BACKGROUND: Hospitalizations due to decompensation are a frequent problem in treating patients with congestive heart failure (CHF). Continuous impedance measurement via implantable devices may detect pulmonary fluid accumulation due to worsening CHF. An acoustic alert might allow an earlier treatment of impending decompensation. An algorithm that implemented impedance measurement into clinical decision making in treating CHF patients was evaluated. METHODS: Forty-two CHF patients (ejection fraction: 27 +/- 6%; New York Heart Association 2.9 +/- 0.6) with cardiac resynchronization therapy and automatic impedance measurements were included. Upon an alert, a stepped therapy was initiated: category (1) overt decompensation, hospitalization; category (2) worsened CHF, increase of diuretics; category (3) no CHF worsening, brain natriuretic peptide (BNP) measurement, elevated BNP: increase of diuretics, normal BNP: no specific treatment. RESULTS: During 18 +/- 4 months, 45 alerts were treated according to the algorithm. Eleven category 1 alerts led to hospitalization; 21 category 2 and 11 category 3 patients (elevated BNP) were treated conservatively. Two category 3 alerts (normal BNP) received no treatment. CONCLUSIONS: Automatic impedance measurement can be integrated into CHF management. BNP measurement restricted to patients with alert but without clinical signs of worsened CHF may prevent premature therapy escalation.
BACKGROUND: Hospitalizations due to decompensation are a frequent problem in treating patients with congestive heart failure (CHF). Continuous impedance measurement via implantable devices may detect pulmonary fluid accumulation due to worsening CHF. An acoustic alert might allow an earlier treatment of impending decompensation. An algorithm that implemented impedance measurement into clinical decision making in treating CHFpatients was evaluated. METHODS: Forty-two CHFpatients (ejection fraction: 27 +/- 6%; New York Heart Association 2.9 +/- 0.6) with cardiac resynchronization therapy and automatic impedance measurements were included. Upon an alert, a stepped therapy was initiated: category (1) overt decompensation, hospitalization; category (2) worsened CHF, increase of diuretics; category (3) no CHF worsening, brain natriuretic peptide (BNP) measurement, elevated BNP: increase of diuretics, normal BNP: no specific treatment. RESULTS: During 18 +/- 4 months, 45 alerts were treated according to the algorithm. Eleven category 1 alerts led to hospitalization; 21 category 2 and 11 category 3 patients (elevated BNP) were treated conservatively. Two category 3 alerts (normal BNP) received no treatment. CONCLUSIONS: Automatic impedance measurement can be integrated into CHF management. BNP measurement restricted to patients with alert but without clinical signs of worsened CHF may prevent premature therapy escalation.