S22

Journal of Cardiovascular NursingVol. 20, No. 5S, pp S22–S30 ❘ © 2005 Lippincott Williams & Wilkins, Inc.

Julie A. Shinn, MA, RN, CCRN, FAANStanford University Medical Center, Stanford Hospital and Clinics,Calif.

Corresponding authorJulie A. Shinn, MA, RN, CCRN, FAAN, Stanford University MedicalCenter, Stanford Hospital and Clinics, Room H0105 MC 5221, 300Pasteur Dr, Stanford, CA 94305 (e-mail: jshinn@stanfordmed.org).

Implantable LeftVentricular Assist DevicesJulie A. Shinn, MA, RN, CCRN, FAAN

The first bridge to transplant with a left ventricular assist device (LVAD) was preformed over 20 years ago. Since that time, the devices have continued to evolve and now patients arebeing supported with devices as an alternative to transplantation. The primary indication ofend-stage heart failure remains the same but increased knowledge about patient selection, thetiming of implant, and patient management have contributed to improved outcomes withdecreasing adverse events. Multiorgan failure, right ventricular failure, bleeding, infection,thromboembolism, and device malfunction continue to be the most serious threats to long-term survival in these patients. Despite that, patients who do well are now able to bedischarged from the hospital and resume relatively normal lives with the devices. The articlereviews 3 of the most widely used LVADs for bridge to transplant therapy: the Thoratec;HeartMate vented electric; and the Novacor Left Ventricular Assist System. Indications,mechanism of operation, clinical problems, and out-of-hospital preparation are outlined.

KEY WORDS: alternative to transplantation, heart failure, ventricular assist devices

Considerable progress has been made in the lastdecade in treatment of congestive heart failure.Newer pharmacologic agents, biventricular pacing,and implantable defibrillators have enabled cardiolo-gists with the ability to maintain patients in a com-pensated state with good quality of life for long peri-ods of time. Surgical techniques of revascularization,valve repair, or replacement and surgical remodelingof the left ventricle also have extended the lives ofmany patients. Despite these advances, the 5-yearmortality rate for heart failure continues to be about70%.1 Approximately 5 million Americans haveheart failure today and after the age of 65 years, itaffects approximately 10 per 1000 of the population.1

Heart failure is the cause of approximately 250,000deaths per year in the United States.2 As heart failureprogresses and becomes more refractory to treatment,subsequent secondary organ dysfunction develops asa result of chronic poor organ perfusion. At this end

stage of the clinical syndrome of heart failure, there isa 1-year mortality of more than 50%.3

Heart transplantation is a treatment option formany end-stage heart failure patients but with donorsupply limited to approximately 2500 organs per year,it is clearly not an option for most patients.4 Youngerpatients without comorbidities are the most likely can-didates for heart transplantation. They are also thebest-suited candidates for left ventricular assist device(LVAD) implantation as a bridge to transplantation.The use of LVADs as bridges to transplantation hasbecome a standard treatment option for advancedheart failure. Between January 2002 and May 2004,312 patients in the International Society for Heart andLung Transplantation’s (ISHLT’s) Mechanical Circul-atory Support Database received bridge to transplanttherapy in 48 centers submitting data5; 35 of those cen-ters were in the United States. This number is actuallyan underestimate of all LVAD patients being bridged totransplant as the database is still relatively new andmany centers have yet to enroll.

The goal of most companies that are developingLVAD devices is to develop products that can serve aspermanent, definitive therapy. The HeartMate ventedelectric (VE) system (Thoratec, Pleasanton, Calif) wasapproved for this use by the Federal Food and DrugAdministration (FDA) for destination therapy in

Implantable Left Ventricular Assist Devices S23

2003 and the Novacor Left Ventricular Assist System(LVAS) (World Heart, Oakland, Calif) is in the finalstages of an evaluation study for premarket approval.It is estimated that 30,000 to 60,000 end-stage heartfailure patients in the United States per year would bepotential candidates for destination therapy.6

The article reviews the clinical indications, clinicalproblems, and out-of-hospital preparation of the 3most widely used LVADs as bridges to transplantationtherapy: the Thoratec; HeartMate VE, XVE (a newerrevision of the VE); and Novacor LVAS devices.

Historical PerspectiveThe first successful bridge to heart transplant with animplantable LVAD was done in 1984 with a NovacorLVAD.7 That patient lived more than 20 years aftertransplantation. Many lives have been saved by thistechnology but it has been difficult to determine whothe best candidates are and who has the greatestchances of survival. Because of the relatively low vol-ume of patients, no one center where LVAD implantsare performed can amass enough data to answer thatquestion well. For that reason, in 2001, a mechanicalcirculatory support device database was establishedby the ISHLT to collect and pool data from multipleinstitutions. The goal of this project is to accuratelyanalyze data to help improve short- and long-termoutcomes and to facilitate the selection of patientsbest suited for this therapy.5 The majority of thepatients who have been entered in the databasereceived LVADs capable of chronic support.

A significant study opened the door for theexpanded use of implantable LVADs as an alternativeto transplantation. The Randomized Evaluation ofMechanical Assistance for the Treatment of CongestiveHeart Failure (REMATCH) was a randomized, con-trolled trial of 129 patients in 20 institutions who wereineligible candidates for heart transplantation. Ofthose, 68 were randomized to receive the HeartMateVE LVAD and 61 patients served as controls, receivingoptimal medical management.8 The importance of thestudy was the demonstration that LVAD implants wereassociated with a relative reduction in the risk of deathof 48% during the entire follow-up period comparedto the medical therapy group (P � .001).8 This studyled to the FDA approval of the device for use as desti-nation therapy in November 2002. Subsequent to that,the Centers for Medicare and Medicaid Services (CMS)approved national insurance coverage for destinationtherapy in October 2003.

Indications for LVAD TherapyThe primary indication for LVAD therapy is end-stage heart failure that is no longer responding to

conventional medical management. If the implant isfor bridge to transplant, the patient would berequired to meet transplant criteria and to already beaccepted as a transplant candidate. Patient selectionis extremely important determinant of patient out-come following LVAD implant.9,10 Prior to implant,the majority of patients will be hospitalized on intra-venous inotropic therapy. Many will require the sup-port of intra-aortic balloon pump therapy, althoughsome centers do ventricular assist device (VAD)implantation without a trial on intra-aortic balloonpump therapy. It is important to time surgery beforerenal and hepatic dysfunction deteriorate and beforepostoperative right ventricular failure will become asignificant problem.10

The best outcomes will be achieved in patientsgoing for implant under elective conditions.Emergency implant is associated with higher mortal-ity.11 Investigators at Columbia University have cal-culated a risk factor score that can be used to predictoutcomes after device implant.12 They identified sev-eral risk factors that place patients at greater risk ofmortality in a consecutive series of 130 patients.Each factor carries a weighted score and a total scoreof more than 5 corresponded to 47% mortality com-pared to 12% mortality in patients with scores ofless than 5. Risk factors included mechanical ventila-tion, previous LVAD, or right ventricular assistdevice used as a bridge to a more long-term VAD,postcardiotomy shock, previous cardiac surgery, acentral venous pressure more than 16 mm Hg as anindicator of the extent of right ventricular failure, aprothrombin time more than 16 seconds, acutemyocardial infarction, and ischemic cardiomyopa-thy.12 The single most important risk factor for mor-tality after LVAD implant is the presence of preoper-ative mechanical ventilation.12 Table 1 summarizesthe risk factors for mortality. At present, no univer-sal criteria exist and patient selection remains a com-plex process.13

Overview of the DevicesThis review cannot possibly outline in detail all of theLVAD devices being used and evaluated today. Manyhave merits but are still in investigational stages requir-ing FDA approval for use in selected centers. Thisreview will focus on 3 devices that are used widely andhave FDA approval for bridge to transplantation.

Thoratec LVAD

The Thoratec LVAD is an external, paracorporeal,pulsatile pneumatic pump. Its placement is achievedvia a median sternotomy on cardiopulmonary bypass.The inflow cannula is placed in the left ventricular

S24 Journal of Cardiovascular Nursing ❘ September/October 2005

apex and tunneled subcostally to the external pump.Blood is returned via an outflow cannula, again, sub-costally to the ascending aorta. The inflow cannulacan also be placed in the left atrium but less optimalflows will be a disadvantage and that is not the usualapproach if the patient is being bridged to transplant.Figure 1 illustrates the placement of this pump. Theblood sac is a smooth, polyurethane material, whichwill necessitate that the patient be anticoagulated ini-tially with heparin and later converted to warfarinsodium. A console or the driver of the pump is con-nected via a pneumatic hose that will deliver alter-nating pressure or vacuum inside the rigid shell hous-ing the blood sac. With pressure, the blood sac col-lapses, causing ejection to occur. Application of vac-uum creates a pressure gradient assisting filling fromthe left ventricle. Unidirectional flow is maintainedby tilting disk valves in the blood pump. The maxi-mum stroke volume of this pump is 65 millimeters.Its flow or output will depend on how quickly thepump fills. Pump flow rates range from 1.3 to 7.2mm per minute.14

The cannula to and from the pump are coveredwith a Dacron material that allows for endothelialcell ingrowth, which effectively seals off the tract ofthe cannula into the mediastinum, preventing migra-tion of bacteria from the skin into the mediastinum.This act of sealing also allows the patient consider-able mobility after this process is complete. It usuallytakes up to 10 days for sealing to be completed.

An obvious advantage of this pump is that it canbe placed in smaller individuals such as women oradolescents. Its paracorporeal position does notrequire the larger body habitus needed for the totallyimplantable LVADs. This pump requires a body sur-face area of 1.3 m squared versus the totallyimplantable pumps, which require a body surfacearea of 1.5 meters squared. Another advantage of thesystem is that a second pump can easily be added tosupport the right ventricle when indicated.

Two modes of operation are commonly used withthis pump. It can be run at a fixed rate determined byclinicians or in an automatic mode. The automaticmode operates by activation of a switch signaling afull blood sac. This switch initiates the filling of thepump housing with compressed air, which forces theblood out of the blood sac, and ejection commences.When ejection is complete, vacuum is applied andfilling of the blood sac begins. Fill to empty is a termthat is used to describe this method of operation. Thepreferred trigger for pump ejection is when the pumpreaches full fill. It can be run at a fixed rate, but it ispreferred to have the blood sac fully fill to prevent thepossibility of thrombus formation with in the bloodsac. Any LVAD that operates in such a method will beasynchronous with the native heart rate and willincrease or decrease its rate on the basis of the rate offilling. In this way, the pump rate can increase toaccommodate the patient’s activity level and increasedrate of venous return to the heart. Pump rates typi-cally decrease when the patient is at rest. Any pumpusing a fill to empty mode will be totally dependenton the patient’s preload status for adequate pumpoutput and support of systemic circulation.

Although the Thoratec pump is pneumatic, elec-trical power must be available at all times to run thecompressors. A portable console that can run thepump on battery power is available. Each fullycharged battery with the portable console can sup-port a patient on an LVAD for up to 80 minutes.With a pneumatic system, ventricular assist can bemaintained utilizing manual pumping to compressthe blood sack in the event of a console malfunction.Patients being maintained on the portable systemcan be discharged from the hospital to wait fororgan donation. The Thoratec VAD system wasapproved for use as a bridge to transplant device in1995.14

TABLE 1 Risk Factors for Mortality After LVADImplant*

■ Mechanical ventilation■ Previous LVAD or RVAD used to bridge to a long-term VAD■ Postcardiotomy shock■ Redo surgery■ CVP � 16 mm Hg■ Prothrombin time � 16 s■ Ischemic cardiomyopathy

*LVAD indicates left ventricular assist device; RVAD, right ventricularassist device; VAD, ventricular assist device, and CVP, central venouspressure.

FIGURE 1. Placement of the Thoratec left ventricular assistdevice is illustrated depicting the paracorporeal placementwith the pump outside the body and with the inflow andoutflow cannula traversing subcostally to the left ventricleand ascending aorta respectively. This is an example ofbiventricular support. Reprinted with permission fromThoratec Corp, Pleasanton, Calif.

Implantable Left Ventricular Assist Devices S25

A newer version of this pump is the ThoratecInternal Ventricular Assist Device (IVAD), which is asmaller version of the same pump, allowing it to beimplantable. A wire for the electrical signal of “full fill”and the pneumatic hose remains external. The princi-ples of pump operation are the same. The advantagesare increased patient mobility, esthetics, and the reim-bursement category for implanted pumps. Also, theesthetics of living outside the hospital with animplanted pump is more appealing. This particularpump was approved for use by the FDA in August2004.

HeartMate VE and XVE Pump

The HeartMate VE or XVE is a totally implantableelectrically driven pulsatile LVAD. The pump is mostoften positioned preperitionally in the left upperabdominal quadrant or alternatively intraabdomi-nally. Implantation requires a median sternotomythat extends to the umbilicus and the use of car-diopulmonary bypass. This 2-compartment surgeryincreases the risk of infection, which is greater in thiscase than when the device is isolated to the chestcavity alone. The pump is made of medical gradetitanium. The inflow cannula is positioned in theapex of the left ventricle, bringing blood into thepump. Blood is returned to the ascending aorta fromthe outflow cannula of the pump. Constructedporcine tissue valves maintain unidirectional flow. Apercutaneous lead from the pump is tunneled subcu-taneously to exit the body in the right upper quad-rant just above waist level. The lead carries electricalenergy to the pump and brings data from transducerslocated in the pump, alerting clinicians or the patientto changes in pump rate, stroke volume, or pumpoutput. This pump has a flow rate capability inexcess of 10 liters per minute.15

As the pump fills, air must be displaced from thepump housing to the outside and must reenter as airspace is created during pump ejection. The air ventthat facilitates these maneuvers can also be attachedto a hand pump for emergency manual pumping ifthe electrical mechanism fails. If the situation is unre-pairable, these pumps can be converted to pneumat-ically driven systems requiring that dischargedpatients be readmitted to the hospital. Figure 2 illus-trates the design and positioning of the HeartMateVE LVAD. Figure 3 is a picture of both the Thoratecand HeartMate Pumps.

HeartMate’s percutaneous lead is covered with aDacron material that will allow for tissue ingrowththat, when intact, will protect the patient from infec-tion. The lead is connected to a system controller,which runs and monitors the pump for alarm condi-tions. The controller is attached to a source of AC

power or 2 portable lead acid batteries that are capa-ble of 4 to 6 hours of untethered support. Controllerscan be clipped to a belt and batteries can be carriedin a shoulder holster or waist pack. External compo-nents are relatively light weight, which allows thepatient a broader range of activities than the Thoratecpump.

FIGURE 3. Thoratec and HeartMate sites.

FIGURE 2. This illustration depicts the implantableHeartMate pump with the wearable controller and battery.Reprinted with permission from Thoratec Corp, Pleasan-tation, Calif.

S26 Journal of Cardiovascular Nursing ❘ September/October 2005

As with the Thoratec, the HeartMate can either beoperated at a fixed rate or in the automatic mode inwhich ejection occurs when the pump is fully filled.Pump rates in the automatic mode are determined byhow quickly the blood pump is filled by the left ven-tricle. A unique characteristic of the HeartMate is thetextured interior surface of the blood pump thatencourages endothelial cells to cover the surface cre-ating a biologic lining. As a result of this smooth bio-logic lining interfacing with the blood, patients donot need to be maintained on warfarin for anticoag-ulation as do all other with LVAD devices. Antiplatelettherapy is all that is required.16

Novacor LVAS

Like the HeartMate, the Novacor pump is animplantable, electrically driven, pulsatile LVAD. It isone of the most reliable systems of circulatory sup-port.17 In fact, it comes with a 3-year warranty andit is the only pump to offer such a warranty. Long-term support capability has been demonstrated. ByKaplan-Meier analysis, excluding percutaneous leaddamage, more than 86% of pumps are in service at 3 years.18 With long-term use, eventual wearing ofmechanical parts and the porcine tissue valves canoccur, requiring elective replacement. In 10 patientswho have lived longer than 3 years on pump support,3 patients required pump replacement.18 TheNovacor is still in the process of acquiring FDAapproval for destination therapy use in the UnitedStates. The positioning of the pump in the left upperquadrant of the abdomen is similar to the techniqueof the HeartMate VE placement with the same con-nections to the left ventricle and aorta. A percuta-neous lead is also used to bring electrical power tothe pump, to send hemodynamic information fromthe pump to the system monitor and to provide aroute for airflow in and out of the pump housing.Figure 4 illustrates the Novacor pump placement.Pump flow rates of 10 liters per minute are achiev-able with this LVAD.

Modes of operation are similar to the HeartMate aswell. It can either be programmed to run in a fixedrate mode or in a fill to empty mode. The rate of bloodfilling the pump determines when ejection occurs inthe later mode. When it is set to trigger when fillingdecreases by 100%, it runs in fill to empty mode. Thatpercentage can be decreased to allow for synchroniza-tion with the native heart systole. In a bridge to trans-plant patient or eventually a destination therapypatient, that feature is not necessary. It might be usedin a patient who might potentially be weaned from thepump. Explantation of LVADs following myocardialrecovery after LVAD support has been reported insmall numbers of patients.19–21

The Novacor pump is currently being evaluated asa destination therapy device under an FDA con-trolled trial. The trial, Investigation of Non-Transplant Eligible Patients who are InotropeDependent or INTrEPID, is completed but the resultsare yet to be released.

Unlike HeartMate, the Novacor requires antico-agulation with heparin initially followed by conver-sion to warfarin sodium. In addition, antiplateletagents are required to prevent thrombotic events andtransient ischemic phenomena. These events are mostlikely to occur in the first 30 days postimplant.17

Initially, stroke was a greater problem in the NovacorLVAS but modifications in the outflow graft havebeen made. As a result, the incidence of these adverseevents is similar in both pumps.17

Like the HeartMate pump, the batteries and con-troller can be carried or worn by the patient. Patientscan be untethered from AC power for up to 5 hoursper battery. Battery support duration is only limitedby the number of charged batteries a patient has inhis or her possession.

ComplicationsThere are several postoperative complications to bevigilant for and are seen in all of the previously

FIGURE 4. An implanted Novacor left ventricular assistdevice is depicted showing pump placement and the wear-able controller and batteries. Reprinted with permissionfrom World Heart, Inc, Oakland, Calif.

Implantable Left Ventricular Assist Devices S27

described LVAD populations. The causes of deathfollowing device implant are primarily multiorganfailure, renal failure, right ventricular failure, andinfection.10 Bleeding and thrombembolism are 2other important complications causing postoperativemorbidity. Patient selection and timing of the deviceimplantation appear to be critical factors that deter-mine successful outcome.10 As data on complicationsand patient characteristics accumulate, risk factorsfor these complications will be better understood andoutcomes should improve over time.

Multiorgan Failure

Patients with end-stage heart failure have many pre-operative conditions that may contribute to the devel-opment of multiorgan failure postoperatively.Preoperative low perfusion may be evident and maypersist postoperatively after the insult of cardiopul-monary bypass and the surgery itself. Preoperatively,patients may be supported with mechanical ventila-tion and may already have some degree of pulmonarydysfunction. Often, creatinine, blood urea nitrogen,and bilirubin levels are elevated, suggesting renal andhepatic dysfunction. Nutrition is often inadequateand is very likely an underestimated risk factor inmechanical support.10 The more severe the level ofsecondary organ dysfunction is preoperatively, thegreater the risk of postoperative multiorgan failure.Multiorgan failure accounted for 27% of deathsreported to the ISHLT circulatory support database.5

Right Ventricular Failure

Right ventricular failure is a concern in any patientwho has elevated pulmonary and central venous pres-sures in the preoperative setting. The hallmark of rightventricular failure in the postoperative period will beelevated central venous pressures, an empty left ventri-cle, and a decrease in device outputs.15 These patientswill require right ventricular unloading with vasodila-tors and/or nitric oxide and prolonged inotropic sup-port of the right ventricle. Given time, the right ventri-cle usually recovers as left sided unloading remainsconstant with the assist of the LVAD. An occasionalpatient may require temporary assist from a short-termright ventricular assist device. Continual biventricularsupport might be employed in a bridge to transplantpatient with either the implanted or paracorporealThoratec but this would not be a suitable option for adestination therapy patient. That type of patient wouldneed to utilize a temporary support device.

Bleeding

Postoperative bleeding is common and there aremany factors that contribute. Prolonged cardiopul-

monary bypass time, anticoagulation, and extensivesurgical dissection, especially with the Heartmate VEand Novacor, will contribute to increased risk ofbleeding. Preoperative hepatic dysfunction will alsoadd to the risk of coagulopathy. Patients who havehad previous cardiac surgical interventions will bemore prone to bleeding because of excessive scarringin the mediastinum and the need for more surgicaldissection. Excessive bleeding has been as frequent as20% and 50% with the Thoratec, HeartMate, andNovacor devices but has decreased with greaterdevice experience.15 The ISHLT database reports a27.8% incidence of postoperative bleeding.5

Infection

Infection is a serious complication that occurs fre-quently. Patients are vulnerable to the usual postoper-ative infections seen in any postsurgical patient suchas line sepsis, pneumonias, and urinary tract infec-tions.15 These infections may have been present pre-operatively because of the debilitated nature of thepatients’ conditions. Device-related infections canoccur in the percutaneous driveline, the pump pocket,from the pump in the form of endocarditis, or in theblood stream. Percutaneous driveline infections arethe most common. The most common pathogens arestaphyloccus, pseudomonas, enterococcus, and can-dida.22 Pocket infection rates range from 11% to24% for the HeartMate and Novacor LVADs, witheven greater rates of driveline infections, which rangefrom 18% to 30%.23 The ISHLT circulatory supportdatabase reports a 32.5% incidence of infection andthat infection is responsible for 7% of deaths for alldevices in the database.5 Infection does not present anobstacle to successful transplantation as long as it isidentified and appropriately treated. In theREMATCH trial, the probability of infection of theLVAD device was 28% within 3 months of implantand was the leading cause of death.8 Once deviceinfection occurs, it can usually be controlled for someperiod of time but often reoccurs, causing late mor-tality and morbidity.22

As we enter the era of destination LVAD therapy,device-related infection is a major concern. In per-manent implants, chronic infection will decreasequality of life, increase the cost of care, and mayrequire replacement of the device.24 Along with theusual practices to prevent infection, steps to preventinoculation of the pumps and components in theoperating room include limiting room traffic, avoid-ing opening the LVAD sterile packaging prematurely,assembling the pump in a low traffic area, usingdummy pumps to size and fashion the abdominalpockets, using antibiotic-soaked sponges to covercomponents prior to implant, using antibiotic

S28 Journal of Cardiovascular Nursing ❘ September/October 2005

irrigations prior to closure immobilizing the percuta-neous lead with a secure dressing, and abdominalbinder prior to leaving the operating room.22

Immobilizing the percutaneous lead, power cables,or drivelines will be crucial to the healing of the site.Tissue ingrowth into the velour covering of the leadsor drivelines will take about 10 to 14 days to be com-plete. If the sealing of the track is disrupted by move-ment or tension on the lead or driveline during thistime, it may never heal completely. Incomplete heal-ing will allow microorganisms to be able to track upthe tunneled lead to the pump pocket or the medi-astinum. Meticulous dressing changes need to be per-formed daily and their importance emphasized topatients and caregivers as they begin to assume theirown care.

Thromboembolism

Death from stroke accounts for 10% of all deaths inthe ISHLT database.5 Neurologic dysfunctionoccurred in 14% of patients.5 All of these deviceshave a blood-device interface that harbors the poten-tial for thrombus formation. With the exception ofthe HeartMate, all of the described devices requireanticoagulation with warfarin and antiplateletagents. The HeartMate’s biologic lining, describedearlier, precludes the need for anything other thanantiplatelet therapy.16 The Novacor device has previ-ously been associated with higher rates of stroke thanthe other devices. However, modifications in the out-flow graft have lead to a thromboembolic rate simi-lar to the HeartMate device.17 A recent studyreported the incidence of thromboembolic eventsbefore modifications at 23% compared to an inci-dence of 6% after modifications.17 In a large multi-center trial of 280 patients supported with theHeartMate VE, thromboembolic events occurred in12% of patients; 6% were deemed device related.16

Neurologic impairment following thromboembolicevents range from transient central nervous system orneuromuscular disorders to permanent disability.

Device Malfunction

It is not unexpected that some degree of device fail-ure or malfunction occurs over time with mechanicaldevices. Moving mechanical parts incur wear. Motorfailure, wear on bearings, wear on valves, and break-ing of external components because of wear and tearby patients have all contributed to device failure andmalfunction. The probability of device failure by 24 months in the REMATCH trial was 35% requir-ing replacement of the device in 10 patients.8 In thebridge to transplant population at ColumbiaUniversity, device malfunction that was considered

clinically significant only occurred in 5.7% of thesingle lead HeartMate VE patients (N � 174), with amean support duration of 65.3 days.25 In a report of277 patients (137 HeartMate VEs, 81 pneumaticHeartMates and 57 Novacors), there were 21 devicefailures only one of which was a Novacor.26 TheNovacor has demonstrated a very high degree of reli-ability and durability and impending wearing out ofthe pump can be diagnosed at least 2 months beforeanticipated potential failure.11 This wear on pumpswill become a problem that will need to be furtheraddressed in the destination therapy patients.

Discharging Patients From the Hospital

Following transfer from the intensive care unit,patients and caregivers begin training for eventualdischarge from the hospital. Outpatient care requiresreliable compliance by the patient and caregiver, con-tinued education with constant reinforcement, andadequate anticoagulation monitoring.27 Table 2 out-lines teaching topics unique to LVADs that patientsmust understand and be competent in before dis-charge can occur. These issues are crucial because alack of understanding or an inability to demonstrateproper device management could lead to morbidityand possibly endanger a patient’s life. It is recom-mended that a teaching documentation tool with spe-cific competencies be used to track patient and care-giver progress.28 Patients receive written informationin the form of patient handbooks from the variouscompanies. Teaching the patient to become inde-pendent occurs on a daily basis with instruction byLVAD coordinators and continual reinforcement bythe unit staff nurses. To reinforce the importance of

TABLE 2 Preparation for Discharge: PatientTraining and Education Topics Unique to LVADTherapy*

■ Components of the LVAD system and their functions■ Care and maintenance of the equipment■ Driveline care and dressing changes■ Battery management■ Safety precautions■ Showering techniques■ Response to alarms■ Response to emergencies■ Execution of hand pumping (Thoratec and HeartMate

only)■ Changing controllers and cables■ Conditions requiring notification of the physician or device

coordinator■ Anticoagulant monitoring■ Use of the medical center paging system■ Dealing with household power failures■ Contact with emergency medical services

*LVAD indicates left ventricular assist device.

Implantable Left Ventricular Assist Devices S29

what is being taught, some centers use signed con-sents or contracts to emphasize to the patients andcaregivers the importance of following the protocolsor instructions.29,30

In preparation for eventual discharge, patients aregradually introduced to experiences outside the hos-pital. Initially, they are accompanied outside the hos-pital for various excursions by professional staff. Asthey become more confident and competent, they areencouraged to make out-of-hospital excursions inde-pendently. Some institutions encourage an overnightstay out of hospital prior to discharge. With dedi-cated, daily teaching and reinforcement, patientswho do well from a medical point of view and haveadequate caregiver support can be discharged asearly as 14 days postimplant.28 Many institutionsrequire that patients reside within a limited radius ofthe hospital, which is usually a 2-hour drive, whileother institutions will allow patients to go home evenif that distance exceeds a 2-hour limit. In this case,the patient and caregiver must reside locally, demon-strate competence, and have no outstanding medicalissues before being allowed to live outside a closeradius to the medical center. It is important that themedical center personnel advise the patient’s localphysician and emergency medical personnel of anLVAD patient’s presence in the community. Trainingand education of these individuals is recommendedin the event of emergencies in the community.28

These individuals include staff in the local emergencydepartment, the fire department, and paramedics andemergency medical technicians who are likely torespond to an emergency in the home.

As we enter the era of destination therapy, we willsee more older patients receiving LVAD implantsbecause transplantation is not an option for them.Older age has some potential impacts on patient teach-ing and discharge planning. Older patients may needmodification in the discharge teaching plan because ofcognitive impairment, physical limitations such asarthritis, and an inherent discomfort with technology.28

These patients may need a 24-hour companion whereyounger patients do not. This requirement will putmore strain on the family or caregivers and plans willneed to be in place to address this issue.29

ReimbursementThe CMS have recently made significant changes in thereimbursement for implantable LVADs and for theapproved external LVADs. These figures are containedin the Medicare Hospital Inpatient ProspectivePayment System Final Rule for fiscal year 2005 (www.cms.hhs.gov/providers/hipps/frnotices.asp).

Basically, the new rule, effective October 1, 2004,improves payment for destination therapy and bridge

to transplant. CMS has reclassified implants ofimplantable pulsatile LVADs from diagnosis relatedgroup (DRG) 525 to DRG 103. This is significantbecause DRG 525 previously reimbursed substantiallyless than DRG 103, which is the category for hearttransplantation. The LVAD implant surgery and sub-sequent recovery is often more complex and associ-ated with a longer length of stay, both in the intensivecare unit and the hospital in general. Now, patientswho are discharged longer than 30 days after LVADimplant and readmitted for a later heart transplant areable to be classified as DRG 103 a second time, allow-ing for a second payment for the transplant surgeryand hospitalization. The base reimbursement rates forDRG 525 are $56,500 and $97,200 for DRG 103.

Examples of how reimbursement would occurwith different types of LVADs can be illustrated usingthe Thoratec extracorporeal device versus animplantable device. A Thoratec extracorporeal LVADwould fall into category 525 but if the heart trans-plant occurs during the same hospital admission,reimbursement would be based on the transplantDRG 103. If the Thoratec patient is discharged formore than 30 days and readmitted for transplant, thereimbursement rate for the first admission would bebased on DRG 525 and on DRG 103 for the secondadmission. A patient who receives an implantedLVAD would be classified under DRG 103. If theheart transplant occurs during the same admission,reimbursement for both surgeries is based on thatone DRG. If the same patient is discharged with theimplanted LVAD for greater than 30 days and read-mitted for a later heart transplant, the institution willbe reimbursed based on 2 DRG 103s. These changeswill likely open the door for wider use of LVAD ther-apy by more institutions.

ConclusionAlthough an attempt has been made to present anobjective overview of the devices described in thearticle, the current methods of recording, analyzing,and publicizing information regarding adverse eventsand outcomes are often not standardized sufficientlyto allow for meaningful comparisons of the devices.Pooling of these data in an international databasewhere standardized definitions are employed shouldhelp with future, more accurate comparisons.

We are on the verge of seeing a variety of end-stage heart failure patients being supported on long-term assist device therapy. Many pumps havenow or will shortly be proven to be effective and reli-able for long-term support of these patients with anacceptable complication rate. Because of changes inreimbursement rates, the use of these devices willbecome more widely accepted. We still have a lot to

S30 Journal of Cardiovascular Nursing ❘ September/October 2005

learn about appropriate candidate selection andwhen we do, the therapy will become more cost-effective. This patient population represents anopportunity for advanced practice nursing to step inand provide direction for this new population ofindividuals who will need chronic care and support.

REFERENCES1. Jessup M. Mechanical cardiac-support devices-dreams

and devilish details. N Eng J Med. 2001;345(20):1490–1493.

2. deRose JJ, Argenziano M, Sun BC, Reemtsa K, Oz ME.Implantible left ventricular assist devices—An evolvinglong-term cardiac replacement therapy. Ann Surg. 1997;226:461.

3. Bank AJ, Mir SH, Nguyen DQ, et al. Effects of left ven-tricular assist devices on outcomes in patients undergoingheart transplant. Ann Thorac Surg. 2000;69:1369–1375.

4. Bond AE, Nelson K, Germany CL, Smart AN. The leftventricular assist device. Am J Nurs. 2003; 103(1):32–40.

5. Deng MC, Edwards LB, Hertz MI, et al. Mechanical circu-latory support device database of the international societyfor heart and lung transplantation: second annual report—2004. J Heart Lung Transplant. 2004;23(9):1027–1034.

6. Copeland JG. Multicenter bridge to transplantation withthe HeartMate assist device: evaluation from another per-spective. J Thorac Cardiovasc Surg. 2003;125(2):228–230.

7. Portner PM, Oyer PE, McGregor CGA, et al. First Humanuse of an electrically powered implantable ventricularassist system. Artif Organs. 1985;9:36–50.

8. Rose ER, Gelijns AC, Moskowitz AJ, et al. Long-term useof a left ventricular assist device for end-stage heart fail-ure. New Eng J Med. 2001;345(20):1435–1443.

9. Aaronson KD, Himanshu P, Pagani FD. Patient selectionfor left ventricular assist device therapy. Ann Thorac Surg.2003;75(6):529–535.

10. Miller LW. Patient selection for the use of ventricularassist devices as a bridge to transplantation. Ann ThoracSurg. 2003;75:S66–S71.

11. Deng MC, Loebe M, El-Banayosy A, et al. Mechanical cir-culatory support for advanced heart failure: effect ofpatient selection on outcome. Circulation. 2001;103:231–237.

12. Rao V, Oz MC, Flannery MA, Catanese KA, ArgenzianoM, Naka Y. Revised screening scale to predict survivalafter insertion of a left ventricular assist device. J ThoracCardiovasc Surg. 2003;125(4):855–862.

13. Hunt SA, Frazier OH, Myers TJ. Mechanical circulatorysupport and cardiac transplantation. Circulation. 1998;97:2079–2090.

14. Farrer DJ. The thoratec ventricular assist device: a para-corpeal pump for treating acute and chronic heart failure.Semin Thorac Cardiovasc Surg. 2000;12:243.

15. Kukuy EL, Oz MC, Haka Y. Long-term mechanical circu-latory support. In: Cohen LH, Edmunds LH Jr, eds.Cardiac Surgery in the Adult. New York: McGraw-Hill;2003:1491–1506.

16. Frazier OH, Rose ER, Oz MC, et al. Multicenter clinicalevaluation of the HeartMate vented electric left ventricu-lar assist system in patients awaiting heart transplanta-tion. J Thorac Cardiovasc Surg. 2001;122(6):1186–1195.

17. Strauch JT, Spielvogel D, Haldenwang PL, et al. Recentimprovements in outcome with the Novacor left ventricularassist device. J Heart Lung Transplant. 2003;22(6):674–680.

18. Novacor LVAS Physicians’ Manual. Oakland, CA: WorldHeart Inc; 2004:7-111-7-12.

19. Frazier OH, Myers TJ. Left ventricular assist system as abridge to myocardial recovery. Ann Thorac Surg. 1999;68:734–741.

20. Hurtzer R, Muller J, Weng Y, Wallukat G, SpiegelsbergerS, Loebe M. Cardiac recovery in dilated cardiomyopathyby unloading with a left ventricular assist device. AnnThorac Surg. 1999;68:742–749.

21. Young JB. Healing the heart with ventricular assist devicetherapy: mechanisms of cardiac recovery. Ann ThoracSurg. 2001;71:S210–S219.

22. Holman WL, Rayburn BK, McGiffin DC, et al. Infectionin ventricular assist devices: prevention and treatment.Ann Thorac Surg. 2003;75:S48–S57.

23. Minami K, El-Banayosy A, Sezai A, et al. Morbidity andoutcome after mechanical ventricular support usingThoratec, Novacor and HeartMate for bridging to hearttransplantation. Artif Organs. 2000;24:421–426.

24. McCarthy PM, Smedira NO, Vargo RL, et al. One hun-dred patients with the HeartMate left ventricular assistdevice: evolving concepts and technology. J ThoracCardiovasc Surg. 1998;115:904–912.

25. Morgan JA, John R, Rao V, et al. Bridging to transplantwith the HeartMate left ventricular assist device: theColumbia Presbyterian 12-year experience. J ThoracCardiovasc Surg. 2004;127(5):1309–1316.

26. Navia JL, McCarthy PM, Hoercher KJ, Smedira NG,Banbury MK, Blackstone EH. Do left ventricular assistdevice (LVAD) bridge-to-transplantation outcomes predictthe results of permanent LVAD implantation? Ann ThoracSurg. 2002;74(6):2051–2062; discussion 2062–2063.

27. Schmid C, Hammel D, Deng MC, et al. Ambulatory careof patients with left ventricular assist devices. Circulation.1999;100:II-224.

28. Andrus S, Dubois J, Jansen C, Kuttner V, Landsberry N,Lukowski L. Teaching documentation tool: building a suc-cessful discharge. Crit Care Nurse. 2003;23(2):1–10.

29. Bond EA, Bolton B, Nelson K. Nursing education andimplications for left ventricular assist device destinationtherapy. Prog Cardiovasc Nurs. 2004;19(3):95–101.

30. Moroney DA, Powers K. Outpatient use of left ventricularassist devices: nursing, technical and educational consid-erations. Am J Crit Care. 1997;6(3):355–362.