Background: Previous studies have demonstrated that intrathoracic impedance monitoring (IIM) is associated with fluid overload. However, it remains unclear whether this new technology can predict heart failure (HF) before deterioration. Whether fluid status based on IIM predicts HF in patients with left ventricular (LV) systolic dysfunction was investigated.
Methods and results: A prospective clinical observational study of 123 patients implanted with IIM-capable cardiac devices was carried out. The primary endpoint was the positive predictive value (PPV) at 12 months. Secondary endpoints were a correlation between onset of HF and IIM, optimal threshold of fluid index and duration between the alert and HF. Complete follow-up clinical data were obtained from 111 patients. During the observational period, 168 alerts were confirmed from 68 patients. In patient-based analysis (alert-based analysis), PPV was 33.8% (33.9%). Sensitivity, specificity and false positive was 67.6% (83.8%), 49.4% (28.4%) and 50.6% (71.6%), respectively. Mean duration between the alert and HF event was 21.4 ± 6.1 days. On multivariate logistic analysis, maximum fluid index, LV ejection fraction and atrial fibrillation were independent predictors of HF events. The optimal cut-off value determined by receiver operating characteristic curve was 114-ohm·day with sensitivity and specificity of 89.5% and 73.0%, respectively.
Conclusions: IIM-based fluid index in patients with HF due to LV systolic dysfunction was effective in predicting worsening HF.