FAST Enough? Intrathoracic Monitor Bests Weight Gain in Predicting Worsening Heart Failure

September 15, 2009

September 15, 2009 (Boston, Massachusetts) — An electronic monitor that tracks fluid accumulation in the chest, piggybacked onto a heart-failure patient's already-indicated implanted device, is more sensitive than low-tech but guidelines-recommended daily checks of the bathroom scale in predicting a growing risk of decompensation requiring a hospital or clinic visit, reported investigators here at the Heart Failure Society of America 2009 Scientific Meeting [1].

The OptiVol Fluid Status Monitoring System (Medtronic, Minneapolis, MN), FDA-approved five years ago and available since then on some of the company's implantable cardioverter defibrillators (ICDs) and cardiac resynchronization therapy-defibrillator (CRT-D) devices, measures intrathoracic impedance as a surrogate for fluid status in the chest. The idea tested in the Fluid Accumulation Status Trial (FAST) was whether a drop in intrathoracic impedance can point to potentially decompensating, potentially treatable fluid buildup before it causes symptoms.

Dr William T Abraham

A number of electronic monitors of fluid volume or pressures are being explored as "ways to proactively keep patients out of the hospital" by tracking physiologic changes that signal a need for treatment before they can lead to decompensation, observed Dr William T Abraham (Ohio State University, Columbus) for heartwire . For patients who already have indications for an ICD or CRT-D device, "it makes a lot of sense to combine them with monitoring technologies," he said.

"The current standard of care, daily weight monitoring, really performs very poorly when held up to rigorous study, and intrathoracic impedance monitoring performs much better," said Abraham. "So if we're going to continue to advocate daily weight monitoring, we should also now add intrathoracic impedance monitoring to our armamentarium. I wouldn't base a major clinical decision on either of them, but intrathoracic impedance is another piece of the puzzle that we can put together in patient assessment."

As presented by Abraham, the study's principal investigator, FAST followed 156 patients with a history of or current HF symptoms for a mean of about 18 months while they were monitored with the OptiVol system and tracked any weight changes; 85% were in NYHA class 2-3 at baseline and nearly all the remainder were in NYHA class 1.

"Fluid-index threshold crossings," the impedance signals predicting a heart-failure event, occurred 417 times in 116 of the patients (74%), or at a rate of 2.08 per patient per year. There were 890 changes in weight reaching the prespecified warning level (at least three pounds gained in one day or at least five pounds over three days) in 94 patients (60%), a rate of 4.43 per patient per year, Abraham reported.

Frequent nonevents have the potential to blunt the system responsiveness--if you keep seeing things go up, and nothing is happening, it would be like crying 'wolf.'

Of the 65 adjudicated heart-failure events that occurred in 31 patients, 48 were forecast by the OptiVol monitor and 13 were detected by weight changes. Forty of them "were successfully detected by intrathoracic impedance monitoring but not by daily weight monitoring," he said; weight changes caught five events that were missed by the device.

Weight changes had an adjusted sensitivity of only 22% for adjudicated acute heart-failure episodes, compared with a sensitivity of 76% for intrathoracic impedance monitoring (p<0.0001). False alarms--that is, a weight change or impedance signal warning of a heart-failure event that didn't materialize--were half as common with OptiVol monitoring (1.9 vs 4.3 per patient per year for weight change; p<0.0001).

Interestingly, Abraham said, weight data were available from the patients for only three of every four days of the follow-up; intrathoracic impedance data were available virtually all the time. "Even in the setting of a clinical trial, adherence to daily weight monitoring is less than desired." In an analysis that excluded days on which weight data were missing, "the performance of weight change was similarly poor" relative to how it fared overall.

"We need to revisit and question the reliability and effectiveness of weight monitoring, because in fact it is limited," Abraham said in his presentation. Yet as a heart-failure maintenance strategy, it's firmly entrenched in the guidelines and is actually a performance measure advocated by the cardiology societies and federal regulators. It can help guide diuretic and other therapies over the short term but is widely regarded as poorly sensitive for worsening heart failure in long-term management, as heart-failure patients over time can gain weight from fat or become cachectic and lose significant muscle mass, he observed.

Dr Lynne W Stevenson

"Scales can't tell the difference between fluid that's intravascular or extravascular and certainly can't tell the difference between fat and muscle mass," agreed Dr Lynne W Stevenson (Brigham and Women's Hospital, Boston, MA) in her invited commentary delivered after the FAST presentation. Clearly another type of measurement is needed for following heart-failure patients over the long term.

"It has to be something that does not rely on heart-failure specialists but will be able to be translated out into the community, where general physicians and perhaps patients can respond to it," she said.

"But what is the right measurement? In [FAST], intrathoracic impedance detected 74% of events, but it also detected 369 other events," Stevenson said. Physiologic changes are hard to interpret because spontaneous changes that don't cause events can also occur. "It's crucial that 88% of impedance crossings in this study were not associated with a following event. And responding to these [with increased diuretics] could worsen renal function and [cause] electrolyte derangements. Perhaps more concerning, frequent nonevents have the potential to blunt the system responsiveness--if you keep seeing things go up, and nothing is happening, it would be like crying 'wolf.' It will be difficult to decide when we actually do need to intervene."

Changes in weight, she said, "are not the right signal for chronic outpatient surveillance, [but] I don't think it's time to throw them away yet. They clearly have helped us. Perhaps we have to be more concerned, however, that the patient who feels bad with a stable weight may yet be deteriorating, and in a patient whose diuretics are not decreasing a weight gain may in fact signal gained fat and not fluid."

FAST was sponsored by Medtronic. Abraham reports receiving research grants and/or consulting fees from Medtronic; honoraria for speaking from Medtronic and Boston Scientific; and consulting or being on an advisory board for Medtronic, Biotronik, and St Jude Medical. Stevenson reports receiving research grants from CardioMEMS and Medtronic and honoraria for speaking from Medtronic, from which she also reports having been a consultant or being on an advisory board.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.