We studied 82 chronic HF patients implanted with a Medtronic ICD/CRT-D (mean age 66卤12 years, 62 men, ICD in 44 and CRT-D in 38, left ventricular ejection fraction 37卤14%). During follow-up, we analyzed OptiVol alerts, which were sent by monthly scheduled transmission and by unscheduled CareAlerts transmission for threshold-crossing events. The elevation patterns of OptiVol Fluid Index (OVFI), including the duration of threshold crossing, time interval from threshold crossing to peak OVFI, and peak OVFI, were compared between true decompensated HF and false-positive events.
During a mean period of 861卤239 days, 612 OptiVol alerts were transmitted from 70 of the 82 patients (85.4%). These 612 alerts were derived from 298 episodes that potentially represented decompensated HF (2.1 alerts per episode). Of these 298 possible HF events, 119 (40%) were judged as true decompensated HF based on clinical findings, while the other 179 events (60%) were false positives, including unexpected OVFI elevation (n=83), paroxysmal atrial fibrillation (n=37), excessive exercise (n=18), non-sustained ventricular tachycardia (n=10), and other causes. The analysis of the OVFI elevation pattern revealed that both the duration of threshold crossing and the peak OVFI value were significantly related to decompensated HF. The cutoff value for the duration of threshold crossing of OVFI for the diagnosis of HF was 21 days, showing 87% sensitivity and 85% specificity. The cutoff value of OVFI peak was 136.0 惟, showing 90% sensitivity and 92% specificity.
Not every OptiVol alert is indicative of decompensated HF. A threshold-crossing OVFI for >21 days and an OVFI peak>136 惟 are both strongly indicative of decompensated HF, and may be useful in the telediagnosis of HF.