RESUMO
PURPOSE: Pulmonary edema (PE) is associated with fluid accumulation in the lungs. Device-based impedance measurements have been used to detect fluid overload prior to hospitalization. However, studies have reported a high false positive rate (FPR). The objective of this study was to develop and test a new multivector impedance-based algorithm that reliably tracks PE clinical events. METHODS: We enrolled patients with implanted CRT-Ds in 23 US centers within 2 weeks of device implant. Six-vector impedance data was collected automatically by the CRT-Ds every 30 min during emergency department visits/hospitalizations and every 2 h at all other times. Detection algorithms for cardiac resynchronization therapy defibrillator (CRT-D) and implantable cardiac defibrillator (ICD) devices were developed using those impedance vectors that would be available in corresponding devices and retrospectively evaluated. RESULTS: There were 75 patients (69 % male), mean age 66 ± 12 years, with a LVEF of 23 ± 6 % and QRS of 149 ± 25 ms. Twenty-one major clinical events occurred over 8.2 ± 2.6 months of follow-up time. CRT-D vector combinations resulted in a sensitivity of 71.4 % (95 % confidence interval 47.8-88.7) and a FPR of 0.56 (0.30-0.94) false positives per patient-year (FPs/pt-yr); ICD vector combinations resulted in a sensitivity of 61.9 % (38.4-81.9) and a FPR of 0.63 (0.36-0.90) FPs/pt-yr. In comparison, the single-vector RVCoil-Can implementation of this algorithm resulted in a sensitivity of 57.1 % (34.0-78.2) and a FPR of 0.74 (0.44-1.12) FPs/pt-yr. CONCLUSIONS: This multivector impedance algorithm was effective in tracking PE clinical events in this patient population. Additional studies are needed to prospectively evaluate the performance of this algorithm in a larger population.