Abstract 12281: Impedance is an Unreliable Surrogate for Left Atrial Pressure in Ambulatory Patients with Heart Failure
Objectives: With the availability of CRT-D systems capable of measuring intrathoracic impedance (Z) from multiple vectors there was promise for deriving a reliable surrogate for left atrial pressure (LAP) in ambulatory patients with severe heart failure (HF). This study quantifies the long term relationship between Z and LAP including the effects of changing volume status, posture and non-hemodynamic factors on Z and evaluates the hypothesis that reliable estimates of LAP may be derived from Z.
Methods: Patients with NYHA III symptoms and prior decompensated HF (N=17, age 64±10, EF 29±10) were implanted with a LAP monitor and a CRT-D capable of recording Z from 6 vectors. LAP was monitored 3x daily for 27±10 months. Vector based Z measurements were linearly transformed into electrode based Z. A 3D accelerometer tracked posture-dependent events in 5 cases. LAP and Z were compared by linear regression.
Results: LAP vs. Z correlated poorly (R2=0.54±0.17; range: 0.25 to 0.78) for best overall vector/electrode chosen. At best, Z from LVring, RVring and Can electrodes had R2>0.5 in only 50% of cases. Patients with rapidly changing LAP due to mitral regurgitation or ischemia (n=5) had poorer correlations compared to with volume-related changes in LAP (figure). Generator change, lead dislodgement and revision (n=1), artifacts and electronic noise affected Z independent of LAP. All patients had complex time-dependent postural changes in Z with a rapid component affected by local tissue contact and slower changes by fluid redistribution.
Conclusions: Z, whether based on vectors or individual electrode estimates, correlated weakly with LAP. Z was primarily influenced by local tissue contact and redistribution of fluid. Z more closely followed LAP when changes were due to large volume overload or diuresis. Overall, Z is an unreliable surrogate for long term estimation of LAP in ambulatory HF patients because it is substantially influenced by confounding non-hemodynamic factors.
- © 2012 by American Heart Association, Inc.