BACKGROUND: Infection rates for ventricular assist devices (VADs) are improving but the presence of infections remains problematic. Intraperitoneal pump placement may decrease the rate of infection. METHODS: This study is a single-institution, retrospective comparison of infections for an extraperitoneal vs an intraperitoneal position with a polytetrafluoroethylene (PTFE) sheet separating the VAD from visceral organs. Implants were performed by a single surgical team from January 1, 2002 to May 31, 2006 (n = 65) for bridge-to-transplant or destination therapy. Patients with paracorporeal pumps (17 patients) or implant times of <30 days (n = 7; 4 died, 3 ongoing, 0 with pocket infections) were excluded. Thirty-five pulsatile left VADs were analyzed, whereas 6 rotary VADs were excluded from analysis. Twenty-three pulsatile VADs were placed in an extraperitoneal pocket (pocket patients) in 18 patients (5 patients had more than one left VAD implant), whereas 12 patients had intraperitoneal (IP patients) VAD placement. RESULTS: Infection with the implanted VAD occurred in 8 of 23 pocket implants (mean support 310 +/- 249 days) and 0 of 12 IP implants (mean support 220 +/- 131 days; pocket vs IP infections: p = 0.003). Staphylococcus and Enterococcus accounted for most of the infections. Four infections in the pocket group recurred during VAD support. Two infected pocket patients had dehiscence of the mid-line wound. CONCLUSIONS: Infections of implanted VADs are associated with serious morbidity (e.g., prolonged antibiotic therapy, recurrent device infections and wound dehiscence). Placing VADs in an intraperitoneal position with a PTFE barrier to protect the viscera may reduce the prevalence of these infections.
BACKGROUND: Infection rates for ventricular assist devices (VADs) are improving but the presence of infections remains problematic. Intraperitoneal pump placement may decrease the rate of infection. METHODS: This study is a single-institution, retrospective comparison of infections for an extraperitoneal vs an intraperitoneal position with a polytetrafluoroethylene (PTFE) sheet separating the VAD from visceral organs. Implants were performed by a single surgical team from January 1, 2002 to May 31, 2006 (n = 65) for bridge-to-transplant or destination therapy. Patients with paracorporeal pumps (17 patients) or implant times of <30 days (n = 7; 4 died, 3 ongoing, 0 with pocket infections) were excluded. Thirty-five pulsatile left VADs were analyzed, whereas 6 rotary VADs were excluded from analysis. Twenty-three pulsatile VADs were placed in an extraperitoneal pocket (pocket patients) in 18 patients (5 patients had more than one left VAD implant), whereas 12 patients had intraperitoneal (IP patients) VAD placement. RESULTS: Infection with the implanted VAD occurred in 8 of 23 pocket implants (mean support 310 +/- 249 days) and 0 of 12 IP implants (mean support 220 +/- 131 days; pocket vs IP infections: p = 0.003). Staphylococcus and Enterococcus accounted for most of the infections. Four infections in the pocket group recurred during VAD support. Two infected pocket patients had dehiscence of the mid-line wound. CONCLUSIONS: Infections of implanted VADs are associated with serious morbidity (e.g., prolonged antibiotic therapy, recurrent device infections and wound dehiscence). Placing VADs in an intraperitoneal position with a PTFE barrier to protect the viscera may reduce the prevalence of these infections.