Retention

Heart failure remains one of the leading causes of hospitalisations and mortality worldwide. Despite advances in treatment, patients frequently experience clinical deterioration that leads to emergency visits, repeated hospitalisations and reduced quality of life.

Traditional care models are reactive, relying heavily on hospital-based monitoring. RETENTION addresses this gap by shifting toward proactive, personalised, data-driven patient management outside hospital settings.

RETENTION is based on the hypothesis that continuous remote monitoring can improve heart failure management and that advanced data analytics can enable early detection of deterioration while generating personalised recommendations. The project deploys an integrated digital health ecosystem that collects clinical and real-world data through certified wearable and home-based devices, alongside patient-reported outcomes.

All data are securely transferred to a centralised platform where they are visualised for clinicians and processed using backend machine learning and AI algorithms. By identifying patterns and early warning signs of decompensation, the system supports risk prediction and produces personalised suggestions, shifting heart failure care from a hospital-based model to a data-driven, preventive approach.

• Multicentre Randomised Controlled Trial (RCT)
• 6 hospitals
• 450 patients aged 18-75
• 3 patient subgroups:
  • Heart Failure (HF)
  • Left Ventricular Assist Devices (LVAD)
  • Heart Transplant (HT)
• Two study arms:
  • Data-driven interventions and remote monitoring visualisations
  • Standard clinical care
• Trial duration: 28 months – whit a median follow-up time of 18 months

The project also addresses the historical underrepresentation of women in heart failure trials, ensuring balanced recruitment.

The RETENTION platform collects daily measurements of three type of data, which are:

• Objective clinical information, consisting of vital signs (heart rate, blood pressure measurements, peripheral capillary oxygen saturation, weight, body temperature), and measurement of the device settings in LVAD users.
• Subjective clinical information, referred by QoL and symptoms related questions (presence of leg oedema, shortness of breath, dizziness, etc), as well as medications adherence.
• Real-world data, derived from non-medical sources such as physical activity, quality of sleep and environmental factors in the living environment of the patients (ambient temperature, relative humidity, and pollutants).

Clinical information is also entered manually via a clinicians’ dashboard during scheduled follow-up visits every 4 month, or unscheduled patient visits.

Advanced analytics integrate these multidimensional datasets to detect early signs of clinical deterioration and support personalized interventions.

RETENTION aims to reduce all-cause and cardiovascular mortality, decrease heart failure hospitalisations and emergency visits and improve patients’ quality of life through continuous remote monitoring and data-driven interventions. The project also investigates the added value of integrating environmental and real-world data into heart failure management to enable more personalised and effective care.

Through this integrated, AI-enabled approach, RETENTION seeks to redefine heart failure care by transforming continuous real-world monitoring into clinical insight and patient benefit.