The American Journal of Cardiology -...

The American Journal of Cardiology

Volume 99, Issue 10, Supplement 1, Pages.1-44 (21 May 2007)

1. Title page Page i

2. Faculty List Page iv

3. Faculty Disclosures Page v

4. Introduction Pages S1-S2 W.H. Wilson Tang

5. Fundamentals of Intrathoracic Impedance Monitoring in Heart Failure Pages S3-S10 Li Wang

6. Use of Device Diagnostics in the Outpatient Management of Heart Failure Pages S11-S16 Robin Germany and Christina Murray

7. Integrating Device-Based Monitoring into Clinical Practice: Insights from a Large Heart Failure Clinic Pages S17-S22 Roy S. Small

8. Device Monitoring of Intrathoracic Impedance: Clinical Observations from a Patient Registry Pages S23-S28 John Andriulli

9. Use of Device Diagnostics as an Educational Tool to Improve Patient Adherence Pages S29-S33 Lisa Rathman

10. Key Lessons from Cases Worldwide Pages S34-S40 Li Wang

11. Collaboration Among General Cardiologists, Heart Failure Specialists, and Electrophysiologists: What Are the Barriers? Pages S41-S44 W.H. Wilson Tang

Implantable Device-Based Monitoring ofPatients with Heart Failure:

The OptiVol Users’ Summit 2006

GUEST EDITOR:

W. H. Wilson Tang, MDAssistant Professor of Medicine

Cleveland Clinic Lerner College of Medicine of Case Western Reserve UniversityDirector for Heart Failure Research

Section of Heart Failure and Cardiac Transplantation MedicineCleveland Clinic Foundation

Cleveland, Ohio

This supplement is based on a symposium held April 29, 2006, in Chicago, Illinois.The symposium and publication of the proceedings were supported by an

unrestricted educational grant from Medtronic. Editorial support was provided byHarriet Guthertz, Medical Marketing and Communications, St. Paul, Minnesota, and

by Rete Biomedical Communications Corporation, Ridgewood, New Jersey.

SENIOR EDITORCraig Smith

SENIOR PRODUCTION EDITORMickey Kramer

EDITOR IN CHIEFWilliam C. Roberts, MD

EXECUTIVE PUBLISHERDavid Dionne

PROOF/PRODUCTION EDITORMary Crowell

MAY 21, 2007, VOL 99 (10A)

Implantable Device-Based Monitoring of Patients withHeart Failure: The OptiVol Users’ Summit 2006

Guest Editor

W. H. Wilson Tang, MDAssistant Professor of MedicineCleveland Clinic Lerner College of Medicine

of Case Western Reserve UniversityDirector for Heart Failure ResearchSection of Heart Failure and Cardiac

Transplantation MedicineCleveland Clinic FoundationCleveland, Ohio

Faculty

John Andriulli, DOAssistant Professor of MedicineDirector, Arrhythmia Device ProgramUniversity of Medicine and Dentistry of New

JerseyRobert Wood Johnson Medical SchoolCooper University HospitalVoorhees, New Jersey

Robin Germany, MDAssistant ProfessorDirector, Heart Failure Treatment ProgramUniversity of Oklahoma Health Sciences

CenterOklahoma City, Oklahoma

Christina Murray, MDClinical InstructorHeart Failure Treatment ProgramUniversity of Oklahoma Health Science

CenterOklahoma City, Oklahoma

Lisa Rathman, MSN, CRNPCongestive Heart Failure Clinic, The Heart

GroupAcute Nurse PractitionerLancaster General HospitalLancaster, Pennsylvania

Roy S. Small, MDDirector, Heart Failure Clinic, The Heart

GroupMedical Director, Heart Failure UnitLancaster General HospitalLancaster, Pennsylvania

Li Wang, PhDTechnical FellowMedtronicMinneapolis, Minnesota

Faculty Disclosures

The authors who contributed to this publication have disclosed the following industry relationships:

John Andriulli, DO, is a consultant for Medtronic and has served on the OptiVol Advisory Board. Hereceived research support from Medtronic for the IMPEDE-HF study.

Robin Germany, MD, has received grant/research support from Medtronic and speaks on behalf ofMedtronic.

Christina Murray, MD, has received honoraria from Medtronic.

Lisa Rathman, MSN, CRNP, speaks on behalf of GlaxoSmithKline, Medtronic, Novartis, and Scios Inc.;and serves as a consultant to GlaxoSmithKline and Medtronic.

Roy S. Small, MD, is consultant to Medtronic and speaks on behalf of GlaxoSmithKline and Medtronic.

W. H. Wilson Tang, MD, serves as a consultant to Amylin Pharmaceuticals, Inc., Amgen Inc.,AstraZeneca, Biosite Inc., Boston Scientific, CV Therapeutics, Inc., F. Hoffman La-Roche, Medtronic, andNovacardia, Inc.; and has received grant/research support from Amgen Inc. and GlaxoSmithKline. He is amember of the Speakers’ Bureau for Takeda Pharmaceuticals North America, Inc.

Li Wang, PhD, is an employee of Medtronic.

(flfaOiciu

tatlbrErrc

dtcoSivvteIIsa

imiihii

O

vC

0d

Introduction
W. H. Wilson Tang, MD
ibmDuppmnoemcimoiotpts

tihtidoni

In November 2004, the US Food and Drug AdministrationFDA) approved a novel heart failure monitoring tool, OptiVoluid status monitoring (Medtronic, Inc., Minneapolis, MN),or use on some biventricular pacemakers. The feature is nowlso available on some implantable cardioverter defibrillators.ptiVol fluid status monitoring allows the measurement of

ntrathoracic impedance from the right ventricular lead to thean of the device. The OptiVol fluid index, a derivative of thempedance signal over time, can serve as an indicator fornderlying fluid accumulation.

Although some physicians have enthusiastically embracedhe concept of this new technology, many have been skepticalnd are exploring how to best use the information to care forheir patients. Until recently, access to OptiVol data has beenimited to direct device interrogation. OptiVol was approvedased on limited published data, without evidence-based algo-ithms on how to use the information stored in the device.arly experience has been limited largely to anecdotal case

eports. Therefore, clinicians who have incorporated intratho-acic impedance data into the care of their patients have be-ome the experts in this new technology.

Clinicians eager to share their experiences with this newevice-based diagnostic tool suggested the creation of a venueo discuss the use of intrathoracic impedance assessment inlinical practice. This grassroots approach parallels the devel-pment of user groups in the computer software industry.ubsequently, �150 participants from around the country,

ncluding practicing cardiologists, electrophysiologists, ad-anced practice nurses, and device nurses, joined those in-olved in the development of the OptiVol algorithms to shareheir experiences and exchange ideas at the first OptiVol Us-rs’ Summit meeting, held April 28–29, 2006, in Chicago,llinois. The OptiVol Users’ Summit, sponsored by Medtronic,nc., has initiated an exciting and much needed dialogue. Thisupplement to The American Journal of Cardiology featuresrticles presented at this inaugural OptiVol Users’ Summit.

The first article, by Dr. Li Wang, whose work has beennstrumental in the clinical development of OptiVol fluid statusonitoring, begins with a description of the fundamentals of

ntrathoracic impedance measurements and the OptiVol fluidndex. Next, Dr. Robin Germany, an electrophysiologist andeart failure cardiologist, and Dr. Christina Murray, discuss thentegration of device-based diagnostics into the outpatient clin-cal setting. Cardiologist Dr. Roy S. Small provides further

Department of Cardiovascular Medicine, Cleveland Clinic, Cleveland,hio, USA.

Address for reprints: W. H. Wilson Tang, MD, Department of Cardio-ascular Medicine, Cleveland Clinic, 9500 Euclid Avenue, Desk F25,leveland, Ohio 44195.

E-mail address: [email protected].

002-9149/07/$ – see front matter © 2007 Elsevier Inc. All rights reserved.oi:10.1016/j.amjcard.2007.02.008

nsight by focusing on the challenges of integrating device-ased monitoring, and particularly intrathoracic impedanceonitoring, into a high-volume private group practice. Next,r. John Andriulli, an electrophysiologist interested in thetility of impedance measurements in the arrhythmia practice,resents the findings of a retrospective analysis of a registry ofatients with implantable devices with intrathoracic impedanceonitoring capabilities. Nurse Lisa Rathman, a heart failure

urse clinician, contributes a presentation highlighting the usef intrathoracic device diagnostics as an educational tool tonhance patient adherence to medication and dietary recom-endations. Dr. Wang follows with an article on a collection of

ases discussing refinements in the approach used to interpretntrathoracic impedance tracings. The supplement concludes withy article on the many challenges that lie ahead for electrophysi-

logists, heart failure cardiologists, and nurses in the successfulncorporation of device-based monitoring and a brief summary ofngoing and upcoming trials of OptiVol fluid status monitoringhat should provide answers to current challenges. It is hoped theseresentations will provide a “nuts and bolts” understanding ofhese device-based measurements and stimulate further discus-ions.

As we await further insights from various clinical regis-ries and results from pivotal studies using device-basednformation to guide clinical management in patients witheart failure, it is very clear that this new format of peer-o-peer learning will provide insights into creative uses ofntrathoracic impedance and other innovative device-basedata and observations far beyond the original intended usesf this technology. This format will be vital as increasingumbers of innovative hemodynamic and remote monitor-ng devices become available for heart failure management.


www.AJConline.org

Topothdai

A

IMttsa

2G The American Journal of Cardiology (www.AJConline.org) Vol 99 (10A) May 21, 2007

raditional clinical trials are unlikely to provide informationn how to incorporate data from these devices into clinicalractice in a timely fashion. Therefore, we will need to relyn interested clinicians and healthcare providers to evaluatehe potential utility and limitations of these new data. It isoped that further interest in the clinical application ofevice-based data and, in particular, intrathoracic imped-nce will be sparked by the discussions initiated at this
naugural users’ summit. p
cknowledgment

thank Kelly Tubbs, Kris Sexton, and Deana Dickenson ofedtronic Inc., for their hard work and dedication in making

he OptiVol Users’ Summit project a reality. I would also likeo thank Harriet Guthertz for editorial assistance on the manu-cripts, and Linda Kelly (Cleveland Clinic, Cleveland, OH)nd Jolene Runkel (Iowa Heart Center, Des Moines, IA) for
articipating in the OptiVol Users’ Summit.

HhAphptyavhisrcespssi

M

0d

Fundamentals of Intrathoracic Impedance Monitoringin Heart Failure

Li Wang, PhD

The primary objective of the first-generation implantable cardiac pacemakers wasto provide critical heart rate support, but these devices did not have any diag-nostic capabilities. In the intervening decades, the number, type, and complexityof implantable devices has greatly expanded. Today, implantable devices not onlyprovide heart rate support but they also provide protection from sudden cardiacdeath with implantable cardioverter defibrillators (ICDs) and reduce symptomsand increase survival with cardiac resynchronization therapy (CRT). Further-more, information on physiologic variables has been collected in patients withimplanted devices for the purpose of providing sophisticated closed-loop optimi-zation of their pacing and defibrillation algorithms. Thoracic fluid statusmonitoring via intrathoracic impedance is the newest device-based diagnosticcapability. For those patients with heart failure who are already targeted toreceive an ICD or CRT with defibrillator implant, the ability to monitor fluidstatus can provide additional insight into the difficult problem of evaluating andmanaging these patients. This article reviews the basics of measuring intratho-racic impedance via OptiVol fluid status monitoring (Medtronic, Inc., Minneap-olis, MN), as well as clinical results regarding the utility of evaluating OptiVolintrathoracic impedance data trends. © 2007 Elsevier Inc. All rights reserved.
(Am J Cardiol 2007;99[suppl]:3G–10G)
tv

pFtllcosgcd

DS

TmtSna

eart failure is a large and growing problem costingealthcare systems billions of dollars each year. Themerican Heart Association (AHA) estimates that ap-roximately 5 million people in the United States haveeart failure, with �500,000 new cases each year.1 Ap-roximately $30 billion was spent in the United States onhe direct and indirect costs of managing heart failure lastear.1 Logically, careful monitoring of fluid status inmbulatory patients with chronic heart failure can pro-ide an early warning of impending decompensation andelp reduce the number of heart failure–related hospital-zations. Regular monitoring of body weight and clinicalymptoms has been advocated for this purpose, but itseliability as a surrogate for clinical stability can behallenging. Likewise, patient characteristics may influ-nce the interpretation of more objective diagnostic mea-ures, such as B-type natriuretic peptide.2 Additionally,ulmonary congestion is difficult to recognize in its earlytages of development because of the late appearance ofymptoms before hospitalization.3,4 A mismatch may ex-st between the signs and symptoms of the patients and

Medtronic, Inc., Minneapolis, Minnesota, USA.Address for reprints: Li Wang, PhD, 7000 Central Avenue NE,

inneapolis, Minnesota 55432.
dE-mail address: [email protected].

heir true cardiopulmonary status, which can delay inter-ention.5,6

Thoracic fluid status monitoring via intrathoracic im-edance is the newest device-based diagnostic capability.or those patients with heart failure who are already

argeted to receive an implantable cardioverter defibril-ator (ICD) or cardiac resynchronization therapy/defibril-ator (CRT-D) implant, the ability to monitor fluid statusan provide additional insight into the difficult problemf evaluating and managing these patients. This discus-ion focuses on the OptiVol fluid status monitoring al-orithm (Medtronic, Inc., Minneapolis, MN), which islinically available in several ICD (Virtuoso) or CRT-Devices (InSync Sentry and Concerto).

evelopment and Operation of the OptiVol Fluidtatus Monitoring Algorithm

he concept of using external transthoracic impedance toonitor pulmonary congestion has a long history, dating

o research sponsored by the National Aeronautics andpace Administration (NASA) at the University of Min-esota in the 1960s and has been validated in animals7,8

nd in humans.9,10 Over the past few decades, several
evice configurations and algorithms have been devel-
www.AJConline.org

ot

cmabctepatBmbcmtt

apsTfliuasAts(

ai

I

Avtctf�app

mtataltimc

FO

T

FMl LA � l


ped, using external electrodes to acquire impedance datao estimate a wide range of hemodynamic variables.

OptiVol fluid status monitoring represents the appli-ation of the concept of transthoracic impedance as aeasure of pulmonary fluid accumulation into implant-

ble devices. Assessment of intrathoracic impedance cane achieved by measuring impedance between the devicease (typically implanted in the left pectoral region) andhe lead in the right ventricle (Figure 1). This vectorncompasses much of the left thoracic cavity. As theatient has worsening heart failure with increasing lefttrial filling pressure caused by left ventricular dysfunc-ion, more fluid is retained in the pulmonary circulation.lood is among the most conductive tissues, whicheans it is among the least resistant tissues in the human

ody. Therefore, as fluid accumulates in the pulmonaryirculation, a reduction in impedance is expected. Theajor advantage is the relatively fixed positions between

he 2 implanted electrodes, which provide more consis-ent and reliable measurements.

Pulmonary congestion starts with vascular congestionnd develops into interstitial congestion. If untreated,ulmonary edema develops in the alveoli and the inter-titium, thereby putting the patient in a congestive state.hus, any type of pulmonary congestion would be re-ected in changes in intrathoracic impedance and is of

nterest in managing patients with congestive heart fail-re. Using the broad concept of Ohm’s law, the imped-nce (or “resistance”) can be measured by delivering amall alternating current between the lead and the device.s stated above, the more fluid in the path of the impulse,

he less resistance (or lower impedance) will be mea-ured. Therefore, anything that changes the fluid status

igure 1. The use of intrathoracic impedance to monitor pulmonary congeN) measures intrathoracic impedance many times a day using the vector

ungs because of worsening heart failure (HF), the impedance decreases.

such as vascular congestion, interstitial congestion, or a

lveolar edema) may be reflected by a reduction in thempedance measurement.

mplant Considerations

pplication of impedance monitoring to implantable de-ice technology requires several considerations duringhe initial implant surgery. Edema in the device pocketan likely lower the impedance measurements becausehe pocket is in the path of the measurement area. There-ore, OptiVol fluid status monitoring is not initiated until34 days after implant because the earlier data may be

ffected by postsurgical edema and inflammation in theocket area and may not reliably indicate worseningulmonary congestion.

Animal data have suggested that the specific place-ent location of the right ventricular lead will not affect

he impedance measurements. Whether the lead is placedt the septum or at the apex, the size of the right ven-ricular coil electrode should be of sufficient size tossess impedance across the intrathoracic region, regard-ess of lead position. Although it is theoretically possiblehat body position can affect impedance measurements,ts impact is unlikely to be significant with an impedanceeasurement averaged over 5 hours, as implemented in

urrent devices.

actors Affecting Intrathoracic Impedance andptiVol Fluid Trend Status

here are several factors that may affect the reliability

ema. The OptiVol fluid status monitoring (Medtronic, Inc., Minneapolis,n the device case and the right ventricular lead. As fluid is retained in theeft atrial; 1 � increase; 2 � decrease.

stion/edbetwee

nd specificity of the intrathoracic impedance monitoring

apscibtimFnmcambtmtct

I

Ws6rdmpast

spDFdifadwns

ins

i

ltepsitaisTis

arflpds(h

gcoM�c

dactptt

ichiathticm

mfih

5GWang/Intrathoracic Impedance Monitoring in Heart Failure

lgorithm. Intrathoracic processes, such as pneumonia orleural effusion, can affect intrathoracic impedance mea-urements. Any pulmonary events that occur in the lungontralateral to the implanted device should not affect thentrathoracic impedance data, although it would have toe a very isolated event to only occur in one lung and nothe other. Similarly, the air volume in the lung can affectntrathoracic impedance. Air itself is an insulator, and theore air there is in the lungs, the higher the impedance.or example, in patients with chronic obstructive pulmo-ary disease, the functional residual volume in the lungsay change over time, so the impedance measurements

an be affected over the long term. Therefore, consider-tion should be given for those diseases or conditions thatay affect the impedance measurements. Furthermore,

ecause intrathoracic impedance can only assess conges-ion within the pulmonary region, the system cannoteasure and track peripheral edema or abdominal ascites

hat occur outside the measurement field. Nevertheless,hanges from the patient’s own reference levels shouldheoretically be unaffected in normal circumstances.

ntrathoracic Impedance Measurements

ith OptiVol fluid status monitoring, impedance is mea-ured every 20 minutes from noon to 5 PM for a total of4 measurements over a 5-hour period. The deviceecords the average of those 64 measurements as theaily impedance value. By averaging many measure-ents over a 5-hour period, the effects of respiration and

osture on impedance are minimized. The system offersresolution of 0.25 �. The impedance measurements

tart after ventricular fibrillation detection is enabled forhe first time, and it cannot be reset.

The OptiVol algorithm aims to track a patient’s fluidtatus over time and provides insight into whether theatient is trending toward a more wet or a more dry state.ata from the Medtronic Impedance Diagnostics in Heartailure Patients Trial (MIDHeFT) study were used toevelop the algorithm for OptiVol fluid status monitor-ng.11 The sampling period was chosen based on datarom MIDHeFT, which showed that intrathoracic imped-nce and the patient’s fluid status were best correlateduring this period. The algorithm is used to determinehether fluid buildup is occurring, and it calculates aumber of values based on intrathoracic impedance mea-urements.

Average daily impedance: A total of 64 intrathoracicmpedance measurements are made each day betweenoon and 5 PM and averaged to arrive at a single mea-urement for the day (Figure 2).12

Reference impedance: Reference impedance is initial-
zed on the 34th day of measurements by averaging the i
ast 4 average daily impedance values. This is the onlyime that averaging is involved in changes to the refer-nce impedance. After initialization, the reference im-edance trend established by the OptiVol algorithmlowly adapts to fluid changes (Figure 2). The referencempedance reflects a relatively slow-moving impedancerend, and is initiated �34 days after implantation tollow time for pocket and lead maturation. At this time,t is expected that pocket wound swelling will haveubsided and will not affect impedance measurements.he reference impedance trend is not programmable and

s a reflection of each patient’s status, with each patienterving as his or her own control.

OptiVol fluid index: Any cumulative consecutive neg-tive deviations in the average daily impedance from theeference impedance are plotted to create the OptiVoluid index (Figure 2). The OptiVol index contains 2ieces of information: the magnitude of impedance re-uction with regard to the reference impedance (mea-ured in ohms), and the sustainability of that reductionmeasured in days). The OptiVol fluid index, therefore,as a measurement unit of ohm-days.

OptiVol threshold: The OptiVol threshold can be pro-rammed to indicate a level that could be of potentiallinical relevance for that patient (Figure 2). This thresh-ld level is nominally programmed at 60 �-days based onIDHeFT data, but it can be adjusted between 30 –180-days based on the fluid tolerance level that the clini-

ian determines is best for that patient .The case illustrated in Figure 2 shows that the average

aily impedance decreased below the reference imped-nce line during decompensation, but it eventuallyrossed back over the reference impedance line. Whenhe average daily impedance is above the reference im-edance line for 2–3 days, the OptiVol fluid index resetso zero. In most cases the OptiVol fluid index will con-inue to rise before it is reset to zero.

The OptiVol algorithm provides some unique capabil-ties. For example, each patient serves as his or her ownontrol. The clinician always compares the patient withis or her own data rather than trying to manage a specificmpedance value. This approach is different from imped-nce measured from external impedance cardiographicechnologies using population-based thresholds. Becauseeart failure is a dynamic condition, the ability to adjusthe reference impedance level automatically is importantn monitoring for changes. Additionally, the physicianan determine the level (threshold) of fluid buildup thatay be of clinical significance for each patient.A more detailed understanding of OptiVol fluid status

onitoring can best be seen in 2 clinical examples. In therst example, the patient was hospitalized for worseningeart failure on September 11, 2001 (Figure 3).13 The
mpedance was relatively stable 2 weeks before hospital-

FtP

ta(

Fil


igure 2. Algorithm to track fluid accumulation. The OptiVol (Medtronic, Inc., Minneapolis, MN) algorithm calculates a number of values to aid inracking intrathoracic fluid accumulation. (A) The average daily impedance is an average of 64 impedance measurements made between noon and 5M. The reference impedance is initialized 34 days after implant by averaging the last 4 average daily impedance values. The trend then adapts slowlyo fluid changes and is a reflection of each patient’s fluid status. (B) Cumulative consecutive negative deviations between the average daily impedancend the reference impedance are plotted to create the OptiVol fluid index, which is a measure of the magnitude and duration of an impedance reduction.
Data on file, Medtronic, Inc.12)
igure 3. Impedance before and during hospitalization for heart failure. As the patient experienced worsening heart failure and eventual hospitalization,ntrathoracic impedance decreased. After diuresis and resolution of heart failure symptoms, intrathoracic impedance increased to the presymptomatic
evel. CHF � congestive heart failure; Pt � patient. (Data on file, Medtronic, Inc.13)

idwI2tcpukpas

Csiwacttisi

rftl

ip

C

Issttap2DdaAvIeufbti

rr

FeT(


zation, but it then started to decrease and continued toecrease until the day of admission as a reflection oforsening heart failure and pulmonary fluid overload.

mpedance decreased approximately 25% over the-week period. If the OptiVol algorithm is applied tohese data, the OptiVol threshold would have beenrossed during this time frame. Aggressive diuresis waserformed over the course of 4 days in the cardiac carenit (CCU), and the patient eliminated approximately 6g of fluid. Average daily impedance increased while theatient was hospitalized over the 4 days; it recovered tolevel similar to that before the patient’s heart failure

tarted to worsen.In the second example, the patient was admitted to the

CU with a fairly high pulmonary capillary wedge pres-ure (PCWP) of about 30 –33 mm Hg (Figure 4).12 Dur-ng the 3.5 days that the patient was in the CCU, PCWPas measured every 2 hours. During this time frame,

pproximately 5 kg of fluid were eliminated, with aorresponding decrease in PCWP to 10 mm Hg at theime of discharge. Additionally, as PCWP was reduced,he average daily impedance levels increased correspond-ngly. The correlation between intracardiac filling pres-ures and impedance levels is highly statistically signif-cant (r � �0.95, p �0.0001).

In summary, both examples illustrate how impedanceeduction reflected worsening fluid overload before heartailure–related hospitalization. After the patients werereated and underwent diuresis, there was a good corre-

igure 4. Correlation between impedance and pulmonary capillary wedgxample was hospitalized with a fairly high PCWP. As diuresis of thehe inverse correlation was fairly high (r � �0.95, p �0.0001). The

Data on file, Medtronic, Inc.12)

ation (r � �0.95) between the increase in intrathoracic t

mpedance and the reduction in intracardiac fillingressures.

linical Studies

ntrathoracic impedance assessment using OptiVol fluidtatus monitoring has been studied in the MIDHeFTtudy.11 The MIDHeFT study used a pacemaker modifiedo accept an ICD lead and special, downloaded softwareo automatically measure and record intrathoracic imped-nce in patients with severe heart failure. A total of 33atients were enrolled, and the mean follow-up time was0 � 8 months. The study was divided into 2 phases.uring the acute phase, patients hospitalized for acuteecompensated heart failure were transferred to the CCUnd received intravenous diuretics and other medications.

pulmonary artery catheter was inserted, and the PCWPalues were measured every 2 hours, whenever possible.ntrathoracic impedance was measured and averaged ev-ry 6 hours. In the chronic phase, patients were followedp monthly during the first year and as needed for heartailure disease management during the second year. Inoth phases, the clinicians and patients were blinded tohe impedance data, and patients were managed accord-ng to normal standard medical practice.

A total of 9 patients were hospitalized 24 times. Self-eported data on symptoms before either an emergencyoom visit or hospitalization was available for 20 pa-

ure (PCWP) during hospitalization for heart failure. The patient in thisprogressed, PCWP decreased, and intrathoracic impedance increased.

represents the linear regression line between PCWP and impedance.

e presspatientcurve

ients. Sequential PCWP values were available from 5

pdbsf�btbp1dcdtsoeh

dciHp1h

Ittnpvcop

D

Tbdrc(mra(w

Fh n.11)


atients with 14 hospitalizations. Briefly, the MIDHeFTata suggest that (1) intrathoracic impedance (measuredy average daily impedance) and intracardiac filling pres-ures (measured by a reduction in PCWP during heartailure hospitalization) were inversely correlated (r �0.61, p �0.001), (2) intrathoracic impedance decreased

efore symptom onset for all heart failure–related hospi-alizations, (3) symptom onset occurred 3.0 � 2.5 daysefore admission, (4) the decrease in intrathoracic im-edance levels preceded symptom onset by an average of5.3 � 10.6 days, and (5) the average decrease in averageaily impedance was 12.3 � 5.3% (p �0.001) and de-reased over an average of 18.3 � 10.1 days (range, 3– 42ays) before hospitalization (Figure 5).11 Although rela-ively small in sample size, the MIDHeFT study demon-trated the possibility of providing an advanced warningf pulmonary congestion at a presymptomatic stage andnabling early interventions with the goal of preventingeart failure decompensation.

The OptiVol fluid status monitoring algorithm wasesigned to be sensitive to changes in impedance thatould possibly lead to a decompensation, while minimiz-ng false-positive findings. When applied to the MID-eFT data, the algorithm detected changes in impedancereceding hospital admission with 76.9% sensitivity and.5 false-positive (threshold crossing without subsequent

igure 5. Comparison of intrathoracic impedance at reference baselineeart failure in 9 patients. (Reprinted with permission from Circulatio

ospitalization) detections per patient-year of monitoring. t

n considering the rate of false-positive results, it is impor-ant to keep the definition in mind. In the MIDHeFT study,here were examples of patients crossing the threshold for aumber of reasons, including dietary nonadherence. Theatients were not hospitalized because the physician inter-ened by increasing the patient’s diuretics. However, be-ause the patients were not actually hospitalized, the thresh-ld crossings were considered false-positive results for theurpose of the study.

ata Access: The OptiVol Example

o fully realize the clinical potential of using device-ased data for heart failure monitoring purposes requireselivery of the right information to the right people at theight time in the right way. Implantable devices are nowapable of generating trend data in summary reportsFigure 6).14 Up to 14 months of continuous trend infor-ation on the average daily impedance in relation to the

eference impedance and on the OptiVol fluid index isvailable as part of the Cardiac Compass Trends reportFigure 6).14 The impedance data are temporally alignedith other parameters useful in the management of pa-

ay before admission for 24 hospitalizations resulting from worsening
and 1 d
ients with heart failure, including daily atrial fibrillation

bnt

DeploTdHmmcNmtb

C

Itf

cpci

Fat Report


urden and ventricular rates, patient activity, daytime andighttime heart rates and heart rate variability, as well ashe percentage of time pacing.

There are several ways of accessing these reports.uring a clinic visit, the device programmer can directly

xtract trend reports by standard device interrogationrocedures. The 90-day trend data can also be down-oaded by the CardioSight Reader (Medtronic, Inc.) at theutpatient clinic for those without a device programmer.his can be achieved via transmission of interrogatedata to a secure server, followed by a fax report of theeart Failure Management Report to the clinic in �10inutes. In the United States, the Heart Failure Manage-ent Report (including the OptiVol fluid index trends)

an also be accessed through the Medtronic CareLinketwork (Medtronic, Inc.) when patients remotely trans-it their device information from home. The report ob-

ained via CareLink provides the same 14 months of dataut without the OptiVol threshold line.

onclusion

ntrathoracic impedance assessment is a promising newool in the management of patients with chronic heart

igure 6. The OptiVol fluid trends report presents up to 14 months of rollnd reference impedance. The report is part of the cardiac compassherapy/defibrillator [CRT-D] devices) and Heart Failure Management

ailure. However, there is much to learn from practicing

linicians on how to best use its capabilities in clinicalractice. It is only through a cooperative journey that wean improve intrathoracic impedance monitoring capabil-ties in order to better suit current clinical practice.

1. Thom T, Haase N, Rosamond W, Howard VJ, Rumsfeld J, ManolioT, Zheng ZJ, Flegal K, O’Donnell C, Kittner S, et al, for theAmerican Heart Association Statistics Committee and Stroke Sta-tistics Subcommittee. Heart disease and stroke statistics—2006update: a report from the American Heart Association StatisticsCommittee and Stroke Statistics Subcommittee. Circulation 2006;113:e85– e151.

2. Daniels LB, Clopton P, Bhalla V, Krishnaswamy P, Nowak RM,McCord J, Hollander JE, Duc P, Omland T, Storrow AB, et al. Howobesity affects the cut-points for B-type natriuretic peptide in thediagnosis of acute heart failure: results from the Breathing Not Prop-erly multinational study. Am Heart J 2006;151:999–1005.

3. Friedman MM. Older adults’ symptoms and their duration beforehospitalization for heart failure. Heart Lung 1997;26:169 –176.

4. Evangelista LS, Dracup K, Doering LV. Treatment-seeking delaysin heart failure patients. J Heart Lung Transplant 2000;19:932–938.

5. Mahdyoon H, Klein R, Eyler W, Lakier JB, Chakko SC, GheorghiadeM. Radiographic pulmonary congestion in end-stage congestive heartfailure. Am J Cardiol 1989;63:625–627.

6. Stevenson LW, Perloff JK. The limited reliability of physical signsfor estimating hemodynamics in chronic heart failure. JAMA 1989;261:884 – 888.

7. Luepker RV, Michael JR, Warbasse JR. Transthoracic electrical im-pedance: quantitative evaluation of a non-invasive measure of thoracic

ormation on the patient’s average daily impedance, OptiVol fluid index,(implantable cardioverter defibrillator and cardiac resynchronization(CRT-D devices). (Data on file, Medtronic, Inc.14)

ing inftrends

fluid volume. Am Heart J 1973;85:83–93.

1

1

11


8. Baker LE, Denniston JC. Noninvasive measurement of intrathoracicfluids. Chest 1974;65(suppl):35S–37S.

9. Sahalos JN, Nicolaidis A, Gotsis N. The electrical impedance of thehuman thorax as a guide in evaluation of intrathoracic fluid volume.Phys Med Biol 1986;31:425–439.

0. Van de Water JM, Mount BE, Barela JR, Schuster R, Leacock FS.
Monitoring the chest with impedance. Chest 1973;64:597–603. 1
1. Yu CM, Wang L, Chau E, Chan RH, Kong SL, Tang MO, Christensen J,Stadler RW, Lau CP. Intrathoracic impedance monitoring in patients withheart failure: correlation with fluid status and feasibility of early warningpreceding hospitalization. Circulation 2005;112:841–848.

2. Data on file. Medtronic, Inc., Minneapolis, MN.3. Data on file. Medtronic, Inc., Minneapolis, MN, 2001.
4. Data on file. Medtronic, Inc., Minneapolis, MN, 2004.

AihcciiIdfmfcp

DO

EhahcvpoT

o

HC7

0d

Use of Device Diagnostics in the OutpatientManagement of Heart Failure

Robin Germany, MD, and Christina Murray, MD

Implantable cardiac devices now go far beyond the functionality of early pacemakersand defibrillators. They provide us with an unprecedented amount of informationabout the day-to-day status of our patients with heart failure, providing insight intothe complex clinical picture of these patients. More advanced technology can be seenon the horizon, and the development of strategies to incorporate this information intoour clinics is needed. Traditionally, information from devices has only been accessedby the physicians implanting the device. However, heart failure physicians and theirpatients can benefit from the information now available and from having ways toaccess and use the data. This article focuses on the successful integration of device-based data into patient care in the outpatient clinical setting. © 2007 Elsevier Inc.
All rights reserved. (Am J Cardiol 2007;99[suppl]:11G–16G)
ii

nRtmptstlavt

T

Mdtsaispu

atFaae

lthough some practices incorporate device or electrophys-ology (EP) nurses into their practices, cardiologists andeart failure physicians are often removed from routineontact with these specialized nurses. Similarly, not allardiology clinics have programmers in their office. In clin-cs where a programmer is available, clinicians have differ-ng degrees of comfort and expertise in using these devices.n the past, it was necessary only to collect data fromevices at routine intervals to check for appropriate deviceunction. In this new era of data integration, device checksay be beneficial at each clinic visit and perhaps even more

requently, based on the patient’s clinical status. Reportsollected at these intervals can be tailored with a focus onatient management rather than just device management.

evice Interrogation in the Non-Electrophysiologistffice

P nurses can be recruited to develop a process to makeeart failure reports accessible to the treating physician andlso to provide support for the heart failure clinic. This canelp ease the transition in cases in which physicians in thelinic are not accustomed to looking at the data from de-ices. Patients can have their devices interrogated beforehysician contact, and the resulting reports can be printedn full-sized paper and placed on the front of the chart.hese reports focus on the heart failure–related data stored

Heart and Lung Center, Heart Failure Treatment Program, Universityf Oklahoma Health Sciences Center, Oklahoma City, Oklahoma, USA.

Address for reprints: Robin Germany, MD, Heart and Lung Center,eart Failure Treatment Program, University of Oklahoma Health Sciencesenter, 920 Stanton L. Young Boulevard, WP-3120, Oklahoma City, OK3104.

tE-mail address: [email protected].


n the device and should not replace ongoing device mon-toring by the implanting physician.

With the availability of an interrogation device that doesot have programming capabilities, such as the CardioSighteader (Medtronic, Inc., Minneapolis, MN), medical assis-

ants can retrieve data from all compatible devices. Theedical assistants can then record this information in the

atient’s chart, just as they would vital signs and medica-ions. If the device is not compatible, a device nurse musttill interrogate the device before the physician visit. Afterhis interrogation procedure becomes routine, it adds veryittle time to the patient visit, yet it provides a tremendousmount of insight. These visits do not replace routine EPisits, but the device nurse does not have to see all patientso obtain heart failure data.

ypes of Data Obtained

any different types of information can be garnered fromevice reports (Figure 1). These data allow patient care to beailored for the purposes of both symptom management andafety, including titration of medications and institution ofnticoagulation therapy. In addition to the usual clinicalnformation, data patterns may emerge on these reports. Inome cases, evidence of medication noncompliance may beresent, and in cases where proarrhythmic drugs are beingsed, this should be carefully assessed.

Cardiac rhythm: Most implantable devices provide thebility to monitor episodes of nonsustained ventricularachycardia. Daytime and nighttime heart rates are tracked.or devices with atrial and ventricular leads, hours spent intrial fibrillation and ventricular rate during atrial fibrillationre recorded. When tracked over time, all of these param-ters allow a more comprehensive approach to medication
itration. Other findings, such as previously unsuspected
www.AJConline.org

pr

ttvdl

gr3pp

h

FapH rillator;


aroxysmal atrial fibrillation, may be present in theseeports.

Heart rate variability: Heart rate variability, defined ashe standard deviation of 5-minute median atrial–atrial in-ervals, can be calculated and recorded over time. Heart rateariability has been shown to be predictive of sudden car-iac death in patients with heart failure. A study that ana-

igure 1. Insight into patient status. The Heart Failure Management Repdditional trends to assess patient status (B). In addition to OptiVol data, thatient’s daily atrial fibrillation (AF) burden, ventricular (V) rate during atrR variability, and percent pacing. ICD � implantable cardioverter defib

yzed heart rate variability in 16-minute electrocardio- h

raphic recordings was able to stratify patients into a high-isk category using low HRV and frequent PVCs. The-year sudden cardiac death rate was 23% when thoseredictors were present, versus 3% in patients without eitherredictor.1

Heart rate variability in a population with heart failureas also been shown to be lower in patients with death or

ludes OptiVol (Medtronic, Inc., Minneapolis, MN) fluid trends (A) andt provides up to 14 months of temporally aligned trend information on theyarrhythmia (AT)/AF, patient activity, day and nighttime heart rate (HR),VF � ventricular fibrillation; VT � ventricular tachyarrhythmia.

ort ince reporial tach

ospitalization, and the decrease in heart rate variability was

szasc

amtau

tmvisDtpdb

1. (Co

13GGermany and Murray/Device Diagnostics in Outpatient Management of Heart Failure

een up to 3 weeks before the event.3 Cardiac resynchroni-ation therapy has been shown to improve heart rate vari-bility, possibly by shifting autonomic control away fromympathetic dominance.2 We have used these data in ourlinic to help determine the frequency of follow-up.

Percentage of atrial/ventricular pacing: The percent-ge of atrial and ventricular pacing is important in theanagement of patients with heart failure. Multiple recent

rials have found worsened left ventricular ejection fractionnd other clinical outcomes with continuous right ventric-

Figure

lar pacing. The Second Multicenter Automatic Defibrilla- r

or Implantation Trial II (MADIT-II) showed a trend ofore frequent hospitalization in the implantable cardio-

erter defibrillator (ICD) group (19.9 vs 14.9, p � 0.09), butt was unclear if this effect was because of longer life spans,hock-related admissions, or right ventricular pacing.4 Theual Chamber and VVI Implantable Defibrillator (DAVID)

rial enrolled 506 patients with ICD indications but with noacemaker indications. Devices were implanted and ran-omized to ventricular backup pacing only at a rate of 40eats per minute (VVI-40) or dual-chamber pacing rate-

nt’d)

esponsive pacing with a low rate of 70 beats per minute

(t(tp�MpamrauNm

ha(ascc

udftisbOt

hIdpaiothf

dtLSrtefb

W

Fuoptatmwdsmwtsca

R

TaavdibqtdatcbrRw

otvawhmtbdptm


DDDR-70) with forced right ventricular pacing. Patients inhe VVI-40 group had fewer primary end point eventsdeath or heart failure hospitalizations) when compared withhe DDR-70 group (hazard ratio [HR], 1.61; p �0.03). Inarticular, those patients with right ventricular pacing40% had worse event-free rates (p � 0.09).5 Study of theode Selection Trial in Sinus Node Dysfunction (MOST)

opulation evaluated complications of heart failure associ-ted with right ventricular pacing. Pacing �40% in DDDRode was associated with an increased risk of heart failure–

elated hospitalizations (HR, 2.99). Furthermore, there waslinear relation between percent cumulative right ventric-

lar pacing and the development of atrial fibrillation.6

ewer devices with managed ventricular pacing algorithmsay be beneficial in avoiding these adverse outcomes.

Patient activity: Patient activity can be plotted as activeours during the day. Studies have documented that patientctivity decreases as a patient approaches hospitalization188 min/day down to 164 min/day, p � 0.028).3 Patientctivity level is also useful in patient education. We oftenhow the patient activity report to our patients to discusshanges in their physical activity and to encourage in-reased activity.

Intrathoracic impedance: Newer and more heart fail-re–specific device-based markers have been and are beingeveloped. For example, intrathoracic impedance can beollowed in some newer devices via a derived fluid index,he OptiVol fluid index (Medtronic, Inc.). Impedance isnversely proportional to pulmonary capillary wedge pres-ures. Pilot studies on this concept documented changesefore heart failure exacerbation requiring hospitalization.7

ngoing studies will determine the prospective validity ofhis concept.

Other measurements and devices: A new implantableemodynamic monitoring system, Chronicle (Medtronic,nc.), which provides continuous intracardiac hemodynamicata about the patient, is currently awaiting regulatory ap-roval. The data have been shown to be reproducible, botht rest and with activity, over time.8 This device is specif-cally designed to provide information on the managementf heart failure and will provide additional information onhe clinical status of the patient when not in the office. Thisemodynamic monitor is only the first in a number of heartailure management devices currently being developed.

Some implantable devices have been paired to externalevices, such as a weight scale and blood pressure cuffhat can be accessed over the Internet. An example is theatitude patient management system (Boston Scientific,t. Paul, MN). These data along with arrhythmia and heartate variability data can be transmitted and monitored overhe Internet with physician notification via e-mail in thevent of significant changes. These data can also be printedrom the Internet database and placed on the patient’s chart
efore a clinic visit. i
hich Patients Should Be Monitored?

or physicians taking care of patients with heart failure, these of device data is not limited to a diagnostic “spot check”f thresholds and impedance data. These data can alsorovide insight into what is happening with patients whenhey are not in the office. Ideally, reports should be collectedt every patient encounter. Otherwise, the clinical utility ofhe information is significantly reduced. Many conditionsonitored by devices are relatively asymptomatic. A patientho developed atrial fibrillation in the last 24 hours may notevelop symptoms of heart failure for a few weeks. Evenignificant episodes of nonsustained ventricular tachycardiaay not be perceptible to the patient. Thus, in our practice,e follow up all patients with compatible devices at the

ime of their clinic visit and at least bimonthly. Lesstable patients or those with recurrent decompensationan be monitored more closely, with checks as frequentlys once or twice per week.

emote Monitoring Using the Internet

he ability to manage device data via Internet sites is a newnd growing field with great promise. Most of the implant-ble cardiac device companies currently offer the ability toiew information over the Internet. However, only certainevices from each company are supported by Internet mon-toring systems. Patients can download device data the nightefore a clinic visit, so the information can be accesseduickly, immediately before the visit. Another useful prac-ice is to monitor the patient between office visits. If theownload shows a problem, the patient can be contacted,nd clinical treatment can be instituted. In some instances,reatment can occur without the patient having to leave theomfort of his or her home. We have found that this capa-ility sometimes allows us to identify the need for a labo-atory test or other test before the patient’s visit with us.emote transmission has also allowed us to monitor patientsho live a great distance from the clinic.Internet access of data also provides many other potential

pportunities for the prevention of heart failure hospitaliza-ion. Our clinic has asked patients who have devices witholume monitoring capability to download device data onceweek to proactively identify clinical problems. Althoughe are awaiting the results of several upcoming trials, it isoped that these data may allow less frequent clinic visits orore information between clinic visits. Additionally, pa-

ients have responded very favorably. They feel they areeing “watched over.” The frequency of downloads mayepend on both the clinic and the severity of the disease. Aatient with New York Heart Association (NYHA) func-ional class II heart failure who has never been hospitalizeday not benefit from frequent examinations of intrathoracic

mpedance, whereas we may request that an NYHA class IV

pt

F

AmfemoOfthw

wma3

tfatt

R

AiCmsvttei

n

Fe

15GGermany and Murray/Device Diagnostics in Outpatient Management of Heart Failure

atient transmit information several times a week to facili-ate care.

ollow-Up Evaluation

mong the best uses of technology allowing remote trans-ission of device data is to decrease the need for such

requent follow-up. Often, patients with heart failure arelderly, cannot drive, or live a significant distance from theanaging clinic. For stable patients, the use of the technol-

gy allows the number of follow-up visits to be decreased.ther patients can then be filtered into the clinic. Follow-up

requency may be determined by specific results found inhe device data. Data points, such as heart rate variability,ave been proposed to help determine the frequency athich patients should be seen.9

Protocols using the data can be extremely helpful asell. We have devised protocols for the use of impedanceonitoring data during a clinic visit (Figure 2) and afterscheduled download from the patient’s home (Figure

igure 2. Use of devices in the office. BNP � B-type natriuretic peptide;xam � examination; HF � heart failure; labs � laboratory tests.

). These protocols provide standardization and utiliza- d

ion for physician extenders as well as those less com-ortable using the information. However, further studiesre needed to determine the utility of such information toitrate drugs remotely and to reduce healthcare utiliza-ion.

eimbursement

lthough the time it takes to access these data is decreas-ng, this extra time may adversely affect a busy clinic.hallenges remain in determining the types of reimburse-ent available for this tailored medical care. A potential

trategy may be to include time and/or complexity of theisit. By looking at rhythm disturbances and volume data,he visit becomes more complex and time consuming. Ifhe data are accessed remotely as part of the routine EPvaluation of the device, this would appropriately bencluded in the normal interrogation charge.

Currently, the physician managing the day-to-day care isot reimbursed for the complex issues associated with HF

Figure 3. Use of device information in the outpatient setting.

isease management. Medicare has recently approved reim-

babty

psmrpc

C

Tphstowsiptw

1

2

3

4

5

6

7

8

9


ursement for remote integration of the device, includingnalysis of the device functioning and thresholds. Specificilling codes, such as those now used for device interroga-ions by EP practices, are still needed for the isolated anal-sis of heart failure device data.

To add to the complexity of reimbursement in the out-atient setting, workflow patterns have to be established. Aseen in the protocol, patients with abnormal volume infor-ation are contacted. If the physician calls the patient di-

ectly, this is coded as a telephone contact. However, a moreractical approach is to have a nurse or physician extenderontact the patient, although this is not reimbursed.

onclusion

echnology has developed incredible ways to help us treatatients. Increasing numbers of patients with heart failureave implantable cardiac devices that can provide us withignificant insights into their management and care. Byapping into this information, we can more effectively treatur patients. However, without creation of easy and simpleays to access and use these data, opportunities to better

erve our patients may be lost. More complex and fascinat-ng devices are on the horizon. Determining ways to incor-orate this into daily clinical practice now will help ease theransition when these devices become standard of careithin the next few years.

. La Rovere MT, Pinna GD, Maestri R, Mortara A, Capomolla S, Febo O,Ferrari R, Franchini M, Gnemmi M, Opasich C, et al. Short-term heartrate variability strongly predicts sudden cardiac death in chronic heart
failure patients. Circulation 2003;107:565–570.
. Adamson PB, Kleckner K, VanHout WL, Srinivasan S, Abraham WT.Cardiac resynchronization therapy improves heart rate variability inpatients with symptomatic heart failure. Circulation 2003;108:266–269.

. Adamson PB, Smith AL, Abraham WT, Kleckner KJ, Stadler RW, ShihA, Rhodes MM, on behalf of the InSync III Model 8042 and AttainOTW Lead Model 4193 Clinical Trial Investigators. Continuous auto-nomic assessment in patients with symptomatic heart failure: prognosticvalue of heart rate variability measured by an implanted cardiac resyn-chronization device. Circulation 2004;110:2389–2394.

. Moss AJ, Zareba W, Hall WJ, Klein H, Wilber DJ, Cannom DS,Daubert JP, Higgins SL, Brown MW, Andrews ML, for the MulticenterAutomatic Defibrillator Implantation Trial II Investigators. Prophylacticimplantation of a defibrillator in patients with myocardial infarction andreduced ejection fraction. N Engl J Med 2002;346:877–883.

. Wilkoff BL, Cook JR, Epstein AE, Greene HL, Hallstrom AP, Hsia H,Kutalek SP, Sharma A, for the Dual Chamber and VVI ImplantableDefibrillator Trial Investigators. Dual-chamber pacing or ventricularbackup pacing in patients with an implantable defibrillator: the DualChamber and VVI Implantable Defibrillator (DAVID) trial. JAMA2002;288:3115–3123.

. Sweeney MO, Hellkamp AS, Ellenbogen KA, Greenspon AJ, FreedmanRA, Lee KL, Lamas GA, for the Mode Selection Trial (MOST) Inves-tigators. Adverse effect of ventricular pacing on heart failure and atrialfibrillation among patients with normal baseline QRS duration in aclinical trial of pacemaker therapy for sinus node dysfunction. Circu-lation 2003;107:2932–2937.

. Yu CM, Wang L, Chau E, Chan RH, Kong SL, Tang MO, ChristensenJ, Stadler RW, Lau CP. Intrathoracic impedance monitoring in patientswith heart failure: correlation with fluid status and feasibility of earlywarning preceding hospitalization. Circulation 2005;112:841–848.

. Magalski A, Adamson P, Gadler F, Böehm M, Steinhaus D, ReynoldsD, Vlach K, Linde C, Cremers B, Sparks B, Bennett T. Continuousambulatory right heart measurements with an implantable hemody-namic monitor. J Card Fail 2002;8:63–70.

. Adamson PB. Continuous heart rate variability from an implanteddevice: a practical guide for clinical use. Congest Heart Fail 2005;11:
327–330.

HUbipupsTAMmippboi

dTfr

aP

r

0d

Integrating Device-Based Monitoring into Clinical Practice:Insights from a Large Heart Failure Clinic

Roy S. Small, MD

Heart failure is a difficult and costly disease to manage in part because the symptomsmay be protean, the physical findings obscure, and the laboratory assessments unre-liable. New implanted physiologic monitors may simplify the care of patients withheart failure, if they can be incorporated into routine clinical practice. Cardiacresynchronization therapy/defibrillators and implantable cardioverter defibrillatorswith continuous intrathoracic impedance monitoring capabilities (OptiVol fluid sta-tus monitoring; Medtronic, Inc., Minneapolis, MN) have recently been introduced andmay provide an early warning of thoracic fluid retention. However, patients who havedevices with this diagnostic capability must be identified, and the device-basedinformation must be accessed systematically, if it is to be used in the disease man-agement process. Ancillary information, such as the Heart Failure ManagementReport that is generated from data stored in Medtronic devices, may facilitaterecognition of disease mechanisms associated with decompensation. The predictivevalue of continuous intrathoracic impedance monitoring with an implantable deviceis still unknown. Thus, therapeutic decisions should be made in conjunction witha clinical assessment. Physicians and other healthcare providers will need tobecome familiar with these devices so they can appreciate their advantages andlimitations. © 2007 Elsevier Inc. All rights reserved. (Am J Cardiol 2007;
99[suppl]:17G–22G)
gpaiahcplstpsimioapt

lurS(wt

eart failure–related expenditures are the single most costlyS government medical expense, accounting for nearly 28illion dollars in direct and indirect costs in 2005.1 Approx-mately 50% of these costs are directly related to the hos-italization of patients with acute decompensated heart fail-re (ADHF).2 Most patients hospitalized with ADHF havereviously been diagnosed with heart failure.3 The readmis-ion rate within 6 months after hospital discharge is 44%.4

he need for frequent rehospitalization for the treatment ofDHF reflects the difficulties in caring for these patients.ost are elderly and have many comorbidities. Manage-ent requires multiple medications, careful education, and

ntensive follow-up to help maintain equilibrium. It may beossible to reduce the frequency of hospitalizations by im-roving outpatient monitoring techniques. The potentialenefits are substantial both in reducing the financial burdenf this disease to society and in improving individual clin-cal outcomes.

Most patients admitted with ADHF have adequate car-iac output and elevated left ventricular filling pressures.5

he clinical and laboratory assessment of patients with heartailure frequently fails to detect the insidious onset of fluidetention, which may then progress to overt vascular con-

Heart Failure Clinic, The Heart Group, Lancaster, Pennsylvania, USA;nd the Heart Failure Service, Lancaster General Hospital, Lancaster,ennsylvania, USA.

Address for reprints: Roy S. Small, MD, The Heart Group, 217 Har-isburg Avenue, Suite 100, Lancaster, Pennsylvania 17603.

CE-mail address: [email protected].


estion prompting hospitalization. Rales and edema may beresent in only a few patients with elevated filling pressures,nd an accurate assessment of jugular venous pressure isnconsistent.6 Up to 40% of patients admitted with ADHFnd severely elevated left ventricular filling pressures mayave chest x-ray findings of no or minimal congestion.7 Theorrelation of plasma B-type natriuretic peptide levels andulmonary capillary wedge pressure is weak and particu-arly unreliable in patients with impaired left ventricularystolic function.8 Although diagnostic models using trans-horacic impedance measurements (impedance cardiogra-hy) have recently been described and may help predictubsequent decompensation,9 their utility has yet to be val-dated in prospective clinical trials. Daily weight measure-ents are helpful and readily obtained but are not a reliable

ndicator of heart failure status.10 Thus, the high frequencyf ADHF hospitalizations is, in part, because of the lack ofn accurate barometer for monitoring patients with a pro-ensity for fluid retention. Intrathoracic impedance moni-oring may be such a tool.

Intrathoracic impedance is inversely correlated witheft ventricular filling pressure.11 This principle has beensed to design the OptiVol fluid status monitoring algo-ithm,12 which has been incorporated into the InSyncentry cardiac resynchronization therapy/defibrillatorCRT-D) device (Medtronic, Inc., Minneapolis, MN)ith standard interrogation capabilities. The newer Vir-

uoso implantable cardioverter defibrillator and Concerto
RT-D (Medtronic, Inc.) have additional wireless capa-
www.AJConline.org

bmTiwguducswpida

IH

Rubacriw�i

vpcfi

eFdtfSgn

tmhdgsewiiaIabattarbmfaarhp

dioSigh


ilities, which can be programmed to automatically trans-it device diagnostics at prespecified routine intervals.hese novel devices require the implementation of clin-

cal procedures to deal with programmed device alerts asell as unsolicited data obtained during routine interro-ations. If device-derived diagnostic information is notsed, then the additional expense associated with theseevices is not justified. Furthermore, the opportunity tose remote physiologic monitoring to reduce hospitalosts and improve patient outcomes will have beenquandered. Clinics that have advanced practice nursesith training in heart failure are ideally suited to incor-orate device-based diagnostics into routine practice. Thentegration of heart failure and electrophysiology (EP)evice follow-up facilitates the use of device-based di-gnostic information.

ncorporation of Intrathoracic Monitoring in theeart Failure Clinic

outine monitoring of intrathoracic impedance may be aseful tool for monitoring patients with chronic heart failurey providing an early warning concerning fluid retentionnd impending decompensation. The Heart Group (Lan-aster, PA) implanted its first CRT-D device with intratho-acic impedance monitoring at Lancaster General Hospitaln December 2004 and currently monitors �500 patientsith a CRT-D implant in the heart failure clinic, including200 with devices featuring intrathoracic impedance–mon-

toring capabilities.The practice model in The Heart Group (a large pri-

ate heart failure clinic) is a physician-directed and nurseractitioner–implemented model. The heart failure clinicoordinates its services with the EP clinic for deviceollow-up. Information derived from device diagnostics

Figure 1. Downloading the device diagnostics using the Car

s shared with both the heart failure and EP services to c

nsure that the data are analyzed and used appropriately.or example, if impedance data indicating a potentialecompensation are discovered during device interroga-ion in EP, the information is forwarded to the heartailure group for analysis and therapeutic intervention.imilarly, heart failure clinicians and members of the EProup meet regularly to discuss the management of CRTonresponders.

Simple and seamless access to device-based diagnos-ic data is a key to optimizing clinical utility. Although itay seem self-evident, if a patient is not identified as

aving a device with monitoring capabilities, or if theevice is not interrogated, then the opportunity to inte-rate the device-based information into the clinical deci-ion-making process will be lost. Therefore, we labelach patient’s medical record in our heart failure clinicith an identifying sticker indicating the presence of an

mpedance-monitoring device. When the patient arrivesn the heart failure clinic, the device is interrogated using

data access tool—the CardioSight Reader (Medtronic,nc.)—as shown in Figure 1. This data reader facilitatesccess to diagnostic information and can be used safelyy trained technical staff. It has read-only capabilitiesnd cannot be used to alter the programmed settings ofhe device. The telemetry antenna of the reader is posi-ioned over the device, which then transmits the data over

standard phone line to a secure server. A detailedeport, the Heart Failure Management Report, is faxedack to the clinic for use during the patient’s visit in �10inutes. This report includes 90-day trend data for the

ollowing variables: intrathoracic impedance data (aver-ge daily impedance and the OptiVol fluid index), atrialrrhythmia profile, ventricular response during atrial ar-hythmias, patient activity index, nighttime heart rate,eart rate variability, and percent atrial and ventricularacing. These trends are aligned across time, which fa-

t Reader. (Courtesy of Medtronic, Inc., Minneapolis, MN)

ilitates recognition of their interrelations. Device interro-

gwpdRr

IP

FtHmha

tcisnclcra

aRcsnpfaNw

paoORacpcToaltt

eart fa

19GSmall/Integrating Device-Based Monitoring into Clinical Practice

ation using this service is quick and easy and is associatedith minimal, if any, cost. The printed report is included in theatient record so that it is available to the clinician for reviewuring the clinic visit. We put the Heart Failure Managementeport in the same category as a vital sign, and as such, it is a

equisite part of a follow-up visit.

nsights in Data Interpretation in Daily Clinicalractice

igure 2 summarizes a case in which a patient was hospi-alized and treated for anemia and gastrointestinal bleeding.ad his device been interrogated earlier, the hospitalizationay have been aborted because the intrathoracic impedance

ad been decreasing 2 weeks before the patient’s eventualdmission.

The Heart Failure Management Report provides addi-ional information that is often beneficial in treating theseomplex cases. The therapeutic benefit derived from CRTs because of the coordination of right and left ventricularystole. Clearly, the loss of consistent ventricular pacingegates the benefits of the therapy. Figure 3 outlines aase in which inappropriate programming resulted in theoss of consistent biventricular pacing and eventual de-ompensation. The etiology of this episode of ADHF iseadily apparent from analysis of the Heart Failure Man-

Figure 2. Patient case 1. CHF � congestive h

gement Report. u

In addition to atrial arrhythmia detection and percenttrial and ventricular pacing, the Heart Failure Managementeport provides a graphic display of patient activity. Thisan be used as an objective measurement of the patient’sense of well-being. Heart rate variability is a marker ofeurohormonal activation in patients with heart failure. Im-aired heart rate variability in patients with chronic heartailure using continuous measurements from a CRT-D isssociated with high mortality and risk for hospitalization.13

ighttime heart rate and patient activity index also correlateith prognosis but may be less sensitive.When interpreting OptiVol fluid status data, it is im-

ortant to consider factors regarding operation of thelgorithm. It is a common misconception that the severityf fluid retention is proportional to the amplitude of theptiVol fluid index after a threshold crossing. It is not.ather, the OptiVol fluid index only indicates that theverage daily impedance values have deviated signifi-antly from the established baseline for an extendederiod. In essence, an “event” has occurred that mayorrelate with the development of vascular congestion.he degree of fluid retention can be assessed by analysisf the raw impedance data. An increased negative aver-ge daily impedance deflection from the established base-ine indicates a more significant episode of fluid reten-ion. After the OptiVol fluid threshold has been crossed,he raw daily impedance data may continue to provide

ilure; EF � left ventricular ejection fraction.

seful information and can be monitored to track im-

pt

cAsttd6ntapocpctp(u

ifmwdp3

brmrdrsra

C

DnpdmfimijtoaApi

Fea


rovement, even while the OptiVol fluid index is abovehe physician-established threshold (Figure 4).

The impedance data should be interpreted within theontext of the patient history and physical examination.lthough each patient serves as his or her own control, the

ensitivity and specificity of continuously monitored in-rathoracic impedance data from an implantable device inhe general heart failure population is still unknown. Theefault OptiVol fluid threshold is set at a nominal value of0 �-days. However, it may be reprogrammed to an alter-ative level. Whether a different threshold setting will affecthe predictive value of the data is unknown. Preliminarynalysis of our own treatment population indicates a false-ositive rate (ie, an OptiVol fluid threshold crossing thatccurs without the development of signs or symptoms ofongestion) of approximately 0.5 episodes per year peratient when set at the default threshold. Therefore, clini-al evaluation of the patient should be performed beforeherapeutic adjustments based on impedance data. False-ositive findings may occur for clinically obvious reasonspouch hematoma, pneumonia, or pocket revisions) or fornapparent causes (Figure 5).

Patients who respond to CRT-D as evidenced by anmprovement in their New York Heart Association (NYHA)unctional class will often require less diuretic therapy toaintain homeostasis. Figure 6 depicts the course of aoman who crossed the OptiVol threshold when heriuretic dose was appropriately decreased after CRT-D im-lantation. The intrathoracic impedance feature initialized

igure 3. Patient case 2. A � atrial; BNP � B-type natruretic peptide; CHchocardiography; EF � left ventricular ejection fraction; P � progratrial-ventricular; V � ventricular.

4 days after implant while the patient was dehydrated m

ecause of excessive diuretics. (Note that the impedanceeference value is not established until approximately 1onth after implant to allow wound healing.) As the patient

eturned to normal volume status after adjustment of theiuretic dose, the change in her thoracic fluid volume waseflected by a decrease in intrathoracic impedance and sub-equent threshold crossing. Thus, this case illustrates why aeflex response to an OptiVol threshold crossing is notppropriate.

onclusion

evice-based intrathoracic impedance monitoring shouldot be considered an “easy button” for the management ofatients with heart failure. Identification of patients withevices with diagnostic capabilities coupled with imple-entation of efficient systems to capture the available in-

ormation remains a challenge. The predictive value ofntrathoracic impedance data derived during continuousonitoring via an implanted device is still unknown. Thus,

ntrathoracic impedance data should be interpreted in con-unction with clinical assessment of the patient. Althoughhe derived OptiVol fluid index may simplify interpretationf the impedance data, the raw thoracic impedance data aremore useful measure of the patient’s fluid volume status.nalysis of the entire Heart Failure Management Reportrovides additional information that can be incorporatednto the clinical assessment of patients and facilitate their

ngestive heart failure; CRT � cardiac resynchronization therapy; echo �� paced atrial-ventricular; RV � right ventricular; SAV � sensed

F � com; PAV

anagement. The report should be available for review

deh

mb

Ffl f absol

FP

21GSmall/Integrating Device-Based Monitoring into Clinical Practice

uring the patient visit. Appropriate interventions requirexperience with using the specific device diagnostics. A

igure 4. Impedance change after device revision. Improved clinical statuuid index helps quantify the change in impedance but is not a measure o

igure 5. Patient case 3. BNP � B-type natruretic peptide; CHF � conges� program.

eart failure clinic is ideally suited to this task. However, t

ethods must be developed to analyze and react to device-ased data, which may be monitored remotely with routine

ected in rising impedance despite persistent index crossing. The OptiVolute impedance. P � program.

rt failure; exam � examination; NYHA � New York Heart Association;

s is refl

tive hea

ransmission to the clinic.

pifWodmd

1

1

1

1

Fv x � ma


With the proper recognition of their limitations, im-lanted physiologic monitors can become important toolsn disease management. The individual clinical benefitsor routine heart failure management are substantial.

hether or not these devices may help abort an episodef ADHF requires prospective study. However, if theseevices can be successfully incorporated into diseaseanagement algorithms, then the financial savings to a

istressed medical system could be enormous.

1. American Heart Association. Heart Disease and Stroke Statistics—2005 Update. Dallas, TX: American Heart Association, 2005.

2. American Heart Association. Heart Disease and Stroke Statisitcs—2004 Update. Dallas, TX: American Heart Association, 2004.

3. Fonarow GC. The Acute Decompensated Heart Failure National Reg-istry (ADHERE): opportunities to improve care of patients hospital-ized with acute decompensated heart failure. Rev Cardiovasc Med2003;4(suppl 7):S21–S30.

4. Krumholz HM, Parent EM, Tu N, Vaccarino V, Wang Y, Radford MJ,Hennen J. Readmission after hospitalization for congestive heart fail-ure among medicare beneficiaries. Arch Intern Med 1997;157:99–104.

5. Publication Committee for the VMAC Investigators. Intravenous ne-siritide vs nitroglycerin for treatment of decompensated congestiveheart failure: a randomized controlled trial. JAMA 2002;287:1531–1540.

6. Stevenson LW, Perloff JK. The limited reliability of physical signs forestimating hemodynamics in chronic heart failure. JAMA 1989;261:

igure 6. Patient case 4. Avg � average; AT � atrial tachyarrhythmia;entricular ejection fraction; NYHA � New York Heart Association; Ma

884–888.

7. Mahdyoon H, Klein R, Eyger W, Lakier JB, Chakko SC, GheorghiadeM. Radiographic pulmonary congestion in end-stage congestive heartfailure. Am J Cardiol 1989;63:625–627.

8. Dokainish H, Zoghbi WA, Lakkis NM, Al-Bakshy F, Dhir M, Qui-nones MA, Nagueh SF. Optimal noninvasive assessment of left ven-tricular filling pressures. Circulation 2004;109:2432–2439.

9. Packer M, Abraham WT, Mehra M, Yancy C, Lawless C, Mitchell J,Smart F, Bijou R, O’Connor C, Massie B. Utility of impedancecardiography for the identification of short-term risk of clinical de-compensation in stable patients with chronic heart failure. J Am CollCardiol 2006;47:2245–2252.

0. Lewin J, Ledwidge M, O’Loughlin C, McNally C, McDonald K.Clinical deterioration in established heart failure: what is the value ofBNP and weight gain in aiding diagnosis? Eur J Heart Fail 2005;7:953–957.

1. Wang L, Lahtinen S, Lentz I, Rakow N, Kaszas C, Ruetz L, Stylos L,Olson WH. Feasibility of using an implantable system to measurethoracic congestion in an ambulatory chronic heart failure caninemodel. Pacing Clin Electrophysiol 2005;28:404–411.

2. Yu CM, Wang L, Chau E, Chan RH, Kong SL, Tang MO, ChristensenJ, Stadler RW, Lau CP. Intrathoracic impedance monitoring in patientswith heart failure. Circulation 2005;112:841–848.

3. Adamson PB, Smith AL, Abraham WT, Kleckner KJ, Stadler RW,Shih A, Rhodes MM, on behalf of the InSync III Model 8042 andAttain OTW Lead Model 4193 Clinical Trial Investigators. Con-tinuous autonomic assessment in patients with symptomatic heartfailure: prognostic value of heart rate variability measured by animplanted cardiac resynchronization device. Circulation 2004;110:

atrial fibrillation; CRT � cardiac resynchronization therapy; EF � leftximum; V � ventricular.

AF �

2389 –2394.

VpnterraIpttifluir

Js

V

0d

Device Monitoring of Intrathoracic Impedance: ClinicalObservations from a Patient Registry

John Andriulli, DO

A distinct advantage of implantable device diagnostics is that the data may be madeavailable on a continuous basis, rather than at sporadic intervals associated withclinical testing. Recently, intrathoracic impedance monitoring has also become avail-able in some implantable devices as an index of congestion and thoracic fluidaccumulation secondary to decompensated heart failure. Despite the potential ad-vantages, new advances in implantable device diagnostic technology also pose im-portant questions regarding their clinical application. To examine these questions andto generate hypotheses, the relation between daily intrathoracic impedance measure-ments and other physical measurements or comorbidities was studied. A retrospectivereview was performed of 25 patients who previously underwent implantation of acardiac resynchronization therapy/defibrillator device with the capability to contin-uously monitor intrathoracic impedance. This limited scope analysis demonstratedthat daily measurement of intrathoracic impedances might reveal the intrinsic rela-tions between heart failure decompensation and the onset of atrial and ventriculararrhythmias. Abnormal patterns of intrathoracic impedance that has increased andplateaued after implant may indicate worsening heart failure. The severity of con-gestive heart failure at the time of interrogation may correlate with device-basedimpedance measurements. Weight, sex, and body index may have a limited impact onimpedance, and the lowest impedances may be seen in older patients. Patients withsevere pulmonary disease may present with unique daily impedance profiles.Finally, daily impedance may have unpredictable relations with other clinicalmarkers of heart failure. In summary, intrathoracic impedance represents aclinically useful diagnostic tool that can increase our understanding of a dynamicdisease state on an individual patient basis. © 2007 Elsevier Inc. All rights
reserved. (Am J Cardiol 2007;99[suppl]:23G–28G)
trdiDcw

pqubvswccisdbul

arious implantable devices, including pacemakers, im-lantable cardioverter defibrillators, and cardiac resynchro-ization therapy (CRT) devices, now contain the capabilityo track several physiologically relevant diagnostic param-ters, including heart rate, heart rate variability, respirationate, atrial tachyarrhythmia (AT) and ventricular tachyar-hythmia (VT) recurrence and duration, symptom markers,nd patient activity, in addition to traditional therapies.ndeed, implantable devices are now either available1 orroposed2,3 that serve solely to provide diagnostic informa-ion without direct therapeutic capability. Recently, in-rathoracic impedance has also become available in somemplantable devices as an index of congestion and thoracicuid accumulation secondary to decompensated heart fail-re.4 A distinct advantage of implantable device diagnosticss that the data may be made available on a continuous basis,ather than at sporadic intervals associated with clinical

University of Medicine and Dentistry of New Jersey, Robert Woodohnson Medical School, Cooper University Hospital, Camden, New Jer-ey, USA.

Address for reprints: John Andriulli, DO, 1020 Laurel Road, Suite 102,oorhees, New Jersey 08043.

E-mail address: [email protected].


esting. Such diligent and automatic recording of clinicallyelevant physiologic information can provide a powerfuliagnostic tool, especially when assessing sudden changesn clinical status. For example, the Medtronic Impedanceiagnostics in Heart Failure (MIDHeFT) trial indicated that

hanges in intrathoracic impedance may provide advancedarning of impending heart failure hospitalization.4

Despite the potential advantages, new advances in im-lantable device diagnostic technology also pose importantuestions regarding their clinical application. In the partic-lar case of intrathoracic impedance, clinicians are justeginning to interpret changes in these measures underarious clinical scenarios. As such, my colleagues and Iought to collect data that would provide insight intohether (1) changes in intrathoracic impedance are asso-

iated with arrhythmias or vice versa; (2) impedancehanges can be characterized over time; (3) the dailyntrathoracic impedance measurements correlate with theeverity of congestive heart failure (CHF); (4) breathingisorders affect daily intrathoracic impedance values; (5)ody mass, weight, sex, or age alter daily impedance val-es; and (6) the association between impedance and bio-ogic or clinical markers of CHF could be determined.

To examine these questions and to generate hypotheses

www.AJConline.org

ostoic(ibEiwpatcdt

IA

o

abmotnchcaAlttcpa

dtaVI

Fa daily i


n the relation between daily intrathoracic impedance mea-urements and other physical measurements or comorbidi-ies, my colleagues and I performed a retrospective reviewf 25 patients from our center who previously underwentmplantation with a CRT/defibrillator (CRT-D) device with theapability to continuously monitor intrathoracic impedanceInSync Sentry; Medtronic, Inc., Minneapolis, MN). This reg-stry protocol was approved by the local institutional reviewoard and adhered to published guidelines for clinical research.ach subject provided written informed consent before partic-

pation. The registry consisted of a single clinic visit duringhich patients were physically assessed. Each subject com-leted a short interview regarding their heart failure symptomsnd medical history. Subsequently, New York Heart Associa-ion (NYHA) functional class was documented, the medicalhart since the time of implantation was reviewed, and deviceiagnostic data were downloaded. The average implant dura-ion for the population was 283 � 124 days.

s Intrathoracic Impedance Associated withrrhythmias or Vice Versa?Intrathoracic impedance and VT: The increased risk

igure 1. Device reported daily impedance and thoracic fluid index from 1fter implant. A statistically significant association was observed between

f VT in the CHF population has been well described.5 In e

ddition, the concept of mechanical–electrical feedback haseen demonstrated in multiple animal6–9 and computerodels10,11 and represents a potential mechanism for the

bserved clinical relation between CHF and VT. Despitehis previous work, the dynamic relation between pulmo-ary congestion/decompensated heart failure and VT in thelinical environment is not well understood. It could beypothesized that pulmonary congestion secondary to in-reased atrial or ventricular loading or “stretching” may bessociated with an increased incidence of arrhythmias.lternatively, AT or VT may also alter loading conditions

eading to increased pulmonary congestion.12,13 Together,hese mechanisms could initiate a vicious circle of intermit-ent or continuous arrhythmias and symptomatic pulmonaryongestion. Monitoring both fluid status and arrhythmias inatients with heart failure may enable better management ofrrhythmias in this population.

Figure 1 shows device-recorded intrathoracic impedanceata recorded from an 80-year-old man with NYHA func-ional class III CHF and a history of ischemic cardiomyop-thy and VT. The subject underwent several unsuccessfulT ablations and was subsequently implanted with an

nSync Sentry device. The device logged a total of 80

with multiple ventricular tachyarrhythmia (VT) episodes during 320 daysmpedance and VT in this patient. VF � ventricular fibrillation.

patient

pisodes of VT in the 320 days after implant. Daily

iuaicdwwdaw�b(tpdt

eJaottietpadATti

CI

Rstfipdn0itTgm

svmattCcblTitvi

DI

Iasiinicdvcv

otd(plihimpppdg

DI

W

25GAndriulli/Device Monitoring of Intrathoracic Impedance

ntrathoracic impedance, ventricular tachycardia/ventric-lar fibrillation episodes, and device-mediated VT shocksre presented in Figure 1. Note that each discrete decreasen the daily impedance (or increase in the fluid index)orresponded with storms of VT and �1 VT shock. Theaily impedance was statistically compared between daysith and without VT shock and between days with andithout VT episodes, using data derived from the storedevice logs (rank sum test). In this patient, the daily imped-nce was significantly lower on days in which a VT shockas delivered versus no shock (54.1 � 3.4 � vs 60.2 � 4.1; p �0.05). There was a significant negative correlation

etween the average VT cycle length and daily impedancer � 0.62, p �0.05). Thus, an association between poten-ially life-threatening VT and decreased intrathoracic im-edance was observed. Further investigation is required toetermine whether the management of pulmonary conges-ion can diminish the incidence of VT.

Intrathoracic impedance and AT: The “chicken or thegg” relation between AT and CHF has been well described.ust as CHF may lead to AT, atrial fibrillation and associ-ted high ventricular rates can induce ventricular cardiomy-pathies.14 My colleagues and I stratified subjects based onhe presence or absence of any device-documented AT inhe 90 days before interrogation. The data in Figure 2ndicate that 7 of 8 patients with a prior documented ATpisode had 7-day median daily impedances �70 �. Fur-hermore, after pooling the data, the mean impedance fromatients with AT was significantly lower than those withoutn AT episode 90 days before interrogation. Based on theseata, it is reasonable to hypothesize that a recent history ofT could result in lower daily impedance measurements.hus, routine assessment of intrathoracic impedance in pa-

ients with a history of AT may help to better understandndividual patterns of AT recurrence.

an Changes in Impedance over Time from Devicemplant Be Characterized?

eported data from the MIDHeFT trial suggested a con-istent nonlinear increase in the mean daily impedancehat began at implantation and plateaued several monthsrom the time of implantation.4 A similar time-dependentncrease in impedance was seen in our registry studyatients (Figure 3). When each patient’s 7-day medianaily impedance measurement was plotted against theumber of days since implant, a positive correlation (p �.06) was noted up to approximately 250 days aftermplant. Subsequent to this time point, no relation be-ween impedance and time from implant was observed.he clinical experience of my colleagues and I has shownenerally lower daily impedance values in patients with
ore advanced heart failure. For example, Figure 4 s
hows data from a patient with very low daily impedancealues immediately after implant. The patient was read-itted to the hospital for worsening CHF symptoms

pproximately 60 days after implant. During hospitaliza-ion, CRT therapy settings were adjusted and diureticherapy was increased. After 5 months of continuousRT therapy, the basal impedance increased signifi-antly. Thus, the raw intrathoracic impedance data maye indicative of the patient having been quite hypervo-emic at the time of implant, despite standard CHF care.his case points to the key clinical utility of the raw daily

mpedance because the fluid index was not helpful forhis patient. The clinical utility of raw basal impedancealues represents an interesting topic for future clinicalnvestigations.

o Breathing Disorders Affect Dailympedance Values?

ntrathoracic impedance measurements taken at the time oftrial fibrillation cardioversion were recently reported to beignificantly lower in patients with a past history of smok-ng, and there was also a trend for higher impedance valuesn those patients diagnosed with chronic obstructive pulmo-ary disease (COPD).15 Because device measurements ofntrathoracic impedance include a significant pulmonaryomponent, it was hypothesized that patients with breathingisorders secondary to COPD with larger residual lungolumes and, hence, less fluid between the right ventricularoil and CRT-D device might have higher daily impedancealues.

A total of 7 patients included in the registry had previ-usly been diagnosed with COPD. Interestingly, these pa-ients encompassed the lowest and the highest 7-day medianaily impedance measurements at the time of interrogationFigure 5). However, a distinct relation between daily im-edance values and history of COPD was not found in thisimited data set. Of note, the 2 patients with COPD present-ng with the lowest daily impedance values also had aistory of AT within the 90 days before interrogation. Thus,t may be hypothesized that patients with COPD have dra-atic “swings” in impedance when other comorbidities are

resent. Additionally, a previous report on the use of im-edance cardiography for the measurement of cardiac out-ut in COPD patients suggested that the estimation of car-iac output via impedance was less accurate in patients withreater disease severity.16

o Body Mass, Weight, Sex, or Age Alter Baselinempedance Values?

eight has previously been shown to have a relatively high
pecificity, but low sensitivity, to worsening heart failure.17

Tfstatsi

ws(aoml

WBH

TmRnps1Iaas

Fti median

Fi


his may be, in part, because patients gain or lose weightor reasons other than decompensation, and that decompen-ation is occasionally unaccompanied by weight gain. Thus,he relation between body weight and intrathoracic imped-nce may not be expected to behave uniformly throughouthe CHF population. Accordingly, the registry data demon-trated no clear relation between weight and raw dailympedance.

In addition, the effects of age on basal impedance are notell understood. My colleagues and I observed a marginally

ignificant inverse correlation between age and impedancep � 0.07) in this patient population. However, because agelso may be correlated with CHF severity, the impact of agen intrathoracic impedance measurements is uncertain anday be better elucidated from analyses of forthcoming

igure 2. Association between atrial tachyarrhythmia (AT) and intrathoraherapy/defibrillator device. (A) Raw impedance values for patients with (n A. Patients with a recent history of AT had a significantly lower 7-day

igure 3. Time dependence of daily intrathoracic impedance measurement.mpedance for days 1–250 but not for �250 days since device implant.

arger clinical databases.

hat Is the Association Between Impedance andiologic or Clinical Markers of Congestiveeart Failure?

he association between impedance and biologic clinicalarkers of CHF has not yet been clinically evaluated.ecently, Vollman and colleagues18 reported a significantegative correlation between changes in intrathoracic im-edance and changes in plasma levels of amino-terminalegment of B-type natriuretic peptide during approximately2 weeks in a series of 52 patients implanted with thenSync Sentry device. However, other data suggest thatbsolute plasma B-type natriuretic peptide levels may not beccurate predictors of variations in heart failure decompen-ation status.

pedance in a population of 25 patients with a cardiac resynchronizationand without (triangles) AT. (B) Box and whiskers representation of datadaily impedance than patients in sinus rhythm.

tically significant correlation was observed between time and intrathoracic

cic imcircles)

A statis

Currently, my colleagues and I are performing a prospec-

ttBireTp

C

Oatsrot

tucWoonpaEpaia

D

FTbk was op

27GAndriulli/Device Monitoring of Intrathoracic Impedance

ive multicenter clinical evaluation of the relations amonghoracic impedance, weight, 6-minute hall walk, plasma-type natriuretic peptide levels, and other secondary clin-

cal parameters. The hypothesis is that an increase in tho-acic impedance will also correlate with improvement inxercise capacity and other clinical markers of heart failure.he trial will enroll 100 patients and monitor them for aeriod of 12 months.

onclusion

bservations obtained through daily intrathoracic imped-nce monitoring have led to multiple novel hypotheses onhe nature of decompensated heart failure. Daily mea-urement of intrathoracic impedance may reveal intrinsicelations between heart failure decompensation and thenset of AT and ventricular arrhythmias. Abnormal pat-

igure 4. Example of daily intrathoracic impedance and fluid index trends ohe very low value of impedance at implant and lack of impedance increasey CRT (ie, the patient was “wet” at implant). Such an impedance patterney diagnostic information rather than the fluid index. In mid May, CRT

erns of thoracic impedance that has increased and pla- a

eaued after implant may indicate worsening heart fail-re. The severity of CHF at the time of interrogation mayorrelate with device-based impedance measurements.eight, sex, and body index may have a limited impact

n impedance, and the lowest impedances may be seen inlder patients. In addition, patients with severe pulmo-ary disease may present with unique daily impedancerofiles. Finally, daily impedance may have unpredict-ble relations with other clinical markers of heart failure.ach of these observations should be the subject of futurerospective evaluation. However, intrathoracic imped-nce represents a clinically useful diagnostic tool that canncrease our understanding of a dynamic disease state onn individual patient basis.

Acknowledgment: The author thanks James Coles,oug Hettrick, and Sara Overgaard of Medtronic, Inc., for

ervolemic patient responding to cardiac resynchronization therapy (CRT).iately after implant may indicate pervasive hypervolemia that was rectifiedplant is somewhat atypical. In this case, the daily impedance provided thetimized and diuretics were prescribed.

f a hypimmed

after im

ssistance with the data analysis.

1

1

1

1

1

1

1

1

1

F(oa


1. Farwell DJ, Freemantle N, Sulke N. The clinical impact of implantableloop recorders in patients with syncope. Eur Heart J 2006;27:351–356.

2. Adamson PB, Magalski A, Braunschweig F, Bohm M, Reynolds D,Steinhaus D, Luby A, Linde C, Ryden L, Cremers B, Takle T, Bennett T.Ongoing right ventricular hemodynamics in heart failure: clinical valueof measurements derived from an implantable monitoring system.J Am Coll Cardiol 2003;41:565–571.

3. Ritzema-Carter JL, Smyth D, Troughton RW, Crozier IG, Melton IC,Richards AM, Eigler N, Whiting J, Kar S, Krum H, Abraham WT.Images in cardiovascular medicine: dynamic myocardial ischemiacaused by circumflex artery stenosis detected by a new implantable leftatrial pressure monitoring device. Circulation 2006;113:e705–e706.

4. Yu CM, Wang L, Chau E, Chan RH, Kong SL, Tang MO, ChristensenJ, Stadler RW, Lau CP. Intrathoracic impedance monitoring in patientswith heart failure: correlation with fluid status and feasibility of earlywarning preceding hospitalization. Circulation 2005;112:841–848.

5. Bigger JT Jr, Fleiss JL, Kleiger R, Miller JP, Rolnitzky LM, Therelationships among ventricular arrhythmias, left ventricular dysfunc-tion, and mortality in the 2 years after myocardial infarction. Circu-lation 1984;69:250–258.

6. Lab MJ. Mechanically dependent changes in action potentials recordedfrom the intact frog ventricle. Circ Res 1978;42:519–528.

7. Franz MR, Cima R, Wang D, Profitt D, Kurz R. Electrophysiologicaleffects of myocardial stretch and mechanical determinants of stretch-activated arrhythmias. Circulation 1992;86:968–978.

8. Hansen DE, Craig CS, Hondeghem LM. Stretch-induced arrhythmiasin the isolated canine ventricle: evidence for the importance of mecha-noelectrical feedback. Circulation 1990;81:1094–1105.

9. Parker KK, Lavelle JA, Taylor LK, Wang Z, Hansen DE. Stretch-induced ventricular arrhythmias during acute ischemia and reperfu-sion. J Appl Physiol 2004;97:377–383.

0. Healy SN, McCulloch AD. An ionic model of stretch-activated andstretch-modulated currents in rabbit ventricular myocytes. Europace

igure 5. Association between daily intrathoracic impedance and chronic obcircles) and without (triangles) COPD. (B) Box and whiskers representatr very low intrathoracic impedance independent of time since implant. Thund warrants additional study.

2005;7(suppl 2):128–134.

1. Kohl P, Hunter P, Noble D. Stretch-induced changes in heart rate andrhythm: clinical observations, experiments and mathematical models.Prog Biophys Mol Biol 1999;71:91–138.

2. Damiano RJ Jr, Tripp HF Jr, Asano T, Small KW, Jones RH, Lowe JE.Left ventricular dysfunction and dilatation resulting from chronicsupraventricular tachycardia. J Thorac Cardiovasc Surg 1987;94:135–143.

3. Wang TJ, Larson MG, Levy D, Vasan RS, Leip EP, Wolf PA,D’Agostino RB, Murabito JM, Kannel WB, Benjamin EJ. Temporalrelations of atrial fibrillation and congestive heart failure and their jointinfluence on mortality: the Framingham Heart Study. Circulation2003;107:2920–2925.

4. Peters KG, Kienzle MG. Severe cardiomyopathy due to chronic rap-idly conducted atrial fibrillation: complete recovery after restoration ofsinus rhythm. Am J Med 1988;85:242–244.

5. Fumagalli S, Boni N, Padeletti M, Gori F, Boncinelli L, Valoti P,Baldasseroni S, Di Bari M, Masotti G, Padeletti L, Barold S, Mar-chionni N. Determinants of thoracic electrical impedance in externalelectrical cardioversion of atrial fibrillation. Am J Cardiol 2006;98:82–87.

6. Bougault V, Lonsdorfer-Wolf E, Charloux A, Richard R, Geny B,Oswald-Mammosser M. Does thoracic bioimpedance accurately de-termine cardiac output in COPD patients during maximal or intermit-tent exercise? Chest 2005;127:1122–1131.

7. Lewin J, Ledwidge M, O’Loughlin C, McNally C, McDonald K.Clinical deterioration in established heart failure: what is the value ofBNP and weight gain in aiding diagnosis? Eur J Heart Fail 2005;7:953–957.

8. Vollman D, Luthje L, Drescher T, Busse S, Zenker D, Schott P,Unterberg-Buchwald C. Device-based monitoring of intrathoracic im-pedance to alert patients with chronic heart failure of cardiac decom-pensation: algorithm performance and correlation with NT-ProBNP

e pulmonary disease (COPD). (A) Raw impedance values for patients withata in A. Patients with COPD (�COPD) tended to have either very high

elation between lung disease and intrathoracic impedance may be complex

structivion of ds, the r

[abstract]. Heart Rhythm 2006;3:S330.

Hipyd�tfep

panmfsipfte(l

pntmtafph

2

0d

Use of Device Diagnostics as an Educational Tool to ImprovePatient Adherence

Lisa Rathman, MSN, CRNP

Implantable cardiac devices are increasingly being used to treat heart failure and mayprovide an opportunity to more effectively monitor a patient’s clinical status andprevent hospitalization by detecting subclinical congestion before progression to acutedecompensated heart failure. This article reviews the use of device diagnostics inclinical practice and discusses their utility as tools in heart failure disease manage-ment, with a focus on the use of a new intrathoracic impedance diagnostic as aneducational tool to improve patient adherence. © 2007 Elsevier Inc. All rights
reserved. (Am J Cardiol 2007;99[suppl]:29G–33G)
ttaspsupcddhi

RA

Ptcpmtstfddsrtnro

hoo

eart failure is a common disorder associated with signif-cant morbidity and mortality.1 It affects nearly 5 millioneople in the United States, with 550,000 new cases eachear.2 Heart failure is the primary or contributing cause ofeath in 286,700 patients per year and accounts for1 million hospitalizations annually.2 Despite advances in

reatment, patients with heart failure remain at high risk forrequent hospitalization, most commonly because of exac-rbation of chronic heart failure.3 Within 6 months of hos-ital discharge, �50% of patients are readmitted.4

Most heart failure–related hospital readmissions are fromreventable causes, such as excess dietary sodium intakend/or medication nonadherence.5 Vinson et al6 found thatearly 50% of readmissions for heart failure were caused byedication or dietary nonadherence. In a study of 585 heart

ailure admissions, Bennett et al7 demonstrated that excessodium intake leading to volume overload was the precip-tant in 59% of admissions. Additionally, as many as 20% ofatients fail to notify their healthcare provider when heartailure symptoms reoccur, until the symptoms are so severehat they require hospitalization.8 Other reasons cited forarly readmission for acute decompensated heart failureADHF) include inadequate discharge planning or fol-ow-up and a failed social support system.9

Are patients really to blame for nonadherence to theirrescribed heart failure treatment regimen? Perhaps they doot clearly understand why it is imperative that they followheir medication, dietary, and symptom management regi-ens. In a study of 113 patients with heart failure referred

o an outpatient heart failure clinic, Ni et al10 found thatlthough all of the patients had a lengthy history of heartailure and were observed by a cardiologist, nearly 40% ofatients reported that they knew little or nothing about theireart failure. In addition, 40% of patients did not recognize

The Heart Group, Lancaster, Pennsylvania, USA.Address for reprints: Lisa Rathman, MSN, CRNP, The Heart Group,

17 Harrisburg Avenue, Suite 100, Lancaster, Pennsylvania 17603.
eE-mail address: [email protected].

hat weighing themselves daily was important. Only onehird of the patients in the study reported that they alwaysvoided salty foods, and 36% of the patients thought theyhould drink a lot of fluids.10 Perhaps patients are not beingroperly instructed on how to monitor their heart failureymptoms. Data from the Acute Decompensated Heart Fail-re National Registry (ADHERE) suggest that only 35% ofatients discharged after a heart failure hospitalization re-eive appropriate discharge instructions.11 The lack of un-erstanding on the part of patients with heart failure of theirisease is associated with symptom exacerbation and repeatospitalizations. Improved patient education and monitor-ng may be a way to improve current outcomes.

ole of the Heart Failure Nurse in an Integratedpproach to Heart Failure Management

atient education is key to decreasing repeat hospitaliza-ions and encouraging patients to make appropriate lifestylehanges. Heart failure nurses play a pivotal role in teachingatients the importance of following their heart failure treat-ent regimens and in making lifestyle changes. In a study

hat examined the effects of a 1-hour, one-on-one teachingession with a nurse educator at discharge compared withhe standard discharge process, Koelling and colleagues12

ound that a patient-focused heart failure education programelivered at hospital discharge led to a 35% reduction ineath or repeat hospitalization at 180 days compared withtandard discharge teaching. Furthermore, there was a 51%eduction in the need for repeat hospitalization in the grouphat received the directed patient education. In the study, theurse reviewed the causes of heart failure and treatmentegimens, including medications, lifestyle changes, andther self-care behaviors with the patient.

In addition to discharge education for the prevention ofeart failure readmission, patients also require appropriateutpatient follow-up and support. Without it, the likelihoodf readmission with heart failure increases. Outpatient dis-
ase management programs have been shown to reduce
www.AJConline.org

hmetht

gsihbopig

U

Amcimcnrditpmnrtpaw

bteviw(pdetdmb

U

DbbdScdctc

awditiefliips

C

whawtihHpvtdomthtshlpd

ad


ospitalization rates and associated costs.13 Initiation of aultidisciplinary heart failure program that includes patient

ducation and close patient follow-up in combination withhe use of appropriate heart failure medications can reduceospitalizations for ADHF by as much as 85% and improvehe functional status of patients.14–16

The goals of outpatient heart failure management pro-rams are not only to reduce the dependence on hospitalervices but also to offer disease management by monitor-ng the patient across the continuum of care. Outpatienteart failure disease management programs can range fromasic interventions, such as telephone follow-up, to the usef comprehensive multidisciplinary teams. Although theserograms are diverse, patient education provided by nursess a common thread and key element of all of these pro-rams.

sing Devices to Improve Connections with Patients

nother common thread in heart failure disease manage-ent programs is the goal of increasing the provider’s

onnection with patients. Devices can play a role in achiev-ng this goal. Devices have been increasingly used in theanagement of patients with heart failure. For example,

ardiac resynchronization therapy (CRT) has been recog-ized as the standard of care for patients with medicallyefractory heart failure and ventricular conduction disor-ers. CRT devices are available either with or without anmplantable cardioverter defibrillator (CRT-D). In additiono providing therapies, these heart failure devices can alsorovide diagnostic information that allows clinicians toonitor patients more closely and effectively. Device diag-

ostics can provide useful clinical information about ar-hythmias. The diagnostic data can also provide insight intohe clinical status of patients with heart failure. For exam-le, a downward trend in heart rate variability or patientctivity or an increase in nighttime heart rate may correlateith worsening heart failure status.17

More recently, intrathoracic impedance monitoring haseen incorporated into certain devices and can be monitoredo assess volume status. A recently completed study by Yut al18 demonstrated that intrathoracic impedance is in-ersely related to pulmonary capillary wedge pressure. Asntrathoracic impedance decreases, pulmonary capillaryedge pressure increases. OptiVol fluid status monitoring

Medtronic, Inc., Minneapolis, MN), which has been incor-orated into selected CRT-D and implantable cardioverterefibrillator devices, uses this concept and may provide anarly warning of an impending ADHF. Information on in-rathoracic impedance, the OptiVol fluid index, and otheriagnostic measures available on the Heart Failure Manage-ent Report can be useful not only in the treatment process

ut can also figure prominently in patient education efforts. O

sing Device Diagnostics to Improve Patient Adherence

evice diagnostics provide another objective data point that cane used in a clinical evaluation. Device diagnostic information cane obtained easily through an in-office or remote reader, whichownloads a report from the device in �10 minutes. The Cardio-ight Reader (Medtronic, Inc.), for example, has no programmingapability and can be used by ancillary personnel to downloadevice diagnostic information. With the remote reader, patientsan download diagnostic information while at home or on vaca-ion. Device data should be obtained at every visit and used in theontext of the patient’s subjective complaints and clinical findings.

Device diagnostics can be used as a tool to improve patientdherence to treatment recommendations. Many times, patientsill forget or deny lapses in diet or medication regimens. Deviceiagnostic data, when shown to a patient, can facilitate moren-depth conversations about lifestyle changes and how they relateo their disease process. In our clinic we regularly share thisnformation with patients and use it as a patient education tool. Forxample, trends in the patient’s intrathoracic impedance, OptiVoluid index, and patient activity level can provide potential insight

nto patient nonadherence with self-care recommendations. Thisnformation can be used as a teaching tool to potentially affectatient behavior. The utility of device-based information can beeen in 3 cases from The Heart Group.

ase StudiesCase 1: The first case involves a 75-year-old man with

orsening heart failure symptoms who was evaluated in oureart failure clinic (Figure 1). Device diagnostics provideddditional objective information to confirm the diagnosis oforsening congestion and also served as a teaching tool for

he patient. After the patient was shown his intrathoracicmpedance report, he admitted that he had stopped weighingimself daily and was not adhering to his sodium restriction.e also reported that his wife had been having some medicalroblems and that, as a result, he was using canned soups andegetables for convenience. Examination of the OptiVol fluidrends report showed that his daily impedance values wereecreasing and that he was approaching the OptiVol thresh-ld. His diuretic dose was increased, and the device infor-ation allowed us to have a discussion about the impor-

ance of daily weight measurements and the resumption ofis sodium restriction. In addition, we discussed alternativeso canned products as well as other low-sodium meal pos-ibilities. When the patient was seen at a subsequent visit,is symptoms were much improved, and he reported fol-owing the recommended lifestyle changes. His daily im-edance values were increasing, and the OptiVol fluid indexecreased.

Case 2: An 82-year-old man reported that he was takingll of his medications faithfully and was not skipping anyoses (Figure 2). Review of his thoracic impedance and
ptiVol fluid index data revealed that the patient had

Fdh

Fst

31GRathman/Device Diagnostics as an Educational Tool

igure 1. Device diagnostics as a recall and teaching tool. AF � atrial fibrillation; CHF � congestive heart failure; COPD � chronic obstructive pulmonaryisease; CRT � cardiac resynchronization therapy; CRT-D � cardiac resynchronization therapy/defibrillator; EF � left ventricular ejection fraction; HF �
eart failure. 1 lb � 0.45 kg.
igure 2. Device diagnostics to improve patient adherence to medications. A The dotted line indicates the physician programmed OptiVol threshold; B Theolid line indicates the daily impedance value. The dotted line indicates the reference impedance. AF � atrial fibrillation; CRT-D � cardiac resynchronization
herapy/defibrillator; EF � left ventricular ejection fraction.

tetmufiadpt

tmIbvpmmarSg

ta

U

DmppwatopdTpwri(p

FCfv


hreshold crossings indicative of worsening heart failurevery 4–6 weeks. These data were reviewed with the pa-ient who subsequently admitted that he ran out of hisedications at the end of each month and did not refill them

ntil he felt poorly. Consultation with his pharmacy con-rmed this refill pattern, and the patient noted he could notfford his monthly medication expenses. This additionalata allowed us to intervene with medication assistance androvided a teaching point on the importance of his medica-ion adherence.

Case 3: In another application of device-based diagnos-ic information, a patient’s activity level can be used toonitor patient adherence with exercise recommendations.

f a patient is encouraged to start a walking program, it iseneficial to show the patient his or her report to give themisual confirmation of their activity and subsequentrogress. A 73-year-old man was seen for a routine appoint-ent in the heart failure clinic. The Heart Failure Manage-ent Report demonstrated that he clearly had been more

ctive. Discussion with the patient determined that the pe-iod in question correlated with his winter vacation to Palmprings, Florida, and that he had been playing golf and

igure 3. Device diagnostics to improve patient adherence to exercise preAD � coronary artery disease; CHF � congestive heart failure; CRT-D

raction; exam � examination; HTN � hypertension; I � interrogate; mentricular.

oing for walks daily (Figure 3). The report served as a p

eaching tool for the patient to reinforce recommendationsbout increasing his exercise.

sing Device Diagnostics as a Disease Management Tool

evice diagnostics can also be used as a disease manage-ent tool because the data are objective (similar to the

atient’s vital signs). The data can be used to follow theatient’s response to treatment or to alert the provider toorsening congestion before the patient becomes symptom-

tic, both of which could potentially prevent future hospi-alizations. However, the clinician must recognize the utilityf the available information and download the data. As wasreviously mentioned, accessing the data is easy and can beone efficiently, either in the office or remotely from home.his allows a greater connection between the patient androvider. Newly released devices will allow the automatedireless download of information from the device, further

educing the need for patient compliance in transmittingnformation. For example, the Concerto CRT-D deviceMedtronic, Inc.) can be programmed to remotely alert theatient’s healthcare provider via phone, fax, or pager about

n. AF � atrial fibrillation; AT � atrial tachyarrhythmia; Avg � average;iac resynchronization therapy/defibrillator; EF � left ventricular ejectionaximum; NYHA � New York Heart Association; P � program; V �

scriptio� cardax � m

arameters, such as episodes of atrial fibrillation, delivered

dwrpc

t“avmnsisaIefbt

C

Dtnwocitptt

1

1

1

1

1

1

1

1

1

33GRathman/Device Diagnostics as an Educational Tool

evice therapies, and battery/lead issues, among others,ithout intervention from the patient. However, currently, a

emote alert for OptiVol fluid status monitoring is not ap-roved by the US Food and Drug Administration (FDA) andannot be programmed.

Although device diagnostics provide a great deal of po-entially beneficial information, they can also result in dataoverload,” with information from remote alerts being sentt varying times during the day. Thus, providers must de-elop systems to deal with these issues. Heart failure diseaseanagement programs that involve a team of physicians,

urse practitioners, and nurses are ideally suited to developystems to handle these issues and provide greater monitor-ng of patients. Electrophysiology and heart failure clinicshould collaborate to share device diagnostic informationbout patients, especially if there are noteworthy changes.n many cases, both clinics are interested in the data gen-rated, and changes in patient treatment may require inputrom both groups. To develop collaborative relationships,oth groups should meet and discuss how they will integratehe data and what types of diagnostic data should be shared.

onclusion

evices have improved the connection between patients andheir healthcare providers. Furthermore, device-based diag-ostics can provide an additional source of objective data,hich can be reviewed with the patient during an office visitr remotely while the patient is at home. This informationan be used to assess patient adherence to the treatment plann addition to serving as a patient teaching tool. Incorpora-ion of devices into a heart failure disease managementrogram allows closer monitoring of patients, with the po-ential to further decrease hospitalizations and improve pa-ient outcomes.

1. Koelling TM, Chen RS, Lubwama RN, L’Italien GJ, Eagle KA. Theexpanding national burden of heart failure in the United States: theinfluence of heart failure in women. Am Heart J 2004;147:74–78.

2. American Heart Association. Heart Disease and Stroke Statistics—2006 Update. Dallas, TX: American Heart Association, 2006.

3. Hunt SA, Abraham WT, Chin MH, Feldman AM, Francis GS,Ganiats TG, Jessup M, Konstam MA, Mancini DM, Michl K, et al.,for the Writing Committee to Update the 2001 Guidelines for theEvaluation and Management of Heart Failure. ACC/AHA 2005
guideline update for the diagnosis and management of chronic heart
failure in the adult: a report of the American College ofCardiology/American Heart Association Task Force on PracticeGuidelines. J Am Coll Cardiol 2005;46:1116 –1143.

4. Aghababian RV. Acutely decompensated heart failure: opportunitiesto improve care and outcomes in the emergency department. RevCardiovasc Med 2002;3(suppl 4):S3–S9.

5. Ghali J, Kadakia S, Cooper R. Precipitating factors leading to decom-pensation of heart failure: traits among urban blacks. Arch Intern Med1988;148:2013–2016.

6. Vinson JM, Rich MW, Sperry JC, Shah AS, McNamera T. Earlyreadmission of elderly patients with congestive heart failure. J AmGeriatr Soc 1990;38:1290–1295.

7. Bennett SJ, Huster GA, Baker SL, Milgrom LB, Kirchgassner A, BirtJ, Pressler ML. Characterization of the precipitants of hospitalizationfor heart failure decompensation. Am J Crit Care 1998;7:168–174.

8. van der Wal M, Jaarsma T, van Veldhuisen DJ. Non-compliance inpatients with heart failure: how can we manage it? Eur J Heart Fail2005;7:5–17.

9. Grady KL, Dracup K, Kennedy G, Moser DK, Piano M, StevensonLW, Young JB. Team management of patients with heart failure.Circulation 2000;102:2443–2456.

0. Ni H, Nauman D, Burgess D, Wise K, Crispell K, Hershberger RE.Factors influencing knowledge of and adherence to self-care amongpatients with heart failure. Arch Intern Med 1999;159:1613–1619.

1. Fonarow GC, Yancy CW, Heywood JT, for the ADHERE ScientificAdvisory Committee, Study Group, and Investigators. Adherence toheart failure quality of care indicators in US hospitals. Arch InternMed 2005;165:1469–1477.

2. Koelling TM, Johnson ML, Cody RJ, Aaronson KD. Discharge edu-cation improves clinical outcomes in patients with chronic heart fail-ure. Circulation 2005;111:179–185.

3. Rich MW. Heart failure disease management: a critical review. J CardFail 1999;5:64–74.

4. Fonarow GC, Stevenson LW, Walden JA, Livingston NA, Steimle AE,Hamilton MA, Moriguchi J, Tillisch JH, Woo MA. Impact of acomprehensive heart failure management program on hospital read-mission and functional status of patients with advanced heart failure.J Am Coll Cardiol 1997;30:725–732.

5. Holst DP, Kaye D, Richardson M, Krum H, Prior D, Aggarwal A,Wolfe R, Bergin P. Improved outcomes from a comprehensive man-agement system for heart failure. Eur J Heart Fail 2001;3:619–625.

6. McAlister FA, Lawson FM, Teo KK, Armstrong PW. A systematicreview of randomized trials of disease management programs in heartfailure. Am J Med 2001;110:378–384.

7. Adamson PB, Smith AL, Abraham WT, Kleckner KJ, Stadler RW,Shih A, Rhodes MM, on behalf of the InSync III Model 8042 andAttain OTW Lead Model 4193 Clinical Trial Investigators. Continu-ous autonomic assessment in patients with symptomatic heart failure:prognostic value of heart rate variability measured by an implantedcardiac resynchronization device. Circulation 2004;110:2389–2394.

8. Yu CM, Wang L, Chau E, Chan RH, Kong SL, Tang MO, ChristensenJ, Stadler RW, Lau CP. Intrathoracic impedance monitoring in patientswith heart failure: correlation with fluid status and feasibility of early
warning preceding hospitalization. Circulation 2005;112:841–848.

Ta(utp

CH

CrreOHIcvptt

vpa

Ca

Ima

A

0d

Key Lessons from Cases Worldwide

Li Wang, PhD

Since OptiVol fluid status monitoring (Medtronic, Inc., Minneapolis, MN) was re-leased to the market in 2004, the manufacturer has been working together withclinicians to better understand the operation and utility of OptiVol in a clinicalsetting. This has been done through completed and ongoing clinical studies andthrough evaluation of case studies submitted by clinicians. In the process, much hasbeen learned about some interesting aspects of OptiVol fluid status monitoring and itsuse with patients with heart failure. © 2007 Elsevier Inc. All rights reserved. (Am
J Cardiol 2007;99[suppl]:34G–40G)
(bnvamect

am

C

Atroiivoqttifltiiiiewtiit

his article presents 6 cases, each of which illustratesn interesting aspect of OptiVol fluid status monitoringMedtronic, Inc., Minneapolis, MN). It is hoped that innderstanding these cases, clinicians will be better equippedo use OptiVol fluid status monitoring in managing theiratients with heart failure.

ase 1: Impedance Reduction Patterns Beforeeart Failure–Related Hospitalization

ase 1 is from a collection of patients. The OptiVol trendseports show the impedance reduction before heart failure–elated hospitalization in 4 different patients (Figure 1). Inach case, impedance decreased, and the patient crossed theptiVol fluid threshold before admission to the hospital.owever, the magnitude of impedance reduction varied.

mpedance reductions in relation to the OptiVol thresholdome in different sizes and forms. Patient 1 had a leftentricular lead dislodgement and experienced a huge im-edance reduction before admission. The impedance reduc-ion for patient 4 was very small but sustained for a rela-ively long period.

Key lesson: Impedance reduction before hospitalizationaries with the particular patient and the reason for decom-ensation, and it involves both the magnitude of reductionnd the duration of the reduction.

ase 2: Impedance Measurements Usingn Epicardial Patch

n this case, intrathoracic impedance measurements wereade between an epicardial defibrillation patch electrode

nd the device case rather than between the right ventricular

Medtronic, Inc., Minneapolis, Minnesota, USA.Address for reprints: Li Wang, PhD, Medtronic, Inc., 7000 Central

venue NE, Minneapolis, Minnesota 55432.
tE-mail address: [email protected].

RV) coil on a transvenous implantable cardioverter defi-rillator lead and the device case, which is typically theorm. As seen in Figure 2, the average daily impedancealues on the OptiVol fluid trends report are very low—round 30 �—rather than the 60–70 � typically seen inost patients. The impedances are very low because the

picardial patch has a much larger surface area than the RVoil. The bigger the surface area of the measuring electrode,he lower the expected impedance.

Key lesson: Absolute intrathoracic impedance values areffected by the sizes and locations of the electrodes used toeasure intrathoracic impedance.

ase 3: Rapid Impedance Changes

review of the definition and function of the reference impedancerend may be helpful in better understanding a case involvingapid impedance change. The reference impedance is initializedn the 34th day of measurements by averaging the last 4 dailympedance values. This is the only time that averaging is involvedn changes to the reference impedance. So the patient’s ownolume state when the reference impedance is initialized—euv-lemic or hyper- or hypovolemic—will be the starting point foruantifying subsequent deviations in impedance. After initializa-ion, the reference impedance will go up and down over time inhe same direction as the daily impedance. But the referencempedance is only allowed to change slightly from day to day. Theuid index will only increase when the daily impedance is below

he reference line, and will be reset to zero if the daily impedences above the reference line for 2–3 days. The reference impedances not programmable. If, for example, the reference impedance isnitialized while the patient is hypovolemic, the OptiVol fluidndex may go up as the patient is improving and approaching auvolemic state. This may be interpreted as fluid buildup and aorsening of the patient’s condition, unless the patient’s history is

aken into account during review of the data. The OptiVol fluidndex can be used to quantify a sustained reduction in dailympedance, but it is not a measure of a specific fluid level. Even-ually, however, the reference impedance will catch up and reflect
he patient’s stable pulmonary and clinical state.
www.AJConline.org

Figure 1. (A and B). OptiVol fluid (Medtronic, Inc., Minneapolis, MN) data trends before and after a heart failure–related hospitalization in 4 different patients. The vertical solid line in Brepresents the time of heart failure–related hospitalization.

35GW

ang/Key

Lessons

fromC

asesW

orldwide

pw

c2

Fla e.

Fph her than


Keeping this information in mind, consider the case of aatient who committed a number of dietary indiscretions

igure 2. Impedance measured between an epicardial patch and the deviocation of the electrodes used to measure intrathoracic impedance. Iccumulation of the difference between the daily and reference impedanc

igure 3. Diureses of a large volume during heart failure (HF) admission.eriod as a result of diuresis during an HF–related hospitalization. The refeas been designed to capture slow changes in intrathoracic impedance rat

hile on a cruise vacation. The patient’s OptiVol fluid index 2

rossed the nominal threshold of 60 �-days on November 3,004 and was admitted to the hospital on November 12,

. The absolute intrathoracic impedance value is affected by the size androgate. OptiVol fluid (Medtronic Inc., Minneapolis, MN) index is an

ly impedance increases dramatically after the patient lost 10 kg in a shortpedance, however, lags behind and increases only slowly. The algorithmlarge, sudden changes.

ce case� inter

The dairence im

004, with severe symptoms of heart failure. He underwent

dscrt

pubapr2lfiit

hipt

aascnpwaOwdphc

ocbado

Fst(

37GWang/Key Lessons from Cases Worldwide

iuresis and lost 10 kg in a very short period. He thenhowed clinical improvement and the daily impedance in-reased dramatically—approximately 25 � (Figure 3). Theeference impedance lags behind but increases at a rela-ively slow rate per the design of the algorithm.

On January 4, 2005, the patient was once again hos-italized because of severe heart failure symptoms. Henderwent diuresis and lost approximately 3 kg beforeeing discharged on January 25, 2005. There was anssociated increase in the daily impedance during thiseriod. After discharge, the daily impedance remainedelatively stable. At the next follow-up, in mid-April005, the patient reported feeling better than he had in aong time, and his physicians and nurses believed he hadnally stabilized. The reference impedance had also been

ncreasing during this time and has now “caught up” withhe daily impedance (Figure 4).

The OptiVol fluid index was flat after the initialospitalization in November 2004, because the referencempedance was below the daily impedance for the mostart. Therefore, the fluid index remained at zero, even

igure 4. Reference impedance after discharge. Eventually the referencetatus. In this case, the patient’s intrathoracic impedance remained stable foo the daily impedance. FU � follow-up; HF � heart failure; Hosp � hospiCourtesy of Medtronic, Inc., 2005.)

hough the daily impedance was decreasing significantly r

fter the initial hospital discharge and before the seconddmission in early January 2005. In the weeks before theecond hospitalization, the reference impedance was in-remented slowly as per the algorithm. Therefore, there iso OptiVol fluid index because the patient’s daily im-edance is higher than the reference impedance. It is onlyhen the daily impedance is below the reference imped-

nce because of progressive decompensation that theptiVol fluid index starts to increase. After the patientas admitted and underwent diuresis in the hospital, theaily impedance increased. Subsequently, the daily im-edance remained stable for approximately 1 month afterospital discharge, and the reference impedance finallyaught up with the daily impedance (Figure 4).

The reference impedance has been designed to capturer follow the slow changes in impedance, not the rapidhanges, which is why the reference impedance lagsehind the dramatic change in the patient’s daily imped-nce. As such, there have been other examples of aramatic decrease in the patient’s daily impedance with-ut changes in the OptiVol fluid index. We are actively

ce will catch up with the daily impedance and reflect the patient’s fluidximately 1 month after discharge, and the reference impedance caught upinterrogate; MD � medical doctor; P � program; RN � registered nurse.

impedanr approtal; I �

esearching potential improvement in this area.

tcar

C

Pfpeifralt2aas

inl

C

Amodrcmtarae

actpcO

C

Foi

Fti


Key lesson: Daily impedance reflects the patient’s sta-us. It may take some time for the reference impedance toatch up when there are rapid changes in daily impedancessociated with significant changes in patient status, eg, as aesult of inhospital intravenous diuretic treatment.

ase 4: Pleural Effusion

leural effusion may or may not be related to heartailure; however, worsening heart failure can lead toleural effusion. In either case, pleural effusion is anotherxample in which one may see a reduction in dailympedance. In this particular case, the patient reportedatigue and had weight gain and limited capacity at aegular follow-up visit. The patient was hospitalized, andfter some analysis, a diagnosis of pleural effusion on theeft side was made. The patient was treated with a pleuralap, and the fluids drained very quickly. The patient lost.8 kg, had better capacity, and had less dyspnea. Imped-nce is affected by rapid fluid removal from the chest, socorresponding rapid increase in daily impedance can be

een on the OptiVol fluid trends report (Figure 5).

Key lesson: If fluid in the chest is removed quickly,mpedance will increase rapidly. Fluid outside the pulmo-ary circulation will also affect impedance measurement, as

igure 5. Fluid removal due to pleural effusion. Intrathoracic impedance wio heart failure. In this example, the patient experienced pleural effusion wncreased after the fluid was removed from the patient’s chest with a pleu

ong as it is in the measurement pathway. c

ase 5: Pneumothorax

similar concept is involved when a patient has pneu-othorax. The patient in this case had a late occurrence

f pneumothorax and was hospitalized. An endocostalrain was performed, and the air was removed. A rapideduction in daily impedance was seen, and the patientrossed the OptiVol threshold (Figure 6). This does notean that the patient had fluid overload. It only means

hat the daily impedance was reduced because the extrair in the chest, which is an electrical insulator, wasemoved. The impedance measurements were accurate,nd OptiVol fluid status monitoring was operating prop-rly.1

Key lesson: As discussed in case 4, impedance will beffected by changes within its measurement pathway. In thisase, the impedance value is affected by the air volume inhe chest. Additionally, it is very important to know theatient and his or her medical history before any judgmentan be made regarding a patient’s fluid status based onptiVol data.

ase 6: Pocket Revision

luid will accumulate in the pocket after the initial implantr a revision because of the surgery. This fluid accumulations reflected in the impedance measurements, as seen in the

ected by fluid changes in the chest whether or not those changes are relatedorresponding decrease in intrathoracic impedance. The impedance valuesFU � follow-up; Pt � patient.

ll be affith a c

ral tap.

ase of a patient who had a pocket revision, because the

FTr

Ffbd

39GWang/Key Lessons from Cases Worldwide

igure 6. Pneumothorax. A rapid reduction in intrathoracic impedance was observed after the patient’s pneumothorax was treated with an endocostal drain.he reduction in impedance was not caused by fluid overload but rather by removal of the extra air in the chest. Air is an electrical insulator, and its removal

esults in decreased intrathoracic impedance. P � program. (Reprinted from Pacing Clin Electrophysiol.1)

igure 7. Pocket revision will affect the daily impedance and the reference impedance and may be viewed as a false positive. The amount of time it takesor impedance values to recover depends on the specific patient. OptiVol fluid (Medtronic Inc., Minneapolis, MN) is an accumulation of the differenceetween the daily and reference impedance. Patient history is important in determining whether the patient has crossed the OptiVol threshold for
ecompensated heart failure or for another reason. P � program.

pDdtptk

apqha

atvcr

C

Asap

maapitt

tflaqtpt

tcapdoml

1


ocket fluid is within the measurement pathway (Figure 7).aily impedance values were relatively stable until theyecreased precipitously, and the patient crossed the OptiVolhreshold after the pocket was revised. Is this really a false-ositive finding? The OptiVol fluid index reacted correctlyo the pocket edema. Again, if the patient’s history isnown, this should not be an issue.

What is the expected length of time for edema to resolvend for impedance measurements to stabilize after theocket is revised? There is no definitive answer to thisuestion. In our clinical experience, my colleagues and Iave seen the impedance stabilize as early as 4–5 weeks ands late as 3 months.

Key lesson: Revising the pocket for any reason willffect impedance measurements, and the amount of time itakes for impedance values to recover and stabilize mayary by patient. Determining whether an OptiVol thresholdrossing is owing to the pocket revision or other causesequires evaluation of the patient’s history.

onclusion

s real-life experiences with intrathoracic impedance mea-urements accumulate, we are also learning how to interpretnd better use the data in clinical practice. The daily im-
edance reflects the patient’s status and may be affected by
ultiple factors, some of which are owing to heart failurend others which are not. A decrease in impedance may notlways mean a clinical decompensation caused by increasedulmonary volume overload. To reach a clinical conclusion,t is necessary to review the OptiVol data, and more impor-antly, the patient data, including the history, signs/symp-oms, and medications.

The reference impedance and OptiVol fluid index are toolso quantify impedance changes but are not a direct measure ofuid status. The reference impedance is designed to capturend follow the slow changes in impedance and is used touantify the impedance reduction. It does not reflect the pa-ient’s ideal fluid status. The fluid index is a measure of im-edance reduction (measured in ohms) and the sustainability ofhe reduction (measured in days), and is reported in ohm-days.

In some cases, the reference impedance will lag behindhe daily impedance, especially when there has been a rapidhange in daily impedance. When the daily impedance isbove the reference impedance, reductions in the daily im-edance will not be reflected in the fluid index, although theaily impedance reduction is a result of pulmonary volumeverload caused by worsening heart failure. Therefore, it isore important to review the daily impedance trend when

ooking at the OptiVol fluid status trend report.

. Timperley J, Mitchell A, Brown P, Betts T. Changes in intrathoracicimpedance from a pneumothorax: insights from an implanted monitor-
ing system. Pacing Clin Electrophysiol 2005;28:1109–1111.

Dupicasdc

swairTitd

Ba

hc

O

vC

0d

Collaboration Among General Cardiologists, Heart FailureSpecialists, and Electrophysiologists: What Are the Barriers?


Although ongoing research has focused on the reliability and accuracy of data derivedfrom implanted devices, important challenges remain, among which is the develop-ment of a framework for handling the data, including delineation of responsibility forinterpretation and management of the data. The barriers to a successful device-monitoring program include difficulties in data access, unfamiliarity with a newtreatment approach, lack of sufficient knowledge and experience in data interpreta-tion, and difficulty incorporating the program into the existing workflow. The dy-namics of collaboration between heart failure and electrophysiology (EP) specialistsmay differ locally and regionally, but a common set of barriers resulting fromtraditional models of care constrains incorporation of device-based data into clinicalpractice. Active dialogue between EP and heart failure specialists that involves thesharing of experiences and practice models is an important first step to overcomingthese barriers. In the future, a “virtual clinic” model of patient care can only berealized with the successful incorporation of remotely accessed device data. © 2007
Elsevier Inc. All rights reserved. (Am J Cardiol 2007;99[suppl]:41G–44G)
hwaowpicmrtctdchdbtiwi

mtmWbafp

ata obtained from implantable devices have severalnique advantages. These devices have the ability torovide objective and individualized data using existingnterrogation techniques with less reliance on patientompliance. However, the successful implementation ofny clinical monitoring strategy for heart failure requirescientific validation of the accuracy and reliability ofiagnostic algorithms, evidence of clinical benefit, andost-effectiveness.

As technology continues to evolve, an important buteldom-discussed challenge is how to handle the data andho should be responsible for its interpretation and man-

gement. Interestingly, patients often welcome remote mon-toring of their clinical status, although payers are ofteneluctant to recognize such efforts as patient encounters.his article broadly discusses some of the barriers involved

n the incorporation of device-based data into clinical prac-ice, potential solutions, future clinical studies, and futureirections for this mode of treatment.

arriers to Collaboration Between Heart Failurend Electrophysiology Specialists

The data access barrier: Specialization in medicineas created isolated “silos” of care with little or poorommunication among physicians treating patients with

Department of Cardiovascular Medicine, Cleveland Clinic, Cleveland,hio, USA.

Address for reprints: W. H. Wilson Tang, MD, Department of Cardio-ascular Medicine, Cleveland Clinic, 9500 Euclid Avenue, Desk F25,leveland, Ohio 44195.

DE-mail address: [email protected].


eart failure. This is a complex problem that variesidely according to the nature and location of practice

nd is perhaps the biggest challenge to the future devel-pment of device-based clinical monitoring for patientsith heart failure. In the early days of cardiac pacing,racticing cardiologists performed most of the devicenterrogations and follow-up. This is still true in manyountries. However, as implantable devices have becomeore sophisticated, many of these tasks have become the

esponsibility of electrophysiology (EP) specialists orheir nurse extenders. Device interrogation itself has be-ome increasingly complicated, particularly with the op-imization of biventricular pacemakers. Simply obtainingata from the device, either at the bedside or remotely,an be daunting to nonspecialists. Indeed, most non-EPealthcare providers are unfamiliar with sophisticatedevices and may have limited experience with device-ased monitoring. Fewer and fewer nonspecialists holdhe primary responsibility of interrogating and managingmplanted devices. As a result, EP specialists are facedith the prospect of establishing systems to meet the

ncreased need for device interrogation and maintenance.

The treatment approach barrier: Despite advances inedical and device therapy, the approach we use when

reating patients with heart failure has not changed dra-atically. Patients are routinely observed in the clinic.hen worsening signs and symptoms occur, patients can

e seen either in the urgent setting or after they have beendmitted to the hospital. This reactive approach stemsrom a fundamental lack of understanding in how stableatients with heart failure deteriorate and decompensate.
ata, such as intrathoracic pressures or impedance deriv-
www.AJConline.org

amtosafdHmi

tOoppomiTirfOss

t

figtdivascnocdimiirc

R

Tciw

Fo


tives, can provide new insight into the hemodynamicanifestations of disease progression. These data allow

he documentation of physiologic fluctuations in advancef the inciting hospitalization event rather than relying onubjective complaints or alterations in weight or edema tossess patient stability. The availability of such dataorces us to rethink our approach to heart failure and toevelop an entirely different proactive treatment model.owever, there are a number of major barriers thatust be addressed before this new approach can be

mplemented.

The knowledge barrier: The threshold of 60 �-dayshat is used to predict heart failure hospitalization withptiVol fluid status monitoring (Medtronic, Inc., Minneap-lis, MN) is an arbitrarily defined limit based on a singleatient series. It is a first-pass attempt to establish a triageoint that balances sensitivity and specificity. As with anyther physiologic processes, intrathoracic impedance infor-ation requires interpretation as part of a spectrum of man-

festations that may vary from individual to individual.herefore, a large knowledge gap still exists in understand-

ng the consistency and reliability of the derived intratho-acic impedance data (Figure 1). Thus, we are still far awayrom reaching a consensus on the frequency with whichptiVol data should be checked, the optimum threshold that

hould be used, and the appropriate clinical responses touch data.

The workflow barrier: New technologies, such as in-

igure 1. Challenges in proactive device-based remote monitoring: manygy.

rathoracic impedance, can be intimidating for both heart a

ailure and EP specialists because of the potential forncreased workload and responsibility, and to some de-ree, liability. EP specialists may be reluctant to implanthese devices if they are the primary recipients of theownloaded data, especially in the case of remote devicenterrogations, because of the added responsibilities in-olved in acting on the results. These actions may includedjustment of diuretics or clinical evaluation beyond thecope of main areas of expertise of the EP specialist. Inontrast, cardiologists, heart failure specialists, andurses who have been trying very hard to find a reliablebjective assessment of fluid status and an index oflinical stability may find themselves facing many hur-les in actually obtaining such device information for usen the clinical evaluation of patients. Additionally, theyay have difficulty in interpreting the results of a device

nterrogation. This labor-intensive process does not eas-ly fit into current workflow patterns and has yet to beeimbursed, making effective implementation even morehallenging.

ecommendations for Successful Collaboration

he ideal collaboration resembles the concept of a “virtuallinic” that EP physicians have successfully incorporatednto their device clinics. This can involve asking patientsith heart failure to track their blood pressure, heart rate,

nding questions. CHF � congestive heart failure; EP � electrophysiol-
outsta
nd body weight as part of their self-monitoring tasks and

pu

hapsdbpo

etfcipiriraaer

flsimub

SbdaMrslrq

F

Tpiorhimcmadptc

scia

TC

M

F

SEP

P

E

MPRS

43GTang/Collaboration Among General Cardiologists, Heart Failure Specialists, and Electrophysiologists

eriodically reviewing their device data as part of the reg-lar clinical assessment.

Start a dialogue: The first and foremost priority is forealthcare providers to meet with each other and discussvailable options. This dialogue can center on individualatient cases and discussion of the choice of devices, or inome cases, how to best evaluate and manage patients withevices that provide potential diagnostic monitoring capa-ilities. This discussion need not be limited to a particularractice or institution; it can also involve a regional collab-ration on use of heart failure disease management clinics.

Learn device interrogation procedures and sharexperiences: Among the most important requirements ofhe proactive use of device-based monitoring is to becomeamiliar with device interrogation procedures and establishollaborative relationships among the different caregiversnvolved in the task of clinical monitoring. The educationalrocess should not be limited to learning how to measurentrathoracic impedance. It should also involve device pa-ameters available from almost any standard pacemaker ormplantable cardioverter defibrillator (ICD), such as heartate, arrhythmic episodes and characteristics (especiallytrial fibrillation), and pacing frequencies. Before a reliablelgorithm for the use of data obtained from devices can bestablished, ongoing discussions and potential questions foresearch projects should be considered.

Incorporate shared responsibilities into the work-ow: Teamwork is the key to successful collaboration. EPpecialists must be reassured that the data obtained frommplantable devices will be used well and that the manage-ent responsibilities will be shared. Organizing heart fail-

re and EP clinics to meet the demands of the workflow can

able 1urrent and ongoing clinical research studies contributing to the knowled

Study P

IDHeFT Proof of concept study that showed inverse correcapillary wedge pressure; sensitivity of 76.9%

AST IDE trial that characterized impedance data in rehospitalizations and diuretic use

ENSE-HF OUS Postmarket trials to provide sensitivity and falseU RegistryARTNERS-HF andOFISSER

Postmarket trials to characterize the relation betw

RECEDE-HF IDE trial to compare the composite of heart failuCardiac Compass with OptiVol* guided-careCardiac Compass with OptiVol in North Ame

xternal researchstudies

Multiple physician-initiated trials to investigate v

EU Registry � European Observational InSync Sentry Study; FASTIDHeFT � Medtronic Impedance Diagnostics in Heart Failure Trial;

rogram to Access and Review Trending Information and Evaluate Correlandomized, Evaluation Using Cardiac Compass with OptiVol in Earlyensitivity of the InSync Sentry OptiVol Feature for the Prediction of He* OptiVol; Medtronic, Inc., Minneapolis, MN.

e among the most challenging aspects of collaboration. a

ome clinics make arrangements so that a patient can visitoth clinics on the same day. Others ask their patients toownload device data before clinic visits so that it is avail-ble for use at the time of their clinic visit. Germany andurray1 and Small2 share their insight on how to incorpo-

ate device data into the clinic in articles elsewhere in thisupplement. However, their arrangements may not be rep-icated easily in other practices because of expertise andequired equipment, and because of the comfort level re-uired on the part of both heart failure and EP specialists.

uture Directions

he use of device-based data in the proactive monitoring ofatients with heart failure will require adequate confidencen the data so that “well-care visits” can be conductednline. It will also require the development of local oregional networks to interpret and distribute device data andeart failure disease management programs that can operaten a scalable fashion. Additionally, it requires that remoteonitoring be reimbursed in much the same manner as

linic-based interrogation of ICDs or pacemakers. Remoteonitoring allows more frequent and convenient follow-up

nd may reduce the overall workload, if algorithms can beeveloped to flag abnormal data for the early detection ofroblems. After establishing these tools, we will then needo determine what treatment strategies can alter the naturalourse of the disease process.

Regulatory agencies are also very sensitive to these is-ues and have expressed different concerns in differentountries. At present, the current indications for the use ofntrathoracic impedance measurements or, for that matter,ny device-based hemodynamic monitoring are limited to

trathoracic impedance assessment

Status

etween intrathoracic impedance and pulmonarycting heart failure hospitalizations

Published

o patients’ clinical status, including Completed, ongoingextension phase

tes from blinded data; utility with alert enabled Ongoing

onitoring data and heart failure–related events Ongoing

italization and all-cause mortality in thent arm versus the control arm blinded to

Planning

aspects of OptiVol Ongoing

d Accumulation Status Trial; IDE � investigational device exemption;R � OptiVol Fluid-Index InSync Sentry Registry; PARTNERS-HF �Symptoms in Patients with Heart Failure; PRECEDE-HF � Prospective,

tion of Decompensation Events for Heart Failure; SENSE-HF OUS �ure Outside of the US.

ge of in

urpose

lation bin dete

lation t

alert ra

een m

re hosptreatmericaarious

� FluiOFISSEation to

Detecart Fail

iding in the assessment of a patient’s heart failure status.

Srss

C

OclioLsc

sa

t

1

2

3


everal registries and outcomes research trials that are cur-ently ongoing or in the planning stages may help demon-trate the potential benefits of treatment strategies guided byuch device-based data (Table 1).3

onclusion

ptiVol fluid monitoring has started a new era in patientare by facilitating effective remote monitoring of physio-ogic data. However, a framework must be developed toncorporate the use of these data into clinical practice. Ane-size-fits-all solution will not work for this problem.ocal and regional dynamics between heart failure and EPpecialists may vary, but traditional models of care pose
onstraints. Active dialogue between EP and heart failure
pecialists is an important first step in developing successfulrrangements for the use of device-based data.

Acknowledgment: I thank Nancy Johnson for providinghe summary table of upcoming clinical trials.

. Germany R, Murray C. Use of device diagnostics in the outpatientmanagement of heart failure. Am J Cardiol 2007;99(suppl):11G–16G.

. Small RS. Integrating device-based monitoring into clinical practice:insights from a large heart failure clinic. Am J Cardiol 2007;99(suppl):17G–20G.

. Yu CM, Wang L, Chau E, Chan RH, Kong SL, Tang MO, ChristensenJ, Stadler RW, Lau CP. Intrathoracic impedance monitoring in patientswith heart failure: correlation with fluid status and feasibility of early
warning preceding hospitalization. Circulation 2005;112:841–848.

The American Journal of Cardiology -...

Documents

Transcript of The American Journal of Cardiology -...