Subcutaneous sensors for monitoring congestion and to reduce heart failure hospitalizations—a viable middle ground between deep implantable intravascular monitoring devices and wearable technologies?

Figures

• 

  Fig. 1. General groups of monitoring options already in place used for therapy cardiac implantable devices (top row), deep implantable intravascular devices (second row, the CardioMEMS sensor) measuring hemodynamic parameters such as cardiac filling pressures, subcutaneous sensors (third row) as less invasive option compared to CIEDs but with similar multi-parametric sensing option, and wearable solutions such as the ReDS (bottom)

• 

  Fig. 2. Average trend data for 52 events and 622 non-events offered by CIED and add-on HeartLogic by Boston Scientific (data reproduced from [45]). Note the similarity of those trends with the deep implantable sensor trend in Fig. 2. Particularly impedance has been effective in capturing fluid accumulation using CIEDs

• 

  Fig. 3. Deep implantable intravascular sensors can record filling pressures directly and have shown benefit in reducing heart failure hospitalization as for instance for CardioMEMS (left, Champion trial results reproduced from [12], each icon represents ten participants, coloured icons represent approximated numbers in multiples of ten) while early intravascular measurements of filling pressure via catheters connected to subcutaneous cans enabled long-term recordings of worsening before events (right, reproduced for recordings taken with Chronicle sensor from [22]). Note that both event and non-event groups ranged well above established threshold with event patients trending up rather marginally in pressure by less than 3 mmHg within 2 weeks prior the event

• 

  Fig. 4. Example cardiac cycle data captured and extracted feature trend data from subcutaneous device by FUTURE CARDIA for three patients and four events (left of blue vertical line, presented by M. Fudim at THT, Boston, 2025). HR, heart rate; LVET, left ventricular ejection period; PEP, pre-ejection period; SCG, Seismocardiography. An example trend of how similar features can be summarised in a single number risk score is shown on the right for the LinQ device extracted for 144 patients and 38 events (blue) and for non-event patients (grey) (reproduced from [35])