| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2003;75:S48-S57
© 2003 The Society of Thoracic Surgeons
a Departments of Surgery and Medicine, University of Alabama at Birmingham, Birmingham, Alabama, USA
* Address reprint requests to Dr Holman, Department of Surgery, University of Alabama at Birmingham, Birmingham, AL 35294-0007, USA.
e-mail: wholman{at}its.uab.edu
Presented at the Heart Failure & Circulatory Support Summit, Cleveland, OH, Aug 22–25, 2002
Abstract
Infection is one of the most important challenges to the use of implanted mechanical circulatory support systems (MCSS), particularly as we enter the era of permanent device use in patients who are not candidates for cardiac transplantation. This paper describes the pathogenesis of MCSS infection, with particular attention to the role of biofilm-forming bacteria. Suggestions are presented for the prevention and treatment of infections in implanted MCSS.
During the past 20 years, mechanical circulatory support systems (MCSS), including total artificial hearts (TAHs) and ventricular assist devices (VADs), have made numerous advances. Several VAD designs have been approved by the federal Food and Drug Administration (FDA) for use as a bridge to cardiac transplantation or pending recovery of the native heart. The recently completed Randomized Evaluation of Mechanical Assistance for the Treatment of Congestive Heart Failure (REMATCH) Trial was a landmark event in MCSS development. This randomized prospective trial showed that the survival of patients with severe heart failure who were not transplant candidates was significantly and substantially better with left VAD support than with optimal medical therapy [1, 2]. However, infection has been a major complication of VAD use since its inception and currently poses an important challenge to the cost-effective long-term use of MCSS [3].
This paper reviews existing information regarding infection as a complication in mechanical circulatory support. The pathogenesis of device infection is discussed, then suggestions are made for the prevention and treatment of device infections. Finally, novel therapies are presented that in the future may provide useful tools for preventing or treating device infection.
Infectious complications in previous reports of MCSS
Reports of VADs and TAHs that mention infection as a complication (Table 1) are numerous yet difficult to summarize due to a lack of uniformity in definitions and statistical analysis. Definitions describing infection vary in several ways. First, the diagnosis of infection may depend on a positive culture alone or on the concomitant presence of a clinical syndrome (eg, fever with leukocytosis) in a patient who receives antimicrobial therapy. Descriptions of infection location likewise vary from trial to trial. Nondevice-related infections (eg, pneumonia or urinary tract infection) are generally grouped together. Device-related infection may include only the blood-contacting surface of a pump and the outer surface of an implanted pump, or it may also include infections of the percutaneous driveline.
|
|
|
There are several other areas that remain to be investigated and that may lead to important improvements in device infection rates. These include research into the effect of heart failure, malnutrition, and cardiac cachexia on host immune defenses, as well as the effect of mechanical pumping and the pump surface itself on the immune system. New information concerning the mechanisms for bacterial adherence to artificial surfaces, biofilm formation, and antimicrobial resistance may likewise prove useful in the future. However, for the present time, prevention of infection by meticulous surgical technique and wound care remains the best method for controlling infectious complications.
Pathogenesis of device infection
As mentioned in the previous section, the majority of intracorporeal device infections are caused by biofilm-forming bacteria and fungi (eg, Staphylococcus sp.). There has been considerable research in this area, and Costerson’s group at the Center for Biofilm Engineering at Montana State University has been a leader in increasing our knowledge of biofilms and device infection. Readers are referred to their publications for additional information [12].
Biofilms (Figs 3, 4) are defined by Costerson as "a structured community of cells enclosed in a self-produced polymeric matrix and adherent to an inert or living surface" [12]. Understanding the development of a biofilm colony from individual microorganisms is important to understanding the prevention and treatment of most device infections. Bacteria or fungi that exist in isolation from one another have a distinct phenotype described as the planktonic form of the microorganism. Planktonic microbial forms that come in contact with the surface of an implanted device can sense the presence of the artificial surface, adhere to this surface, and multiply to form microcolonies. Adhesive substances are secreted as the microcolony grows in size, anchoring the organisms in place. As the density of microbes on the artificial surface increases further, biochemical signals that are sent into the immediate environment are recognized by other organisms of the same species. In the case of Pseudomonas aeruginosa, these signaling molecules appear to be acylhomoserine lactone [12] or DNA [13]. The detection and response to other nearby organisms is termed "quorum sensing," and it is important because quorum-sensing leads to a series of phenotypic changes in individual microorganisms of the biofilm that render the colony resistant to host defenses and antibiotics. Specifically, the organisms in the biofilm secrete a layer of slime (typically a polysaccharide) and downregulate their metabolic activity. The biofilm serves as a physical barrier to attack from leukocytes, antibodies, complement, and antimicrobial drugs. Moreover, the downregulation of microbial metabolism renders the organisms less sensitive to many antimicrobial drugs.
|
|
The study of antimicrobial resistance is another interesting and potentially important field of research. In the case of one common biofilm pathogen, Staphylococcus, the mechanisms for ß-lactam antibiotic resistance are being elucidated [18, 19]. The mechanisms of antibiotic resistance to penicillin and other drugs are complex but, if understood, may lead to new ways to reverse this resistance, thereby substantially broadening the choice of effective drugs for treating infections. Unfortunately, as it stands now, bacteria and fungi seem to be winning the race as new multiple-drug resistant organisms are discovered, including vancomycin-resistant strains of Staphylococcus [20]. The indiscriminate use of antibiotics inside and outside of the hospital may accelerate this process [21].
Is device infection avoidable?
As described in the previous section, certain bacteria and fungi (eg,Staphylococcus and Candida) that enter the wound at the time of surgery are able to survive on the surface of implanted materials for extended periods of time despite the presence of host defenses and systemic prophylactic antibiotics, then cause a septic syndrome weeks or months later. Furthermore, small cutaneous infections (eg, percutaneous driveline infections) or other infections are capable of seeding an implanted MCSS with bacteria or fungi that later cause sepsis due to infection of the outer or inner (blood-contacting) surfaces of an MCSS. Implanted pumps present a large area for microbial colonization and the host is often immuno-compromised. Given these risks, is it reasonable to believe that infection is avoidable?
We believe that available information and related surgical experiences indicate that most device-related infections are preventable. First, if one estimates the incidence of device-related infection from published reports of the prevalence, it becomes apparent that the actual incidence of device-related infection has decreased over the past decade as the mean and median durations of circulatory support increased. The clinical course of patients supported for more than 1 year corroborates this idea. There are patients at several institutions who have survived for more than 1 year (in some cases, 2 or 3 years) on circulatory support devices as a bridge to transplant or as destination therapy without an episode of sepsis or a device related infection.
The REMATCH Trial of destination therapy did not include a sufficient number of patients to demonstrate site-specific differences in the incidence of sepsis with a high level of statistical confidence. However, data from the REMATCH Trial suggested that this might be the case; thus, the methods for avoiding infection at low-infection centers are worth emulating. Experience with another large implanted device, the internal cardioverter-defibrillator (ICD), is analogous to the situation with implanted VADs and provides a stronger rationale for precisely following the protocols of successful institutions. In 1996, Shepard and Epstein published their results from 251 consecutive ICD procedures, including implants using thoracotomy as well as nonthoracotomy approaches. The authors’ literature review found a cumulative infection rate of 3.5% (95% CI, 2.7% to 4.5%) for thoracotomy ICD implants and 2.1% (95% CI, 1.4% to 3.3%) for nonthoracotomy implants. The authors’ infection rate of 0% (95% CI, 0.0% to 1.9%) is significantly lower than the published infections rates, and indicates that an institutional commitment to infection prevention can result in fewer infections.
Prevention of infection in MCSS
Overview
Many things that occur throughout the course of mechanical circulatory support influence whether or not a patient develops an infection. The present discussion focuses on device-related infections and wound infections, and not other infections (eg, pneumonias or urinary tract infections). The section is organized into measures that can affect the incidence of infection during the preoperative, intraoperative, and postoperative time periods. Device design issues that influence the incidence of infection are also addressed.
Preoperative measures
Risk factors that are present before device insertion include preoperative infection, malnutrition, immunosuppressive medications (eg, steroids), mechanical ventilation, and the presence of indwelling catheters (especially for more than 5 to 7 days). It is helpful to correct as many of these conditions as possible before implanting an MCSS. Preoperative infection may be documented by positive cultures, which should be used to guide therapy. Any infections should be cleared before device placement if the wait that is involved does not endanger the patient’s life, and a septic patient should not be taken to the operating room. Infection is often the cause for decompensation of a chronic stage IV heart failure patient, and mechanical support is not the best answer to the high-flow/low-resistance state of sepsis.
A more difficult situation is posed by patients with profoundly impaired circulation who develop a syndrome of acute inflammation that is not related to infection (systemic inflammatory response syndrome) [22]. The leukocyte count in this syndrome is elevated with numerous immature forms (ie, a left shift), yet cultures are negative. These patients are best treated by urgent VAD placement with coverage by routine prophylactic antibiotics. Because this inflammatory state is usually seen only in extreme circulatory insufficiency, waiting to clarify the diagnosis will lead to a worse clinical situation.
The roles of nutrition in affecting outcome for mechanical circulatory support patients and in combating infection have not yet been defined. However, cachexia and inflammation in heart failure are well described [23] and lead to loss of skeletal muscle mass as well as immuno-incompetence. Malnutrition, in particular the cachectic state, is a risk factor for mortality in patients with chronic heart failure independent from New York Heart Association (NYHA) functional class, age, peak oxygen consumption, and left ventricular ejection fraction [24]. The precise risk that cachexia and malnutrition in general pose to patients undergoing a VAD or TAH implant is not known. However, there are numerous anecdotes of patients who have improved circulation yet cannot be rehabilitated, implying that cachexia is an important determinant of outcome. The important question that has not been answered is whether nutritional supplementation either before or after VAD placement can improve results. In the absence of such information, one should assess a patient’s nutritional status periodically and provide supplementation (preferably with enteral feeding) as needed.
Patients become colonized with nosocomial bacteria and fungi during hospitalization. These same microorganisms can infect indwelling tubes and catheters, and cause postoperative systemic infections. The duration of preimplant hospitalization is beyond the surgeon’s control, but indwelling tubes and catheters can be replaced immediately before surgery or soon after device implant to minimize infectious risk.
Skin preparation before surgery should include clipping of hair immediately before surgery and scrubbing with an antimicrobial soap. In the operating room, the patient’s skin is painted with an effective antimicrobial prep that is allowed to dry before placing an adherent plastic drape treated with iodine. The incision is kept above the umbilicus unless the patient is small or very thin and device placement would be impossible without extending the midline incision around the umbilicus. The device pocket must be large enough to prevent wound tension during closure. However, one should avoid an excessively large pocket that leaves space for the collection of blood or other fluids that may become infected.
Prophylactic antibiotics must ensure adequate coverage of Staphylococcus species. Short-term antifungal prophylaxis with a relatively nontoxic drug (eg, fluconazole) should be strongly considered, followed by oral agents (eg, nystatin) if the patient is at risk for oral or esophageal fungal infection [25–27]. The antibiotics suggested for patients in the REMATCH Trial are an example of a staphlyocidal regimen, and include vancomycin, levofloxacin (or trovafloxacin), and rifampin given immediately before the first incision is made.
Nasal carriage of Staphlycoccus aureus has been implicated as a source for S. aureus bacteremia [11] in hospital settings using sophisticated genomic identification techniques. The actual role of nasal S. aureus in device infections is not known, however a study by Perl and associates [28] showed that nasal mupirocin prophylaxis decreased the rate of nosocomial S. aureus wound infections among patients who were nasal S. aureus carriers. Based on the information available to date and the serious consequences of device or wound infection with S. aureus, nasal mupirocin prophylaxis before device implant should be considered.
Intraoperative measures
The most important phase of care for infection prevention is the intraoperative period. Inoculation of the pump or wound with microorganisms at the time of surgery can thwart the best pre- and postoperative care and doom the patient to a device-related infection.
Thorough cleaning of the operating room (terminal cleaning) should take place before the patient enters. The operating room should be equipped with high-efficiency particle arresting (HEPA) filters or the equivalent and have laminar air flow around the operating table. Sealed operating suits as used by orthopedic surgeons may be beneficial, although this has not yet been demonstrated for MCSS implants.
Traffic in the operating room must be strictly limited during the device implant. Any observers should remain in the anesthesia area and not lean over the operative field. They should especially remain well away from the pump itself. Personnel in the operating room should wear headgear that completely covers all hair. Double gloves or a sturdy variety of glove should be used to minimize the possibility of contamination from a glove tear.
The pump is not removed from its sterile packaging until shortly before it is assembled and placed into the patient. After assembly, the pump is filled with saline and covered with laparotomy pads soaked in an antibiotic solution containing vancomycin and gentamicin. Laparotomy pads soaked in antibiotic solution are wrapped around the percutaneous driveline and the outflow graft after it has been preclotted. Laparotomy pads or sterile towels are placed on the edges of the incision so that contact with the cut skin edge is avoided. A dummy pump is used to size the pump pocket so that the implanted pump only passes the incised skin once. The percutaneous driveline is brought out through the right upper quadrant taking care to spare most or all of the right rectus muscle. Sparing the right rectus muscle from injury due to chest tubes or the driveline allows plastic surgeons to mobilize this muscle for wound revision if it is necessary. Serious wound complications fortunately are rare, but they are difficult to manage if tissues for reconstruction are not available or are very limited. Injury to the right rectus muscle can be prevented by tunneling the driveline beneath the rectus muscle and bringing it out at the lateral edge of the rectus. The skin incision should be slightly snug around the driveline, and several surgeons advocate a round rather than linear skin incision. If the exit hole is loose, it should be surgically tightened to create a snug fit that will minimize motion of the driveline in the early postoperative period.
Before leaving the operating room, the percutaneous driveline exit site should be painted with an antimicrobial substance (eg. Betadine, Purdue Pharma LP, Stanford, CN) and held in place with a monofilament skin stitch placed slightly away from the incision. An occlusive dressing is placed that holds the driveline securely in position until a specially modified abdominal binder is fitted to the patient’s percutaneous driveline.
The best possible hemostasis should be obtained before a final irrigation of the mediastinum and pump pocket with sterile saline containing vancomycin and gentamicin. Multiple drains are placed to minimize the possibility of hematoma formation and the wound is closed.
Postoperative measures
Prophylactic antibiotics are ideally discontinued after 48 hours, although there are high-risk situations where they should be continued (eg, a patient who remains on ventilator support with indwelling venous catheters). Chest drains are removed as soon as possible after surgery, leaving closed bulb suction drains in the pump pouch until drainage is less than 50 to 100 mL daily. If there was considerable drainage in the immediate postoperative period, a return to the operating room is indicated to remove clot from the mediastinum and pump pocket. Central lines should receive routine dressing changes and be changed at least every 7 days.
Driveline exit site care during the early postoperative period is performed using aseptic technique and includes a wash with an antimicrobial soap followed by topical treatment with a small amount of topical antimicrobial drug (eg, topical silver sulfadiazine). If healing does not progress satisfactorily, a surveillance culture of the cutaneous exit site may be useful in directing further therapy with topical and systemic antimicrobial agents. Nutrition during this interval is crucial to wound healing, and enteral nutritional therapy may be beneficial in a depleted patient or a patient with poor wound healing. A cachectic patient in a chronic catabolic state may take weeks of nutritional supplementation to return to an anabolic state and begin accreting skeletal muscle mass.
The importance of immobilizing the percutaneous driveline with a belt modified from an abdominal binder cannot be overemphasized to the patient and their family. The patient should wear this at all times, including during sleep. After excellent healing has been accomplished, short periods of time without the belt (eg, when showering) are acceptable, although the patient must avoid sudden motion at the base of the driveline that can tear the delicate interface between the epithelium and the driveline fabric.
In addition to nutrition, the surgeon and staff should aggressively manage serum glucose in diabetics. Maintaining a serum glucose in the range of 100 to 150 mg/dL will decrease the risk of wound infection. In some patients, this level of control will require the use of an insulin infusion.
The blood–artificial surface interaction due to an implanted pump may injure the host immune system [29]. This interaction is thought to cause a T-cell deficit that predisposes the patient to infection, particularly fungal infection. The Columbia group correlated the presence of an implanted left VAD with a higher incidence of fungal infection as compared with comparably ill patients awaiting heart transplantation without VAD support. In view of this information, it is important to periodically examine the patient for signs of fungal infection (eg, thrush or vaginitis) before the infection becomes systemic. Persistent unexplained fever or a very debilitated patient with poor convalescence should prompt blood cultures using a method capable of detecting low levels of Candida species.
Device design
Considerations of infection resistance in pump design have become more important over time as the magnitude of this problem was recognized, especially for permanently implanted devices. There are several approaches to decreasing infection risk by MCSS design.
Axial flow (turbine) and centrifugal VADs are in various stages of development. They do not require as large a pump pocket, as the implanted pulsatile VADs and the smaller pocket is probably less prone to infection. Moreover, the surface area of these pump presents less area for microorganisms to colonize and the absence of valves provides fewer eddy currents to encourage microbial adhesion to the artificial blood contacting surfaces of the pump.
Sealed (ie, completely implantable devices) are now in clinical trials (Fig 5). One putative but important advantage of such a design is absence of a percutaneous driveline that may translate into a lower risk of infection. An alternative approach is the development of drivelines with antimicrobial substances embedded in them [30]. One must be certain that these antimicrobial substances do not also impair normal wound healing.
|
Diagnosis and treatment of device-related infection
Driveline infection
Skin is colonized with a variety of organisms, including Staphylococcus, so that surveillance cultures of the driveline exit site are likely to show microorganisms even in the absence of true infection. The presence of poor healing and local inflammation constitute the diagnosis of driveline infection. This may or may not be associated with systemic signs of infection (eg, fever and leukocytosis). Poor healing, drainage, and local inflammation should be distinguished from the drainage that is due to a deeper wound problem. Such drainage may be a sign of a pocket or deep driveline infection that requires more aggressive surgical treatment.
Healing of an infected driveline exit site may be helped by nutritional supplementation. Antibiotic therapy, both topical and systemic, is beneficial and should be directed at the causative organisms. This includes the use of either hydrogen peroxide diluted in saline or chlorhexidine scrub solutions. In the case of Candida or Staphylococcal infections, specific topical antibiotics should be used (eg, vancomycin diluted in saline for irrigations with Staphylococcal infection). Between treatments, application of a topical antibiotic and application of a dry dressing is beneficial. Surveillance cultures should be performed if steady improvement does not occur. Fungal or Pseudomonas overgrowth may be present, particularly if only anti-Staphylococcus coverage was used.
If the wound appears refractory to healing due to infection at the driveline exit site, replacing the distal portion of the driveline in a new tunnel has been used successfully at several institutions. This places the deeper portions of the driveline at risk for infection, so this maneuver should not be used unless necessary.
Pump pocket infection
Patients with infection of the pump pocket may present with a septic picture or with very few symptoms. The indolent nature of Staphylococcus epidermidis is such that these pocket infections may not occur until months after device placement. The incision is usually intact and not inflammed, although in some instances, a necessitating pocket abscess can be detected by palpation over the incision. Another presentation of a pocket infection is new and persistent drainage from the driveline exit site, especially if the material is sanguinous consistent with an infected hematoma.
We obtain blood cultures in patients who have devices implanted for more than a few weeks if they develop fevers (even low grade) and complain of generally not feeling well. If the cultures are positive for Staphylococcu, then any indwelling lines (eg, chronic central lines) are cultured and may be removed. If the bacteremia persists despite systemic antibiotic therapy, we obtain an ultrasound scan of the pocket. Most pockets have fluid, so this test is of limited benefit. However, if there is very little fluid, it stimulates us to look harder for other sites of infection. If there is copious fluid (especially if a bacteremia is still present despite antibiotics), pocket exploration is undertaken. At the time of exploration, cultures of the fluid are obtained and polymethylmethacrylate (PMMA) beads containing powered antibiotics (typically vancomycin and tobramycin) are left in the pouch. This method adopted from orthopedic surgeons has worked well for us [31, 32]. It is important to check serum levels of the antibiotics to ensure that toxic levels from the bead antibiotics do not occur. A drain is left in the pouch but is removed as soon as possible so that the device will be bathed in high levels of antibiotics. Prefabricated PMMA beads with antibiotics are not commercially available in the United States. The elution characteristics of locally fabricated PMMA beads vary, and in our clinical experience, antibiotic elution lasts about 4 to 8 weeks. We have used chloramphenicol for PMMA beads in unusual situations, and any powered antibiotic that is relatively heat-stable can be considered for use.
Infection of blood contacting surfaces (pump endocarditis)
It can be difficult to distinguish sepsis due to a nonpump source from pump endocarditis. Clues to the presence of an infected pump include septic embolization without evidence of vegetations on the native cardiac valves, new incompetence of the pump inflow or outflow valves in the presence of blood stream infection, and persistence of blood stream infection after adequate antimicrobial therapy and without any other source for infection. The pump housing makes study of the pump using magnetic resonance imaging, computerized tomography, or ultrasound impossible; however, blood samples obtained for quantitative cultures from the left ventricle and the VAD outflow graft may show a difference in the bacterial load. Others have used tagged leukocyte imaging to demonstrate device infection, although routine wound healing and uptake by abdominal organs may obscure radioactivity in the device [33].
Persistent or relapsing bacteremia or fungemia can be successfully managed by cardiac transplantation if the patient is sufficiently stable to undergo this procedure [34–36]. Otherwise, replacement of the entire MCSS is required to treat the infection. This obviously entails substantial risk for death and complications, and is unfortunately the only option that will be available to patients with permanently implanted devices.
Wound infections in patients with implanted MCSS
Patients with serious infections of the midline incision, ranging from exposure of the pump surface to cases of purulent mediastinitis, have been successfully managed in the bridge to transplantation setting [32, 37–39]. This will be a more difficult complication to manage in patients who receive destination MCSS therapy. However, if the device is not exposed, aggressive wound revision and flap coverage may avert more serious wound problems. It is also possible that the addition of omental coverage in some difficult cases of wound infections may at least provide longer palliation from infection in patients with chronically implanted pumps [40]. Seeking the collaboration of a plastic and reconstructive surgeon can provide substantial creative help in these situations.
Comment
Prevention of device-related infection is crucial to the cost-effective use of mechanical circulatory support devices. Preoperative risk factors for device infection are typically difficult to ameliorate or nullify (eg, chronic malnutrition); thus, meticulous surgical technique and postoperative wound care are the most important components of infection prevention. The appropriate and timely use of prophylactic antimicrobial therapy is helpful but cannot eliminate a heavy microbial innoculum that occurs at the time of implantation.
Once device infection occurs, it can usually be controlled for months to years. However, infection often recurs and causes mortality and important morbidity. Published information over the past decade indicates that the incidence of infection is decreasing but the prevalence of infection at some time during the period of mechanical circulatory support remains high at roughly 30% to 40%. The importance of infection was highlighted in the recent REMATCH Trial, in which sepsis was the leading cause of death for patients treated with a left VAD.
The rigorous and uniform application of a protocol for infection prevention, as outlined by the REMATCH investigators, could substantially improve the outcome of VAD patients today. Developments in antimicrobial prophylaxis and treatment of infection together with advances in MCSS design are likely to further reduce device-related infections in the future [49,50].
References
This article has been cited by other articles:
|
|
 |
|
  E. A. Genovese, M. A. Dew, J. J. Teuteberg, M. A. Simon, J. Kay, M. P. Siegenthaler, J. K. Bhama, C. A. Bermudez, K. L. Lockard, S. Winowich, et al. Incidence and patterns of adverse event onset during the first 60 days after ventricular assist device implantation. Ann. Thorac. Surg., October 1, 2009; 88(4): 1162 - 1170. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. F. Padera Jr. and F. J. Schoen Pathology of Cardiac Surgery Card. Surg. Adult, January 1, 2008; 3(2008): 111 - 178. [Full Text]
|
|
|
|
|
|
S. Aggarwal, F. Cheema, M. C. Oz, and Y. Naka Long-Term Mechanical Circulatory Support Card. Surg. Adult, January 1, 2008; 3(2008): 1609 - 1628. [Full Text]
|
|
|
|
|
|
D. D. Yuh, M. Albaugh, S. Ullrich, and J. V. Conte Treatment of Ventricular Assist Device Driveline Infection With Vacuum-Assisted Closure System Ann. Thorac. Surg., October 1, 2005; 80(4): 1493 - 1495. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. W. Stevenson and P. Shekar Ventricular Assist Devices for Durable Support Circulation, August 30, 2005; 112(9): e111 - e115. [Full Text] [PDF]
|
|
|
|
|
|
A. El-Banayosy, L. Arusoglu, M. Morshuis, L. Kizner, G. Tenderich, P. Sarnowski, H. Milting, and R. Koerfer CardioWest Total Artificial Heart: Bad Oeynhausen Experience Ann. Thorac. Surg., August 1, 2005; 80(2): 548 - 552. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Etz, H. Welp, H. H. Scheld, and C. Schmid Near fatal infection of a patient with a left ventricular assist device due to unrecognized fetal death Eur. J. Cardiothorac. Surg., April 1, 2005; 27(4): 722 - 723. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
