartLogic alerts and interventions. Cox proportional hazards models were applied to evaluate the 30-day risk of HF hospitalization following HeartLogic alerts, comparing those who received intervention during the active alert period with those who did not. Results HeartLogic alerts: During 4,290 patient-years, there were 3,829 alerts (~0.9 alerts/patient-year). Of 1,799 patients, 980 (54.5%) experienced at least one HeartLogic alert. Clinical actions: HeartLogic alerts were significantly associated with medication adjustments or new HF prescriptions (OR 1.87, 95% CI 1.38–2.48; p0.001). HF hospitalization: Patients who received timely interventions in response to an HeartLogic alert had a lower risk of HF hospitalization within 30 days (HR 0.32, 95% CI 0.10–0.99; p0.05) compared to those without interventions. Subgroup analyses: The predictive value of HeartLogic held across ICD vs. CRT, baseline LVEF, NYHA class, age, and sex (p0.05 for all interactions). Conclusion In this retrospective analysis of Danish ICD/CRT patients, HeartLogic alerts were strongly associated with outpatient medication changes, and lack of timely intervention was associated with a higher risk of HF hospitalization. Our findings emphasize that real-time access to HeartLogic data may allow for even earlier actions and further improve HF outcomes. Prospective, real-time evaluation in the DanLogic study will help clarify the full potential of HeartLogic monitoring in everyday practice.
Davodian et al. (Sat,) conducted a cohort in Heart failure (n=1,799). Timely clinical interventions during active HeartLogic alert vs. No intervention during active alert was evaluated on Heart failure hospitalization within 30 days (HR 0.32, 95% CI 0.10-0.99, p=<0.05). HeartLogic alerts were associated with medication changes (OR 1.87), and timely interventions during alerts lowered the 30-day risk of heart failure hospitalization (HR 0.32; 95% CI 0.10-0.99; p<0.05).