A prospective, multicenter study of 145 patients who received ventricular assist devices (VAD) found that infection remains a persistent problem, and that depression and elevated creatinine levels are independent predictors of VAD infection. The study was published online Jan. 11 in Circulation.
Patients with advanced heart failure have few treatment options absent heart transplantation. Such patients may receive VADs to provide support until a donor organ is available (bridge to transplant, BTT), or, for those patients for whom transplantation is not an option, as a destination therapy (DT). Although VADs provide lifesaving therapy for such patients, they are associated with infection. In this study, Rachel J. Gordon, MD, MPH, of Columbia University Medical Center in New York City, and colleagues attempted to describe the epidemiology of VAD infections.
The study included 11 participating centers in the U.S. Patients were eligible if they were at least 18 years of age and approved to receive a VAD as either BTT or DT between 2006 and 2009. The final cohort included 150 patients from 11 centers. The mean age at enrollment was 54 years of age, 84 percent were male and 67 percent were white.
The patients received various models of univentricular assist devices (145 patients) and biventricular support devices (five patients). Most patients received continuous flow devices, but 41 percent received at least one pulsatile device. Participants were followed and monitored for infection until explantation or up to one year. Patients who received replacement VADs remained in the study. When a patient was transplanted, a VAD explanted or at autopsy, the devices were examined for signs of infection and cultured, if indicated.
The population approved for VAD is necessarily very sick, and the patients experienced a large number of adverse events prior to VAD infection or study termination, including major bleeding (44 percent), respiratory failure (21 percent), neurologic events (16 percent), renal failure (12 percent) and other complications. A significant number of patients experienced non-VAD infections (39 percent) including urinary tract infections, bloodstream infections, pneumonias and others.
Post-implantation, patients received clinical infection assessments if infection was suspected, and one of the research team, an infectious diseases specialist, examined all such data, including source documentation. The specialist used several predetermined criteria to confirm infection, including positive cultures from one or more sites, purulent fluid, gross or histopathologic evidence of infection upon device removal or pump pocket exploration, excess drainage from the driveline site, and local signs of infection. In the event of a disagreement between the specialist and the center, a second infectious disease specialist reviewed the data and made the determination.
The researchers found that 22 percent of patients experienced VAD infections (one had two separate VAD infections), with an incident rate of 0.1 VAD infections per 100 patient days. The median time to first infection was 68 days, but infection peaked at 18 days post-surgery. The VAD driveline was the site of the infection in the majority of cases (82 percent) although most often other sites were involved in addition to the driveline.
Gram positive organisms were the cause of most of the infections; coagulase negative staphylococci were present in nine infections, Staphylococcus aureas was identified with seven infections, Psuedomonas aeruginosa was present in five, and five infections were polymicrobial. Candida glabrata caused one infection that led to the patient’s death. The other infections were caused by various Gram negative enteric bacteria.
Patients at eight of the 11 participating centers experienced infections, and there was no association between microbiology and clinical center, time to infection and center or time to infection and organism.
The researchers conducted univariate and multivariable analyses and found that a history of depression (adjusted hazard ratio 2.8) and elevated serum creatinine (adjusted hazard ratio 1.7) were independent predictors of VAT infection. No other past or current medical condition increased infection risk. In addition, no factor related to the surgical procedure, including total bypass time, units of blood products received or type of device had an impact on the rate of infection. Nor did the time spent in ICU, time in hospital between implantation and infection diagnosis or study termination. DT and BTT patients had the same rate of infection.
Patients who experienced infection had higher mortality. Of the patients receiving transplants, 33 percent had experienced a VAD infection and 36 percent had not; 27 percent of those who had infections died compared with 21 percent of those who had not had infections. In the multivariable analysis, infection substantially increased one-year mortality (adjusted hazard ratio 5.6).
The researchers noted that pulsatile devices are generally thought to be associated with increased risk of infection; although the results of this study did not support that hypothesis, it may have been underpowered to reveal an association. They acknowledged that study size was a limitation, as was the fact that depression was measured by history rather than a validated index.
“[T]he VAD population suffers numerous adverse events and has a high severity of illness, but depression, elevated creatinine and elevated bilirubin may increase the chance of infection. Further studies of remediable risk factors such as depression and their treatments are warranted,” the authors concluded.