← Back to July 2013

The Trouble with Driveline Infections

Christine E. Koval, MD (a)
Lucy Thuita, MS (b)
Nader Moazami, MD (c)
Maria Mountis, DO (d)
Eugene Blackstone, MD, PhD (b,c)

Cleveland Clinic Foundation:
(a) Department of Infectious Disease
(b) Department of Quantitative Health Sciences
(c) Department of Thoracic and Cardiovascular Surgery
(d) Department of Cardiovascular Medicine

Driveline infections frequently complicate ventricular assist device (VAD) support (1,2,3). Most driveline infections (DLIs) occur at the exit site but can involve tissue anywhere along the driveline tract to the pump pocket. Consensus definitions have been published to help clinicians and researchers describe DLI and the depth of tissue involved (4).

While much is known about driveline infections, little has been published on their "natural" history. We sought to describe DLIs over time in our patients with continuous flow (Heart Mate II) VADs including the onset, risk factors, organisms involved, association with invasive infections and outcomes.

We retrospectively evaluated 181 patients with first time Heart Mate II VADs implanted at The Cleveland Clinic between July 2004 and July 2011 with follow up through December 2011. Forty-four of 181 (24%) developed driveline infections over 309 person-years. Median time from VAD implantation to DLI was 231 days (range 22-862). Cumulative DLIs increased steadily over time on VAD support, as is well recognized from previous reports. However, we identified a spike in hazard for DLI from 2.2% to 10% occurring at 6-9 months following VAD implantation.

We found no risk factors clearly associated with driveline infection and thus could not identify a cause for the incident hazard to temporarily increase 6-9 months after VAD implantation. Trauma to the driveline has previously been associated with DLI, thought due to loss of tissue in-growth at the exit site. There may be behavioral or structural effects that could lead to traumatic injury peaking between 6-9 months after implantation. Our data on trauma was not prospectively collected and could not statistically be associated with DLI, although 20/44 (45%) retrospectively reported driveline trauma that preceded their infection. More prospective methods to collect this data are required.

Pseudomonas aeruginosa was the most frequent infecting organism, causing 28% of DLIs. However, both Staphylococcus aureus and coagulase negative Staphylococcus species also contributed widely, causing 19% and 13% of DLI, respectively. Of particular interest was that Pseudomonas aeruginosa became even more prevalent over time in patients with existing DLI, often replacing earlier infecting organisms as the predominant pathogen. These "superinfections" occurred in 9/44 (23%) of those with DLI at a median of 256 days (range 44-444 days) after the initial DLI. That water-borne biofilm producers, predominantly Pseudomonas and Proteus species, became more prevalent over time suggests a role for showering in driveline infection evolution.

Additionally, it was observed that Pseudomonas aeruginosa contributed to the evolution from superficial driveline to deep driveline infection. Of 14 deep driveline infections, 9 began as superficial driveline infection and progressed over a median 173 days (range 61-362 days). Eight of 9 (89%) of these were due to Pseudomonas aeruginosa (3/8 that were initially identified as Staphylococcal infection).

Surgical management for deep driveline infection occurred in 12/14 patients. All four that were transplanted survived to follow up. However, only 2/8 with omental flaps, incision/drainage, VAD exchange or VAD explant survived. While we found that any driveline infection was associated with an incremental risk for death on VAD (HR 2.15, 95% CI 1.18-3.97, p=0.01), our data highlights that even patients with prolonged and deep driveline infection could be successfully transplanted. Overall, 13/44 (30%) with DLI underwent heart transplantation and only 1 died.

Twenty of 44 (30%) with DLI died over a median of 322 days (range 17-858), indicating the prolonged duration of infection management. Only 10/44 (22%) ultimately died from sequelae attributable to their DLI (sepsis, mycotic aneurysms). Most 29/44 (66%) required admission at least once for DLI, and 18/44 (41%) had at least one episode of bacteremia due to the DLI organism.

All were managed with prolonged durations of either oral or intravenous antibiotics or both (median 171 days). Despite this, only 3/44 (6.8%) developed an antibiotic related complication (1 with Clostridium difficile associated diarrhea, 1 with catheter associated blood stream infection, and 1 with oral candidiasis). Multidrug resistance did, however, develop in 7/44 (16%), all of which were gram-negative pathogens, so ongoing antibiotic pressure is not without its potential downside.

Our data highlight the ongoing importance of driveline infections in patients with continuous flow VADs. Most begin at the driveline exit site and evolve over time. While these infections can often be effectively managed for prolonged periods, they are associated with reduced survival on VAD support. We advocate ongoing efforts to mitigate risk for onset and evolution of driveline infections with targeted investigations of driveline care and maintenance.

Disclosure statements:
Drs. Koval, Mountis, Blackstone and Lucy Thuita have no conflicts of interest to report.
Dr. Moazami receives consulting fees from Terumo and Thoratec, Inc.


  1. Schaffer JM, Allen JG, Weiss ES, Arnaoutakis GJ, Patel ND, Russell SD, et al. Infectious complications after pulsatile-flow and continuous-flow left ventricular assist device implantation. J Heart Lung Transplant 2011; 30:164-174.
  2. Sharma V, Deo SV, Stulak, JM, Durham LA, Daley RC, Park SJ, et al. Driveline infections in left ventricular assist devices: Implications for destination therapy. Annals Thoracic Surgery 2012; 94(5): 1381-1386.
  3. Gordon, RJ, Weinberg AD, Pagani FD, Slaughter MS, Pappas PS, Naka Y, et al. Prospective multicenter trial of ventricular assist device infections. Circulation 2013; 127:691-702.
  4. Hannan MM, Husain S, Mattner F, Danziger-Isakov L, Drew RJ, Corey GR, et al. Working formulation for the standardization of definitions of infections in patients using ventricular assist devices. J Heart Lung Transplant 2011; 30: 361-374.

Share via:

links image    links image    links image