Skip to main content

The Effect of an Antimicrobial Drain Sponge Dressing on Specific Bacterial Isolates at Tracheostomy Sites

Empirical Studies

The Effect of an Antimicrobial Drain Sponge Dressing on Specific Bacterial Isolates at Tracheostomy Sites

Index: Ostomy Wound Manage. 2005;51(1):60-66.

    Patients with tracheostomies frequently experience nosocomial complications such as bacteremia, sepsis, pneumonia, and multi antibiotic-resistant bacterial infections.1 The resultant increased morbidity may require prolonged treatment and cause extended or multiple hospitalizations.

    Dry gauze dressings, including fenestrated gauze or drainage sponges, are used around tracheostomy sites to protect the tube and absorb small amounts of drainage. However, standard gauze is porous and does not provide a barrier to bacteria. Drainage sponges that contain the broad-spectrum antimicrobial component polyhexamethylene biguanide (PHMB) have been available for a number of years. Because PHMB resists bacterial colonization and penetration, drain sponges impregnated with PHMB function as a bacterial barrier.2

    A controlled, descriptive study was conducted to compare the effect of applying an antimicrobial drain sponge dressing to a non-antimicrobial drain sponge on four bacterial pathogens and the resident normal skin flora around mature tracheostomy sites.

Literature Review

    Wound infection of the tracheostomy site frequently occurs following extended periods of intubation. The results of one study3 demonstrated the presence of polymicrobial aerobic-anaerobic flora wound infections in patients hospitalized for long periods and who required a tracheostomy for 3 months to 2 years. Of the 145 isolates recovered (an average of 5.8 isolates per specimen), the most frequent aerobic isolates identified were alpha-hemolytic Streptococci, Pseudomonas aeruginosa, Enterobacter cloacae, and Staphylococcus aureus. The most common anaerobic bacteria were Peptostreptococcus sp., Bacteroides sp., and Fusobacterium sp. It was concluded that the polymicrobial aerobic-anaerobic flora of tracheostomy wound site infections and the frequent presence of beta lactamase-producing bacteria may have important implications for clinical management.

    In another study1 involving 106 patients, the incidence of positive post-tracheostomy blood cultures was 10.4 %, compared to 6.6% for blood cultures in patients who did not have a tracheostomy. Staphylococcus epidermidis was the most common organism cultured (6.6% of post-tracheostomy cultures, compared to 2.8% for other cultures). Bacteremia was cited as a common complication of percutaneous tracheostomy and was caused by organisms from the patients’ own trachea or skin. In this study, no significant differences were observed between patients who did or did not receive prophylactic antibiotics.

    Curtis et al4 conducted a retrospective review of patients who had undergone coronary artery bypass graft (CABG) to determine if postoperative tracheostomy was associated with a higher incidence of mediastinitis and mortality. In 6,057 patients, the mortality of patients with a tracheostomy was 24.7% compared to 5.2% in patients not requiring the procedure. While some patients who required a tracheostomy may already have been infected and the incidence of other confounding risk factors that created the need for the tracheostomy may have been higher in those who had mediastinitis, it was concluded that patients who require a tracheostomy after CABG have a higher incidence of mediastinitis and a higher mortality rate than patients who do not require tracheostomy.


    This prospective, randomized, controlled, descriptive case series compared a sterile nonwoven drain sponge dressing containing an antimicrobial agent to a control, a sterile nonwoven drain sponge dressing without an antimicrobial component. The purpose of the study was to determine if use of an antimicrobial-impregnated drain sponge would impact the presence of four selected pathogens — methicillin-resistant S. aureus (MRSA), E. cloacae, P. aeruginosa, and S. aureus and resident normal skin flora (alpha-hemolytic Streptococci (AHS) and S. epidermidis) around the stoma site.

    The nonwoven antimicrobial-impregnated drain sponge dressing (Excilon® A.M.D. Drain Sponge, TYCO Healthcare Group LP, Mansfield, Mass.) is a 50/50 rayon-polyester blend that was cleared for marketing by FDA 510(k) in August 2001. The drain sponge is impregnated with the antimicrobial agent PHMB (Cosmocil CQ, Zeneca Biocides, Wilmington, Del.), a hetero-disperse mixture of polymers with antimicrobial activity.5-7 Polyhexamethylene biguanide causes an irreversible loss of essential cellular components as a consequence of cytoplasm membrane damage. Micro-organisms cannot adapt and no known resistance to PHMB has been documented.

    The activity of the PHMB allows the drain sponge to resist bacterial colonization within the dressing, reducing bacterial penetration through the dressing.7 Polyhexamethylene biguanide provides a broad spectrum of activity against a wide range of micro-organisms, fungi, and yeast regardless of the presence of organic matter and is highly active against Gram-negative bacteria.2 Furthermore, PHMB has minimal to no odor, is non-foaming, chemically stable, nonvolatile, and has low mammalian toxicity.2,8

    Selected pathogens and isolates. Selection of the four bacterial pathogens was based on a number of factors. First, previous research identified the isolates most frequently recovered from patients who developed tracheostomy site wound infections.3 Second, many infections are caused by bacteria resistant to commonly used antibiotics. The Centers for Disease Control and Prevention (CDC) has estimated that approximately 13,300 Americans died in 1992 and up to 150,000 died during the past decade of healthcare-associated infections caused by such pathogens.9 Third, antibiotic-resistant bacteria, particularly MRSA, pose significant threats to patients in the post-acute care hospital setting where these organisms are difficult to isolate and treat.10 Finally, the experiences of a post-acute care hospital with a ventilator-dependent unit (the study site) played a role in determining which isolates were chosen; a retrospective review of organisms most frequently identified in patients with tracheostomies and infection confirmed that the four selected pathogens were most commonly identified.

    The presence of the six selected isolates (four pathogenic bacteria and two normal flora) was determined using semi-quantitative wound swab cultures. This noninvasive, reliable technique is advocated because it causes little or no patient discomfort.11-14

    Subject recruitment. After receiving approval from the study site’s Institutional Review Board (IRB), the principal investigator recruited patients who met the study criteria; these patients were interviewed and after agreeing to participate and providing informed consent were enrolled. Participants were randomized on a one-to-one ratio (five patients in each treatment arm) according to a schedule created using the Microsoft® Excel 2000 computer program. Although the number of subjects (n =10) was small, the protocol was not designed to show statistically significant differences but rather to describe the effects of the two sponge dressings on the presence of bacteria.

    Inclusion criteria. Participants were eligible for enrollment if they were 18 years of age or older; able to provide informed consent; had a mature tracheostomy (duration >30 days); had an expected length of stay >5 days; experienced light, moderate, or heavy drainage at the tracheostomy site; required daily dressing changes; had no known sensitivity to PHMB; and were not receiving systemic or local antibiotic therapy. Each participant provided written informed consent that included a description of the study methodology, possible risks, potential benefits, alternative therapy, associated costs, compensation for participation or injury, legal rights, and the right to study withdrawal. Subjects’ identities were kept confidential by using an assigned number for tracking study results — no names were printed on data collection instruments or study forms and researchers kept the list of subject names in a locked file.

    Procedure. Each participant received five daily consecutive study-related dressing applications. Cultures were taken at baseline (screening, enrollment, and initial application of the dressing) and at Days 1, 2, 3, and 4 (dressing change and site assessment). At baseline and each daily study-related dressing application, researchers photographed the tracheostomy and surrounding skin. In addition, the investigator documented clinical assessment parameters including pain at the stoma site, condition of the surrounding skin, color and odor of the drainage, and any other pertinent clinical observations.

    The principal investigator performed all study-related dressing applications. The dressing change procedure followed institutional policy which includes cleansing of the peristomal area with a solution of half normal saline/half hydrogen peroxide followed by rinsing with normal saline rinse and inner cannula cleansing if necessary. After cleansing, the semi-quantitative swab cultures were collected using aseptic technique. A rayon-tipped swab applicator was rotated over the peristomal area in a zigzag pattern and the tip of the swab was inserted into a sterile round bottom polypropylene tube containing a non-nutritive, highly conductive transport medium (Starswab II Bacteriology Culture Collection and Transport System, Medegen Medical Products, Gallaway, Tenn.). Following culture collection, the investigator applied either the PHMB antimicrobial drain sponge dressing or the control dressing to the tracheostomy stoma using clean technique.

    All culture samples were immediately sent to a CLIA-certified laboratory for susceptibility testing and analysis. This process was identical for culturing both the targeted pathogenic organisms and the normal skin flora.
Wound culture swabs were inoculated on standard media and streaked in four quadrants. Results ranged from “no growth” to “1 to 4+ growth” of the micro-organism, depending on the number of streaked quadrants that supported bacterial growth. The same technique was used to identify the pathogens and normal skin flora. The investigator recorded results by subject and group until reports on all cultures were completed.

    Throughout the duration of the study, the presence of copious drainage at the stoma site dictated the frequency of dressing changes — eg, daily or as needed for loss of dressing integrity or excessive drainage. All caregivers were trained by the researchers at the first visit by demonstration/return demonstration to perform these additional dressing changes in compliance with the randomization scheme but all study-related dressing changes were performed by the investigator.

    To determine the dressing’s effect on targeted isolates, researchers tracked the number of days each selected isolate was present per study group. The investigator used a two-by-two table, a standard epidemiology reporting method for comparing disease frequencies, to cross-tabulate the presence of a selected isolate.


    All 10 patients (three men and two women in the treatment and four men and one woman in the control group,) completed the study. Ages ranged from 39 to 83 years; treatment group mean age was 58 years, control group mean age was 71.2 years. The duration of time a tracheostomy was present varied from 6 weeks to 36 months; treatment group mean duration was 14 months, control group mean duration was 7.5 months. Study participant’s dependency on the ventilator across both groups was as follows: dependent at all times (n = 3); dependent at night only (n = 4); intermittent use (n = 1); not required (n = 2).

    During the course of the 25-day study, 50 cultures were collected and analyzed. At baseline (on enrollment), three of the 10 subjects tested positive for MRSA. Of these, two were assigned to the control group and one to the antimicrobial drain sponge dressing group. Results show that each of the four pathogens selected for study was present for a greater number of days in the control group than in the antimicrobial drain sponge group (see Table 1). The PHMB dressed sites had an absence of growth of the four selected pathogens for 11 days whereas these pathogens were absent for 6 days in the control-dressed sites.

    Clinical findings. In the control group, the total positive MRSA count (n = 4) included two persons who tested positive at baseline and two subjects who tested positive following enrollment. At least one of these 4 patients tested positive for MRSA every day a culture was taken. One patient randomized to the antimicrobial drain sponge dressing group cultured positive for MRSA at baseline. At the final assessment, this participant had no bacterial growth. No other subjects in the PHMB group tested positive for MRSA during the study. In total, MRSA was recovered for only 3 days in the antimicrobial drain sponge dressing group, versus a total of 11 days from the control.

    Pseudomonas aeruginosa was present in the control group for 10 days versus 3 days in the antimicrobial drain sponge group and the other selected pathogens (E. cloacae and S. aureus) were present for a greater number of days in the control than in the antimicrobial drain sponge dressing group (see Table 1).
Normal skin flora isolates were present in both groups but were present fewer days in the control than in the antimicrobial drain sponge dressing group. Specifically, alpha-hemolytic Streptococci and S. epidermidis were present for 11 days in the antimicrobial drain sponge group versus 2 days in the control.


    This prospective, randomized, controlled, descriptive clinical case series was conducted to evaluate whether the application of an antimicrobial drain sponge dressing would effect the presence of four targeted pathogens prevalent in the tracheostomy patient population. In addition, the effect of the dressings on resident normal skin flora was assessed by culturing for two micro-organisms.

    Semi-quantitative wound swabs were used to recover and identify those isolates present marginal to the subjects’ tracheostomy stomas. However, semi-quantitative microbiology alone cannot predict the risk of sepsis.15,16 Quantifying wound organisms does not address the question of bacteremia nor take into account the cause and extent of the wound in relationship to the patient and underlying comorbidities. Additional variables, including the presence and/or stage of infection or the lingering presence of secretions at the tracheostomy site, must be considered when interpreting culture results.

    The PHMB-impregnated antimicrobial drain sponge dressing has not been found to inhibit the growth of normal flora critical to controlling bacterial balance and maintaining the skin’s barrier function.17 Tempering the presence of pathogens may support the growth of resident normal skin flora. Conversely, having a preponderance of pathogens competing for a nutrient base may limit the capacity of normal skin flora to proliferate and flourish. It is possible that as the antimicrobial drain sponge dressing reduced the pathogenic bioburden, it simultaneously created an environment in which normal skin flora were allowed to thrive.

    This disparity also could be attributed to microbial competitive inhibition, a phenomenon that can occur when the presence of one organism limits the growth of another due to competition for the shared nutrient base. This may help explain the observed absence of slow growing Alpha-hemolytic Streptococci in the control group, suppressed by the presence of MRSA and P. aeruginosa.

    Although a greater number of control than study group patients cultured positive for MRSA, the effects of this result could not be ascertained. The small sample size and the short duration of tracking outcomes limited data analysis and interpretation. However, the observed trends are noteworthy and suggest that a larger subject population study is warranted to determine the impact of a PHMB antimicrobial drain sponge dressing on pathogenic organisms and tracheostomy infection rates. Ideally, the study should include a large sample size and the protocol should be modified to include acute care subjects to determine if the delay in colonization can occur in a new tracheostomy. Future study designs also might benefit from a retrospective data collection effort that compares infection rates and clinical outcomes of subjects receiving the antimicrobial impregnated drain sponge versus those receiving the control.


    One of the multiple factors associated with the increased risk of acquiring antibiotic-resistant bacterial infection is the presence of any indwelling, invasive devices, such as a tracheostomy cannula.18 Polymicrobial aerobic-anaerobic flora are associated with tracheostomy wound infections at stomal sites. Methicillin-resistant S. aureus is not an organism that occurs naturally and is usually facility-acquired. Pseudomonas aeruginosa is neither a component of normal skin flora nor a species found routinely in healthy humans. The presence of either of these pathogens in any healthcare facility has potential serious consequences.
The results of this clinical case series suggest that the PHMB-impregnated antimicrobial drain sponge has an effect on peristomal pathogens and normal skin flora and could be an important adjunct in the control of antibiotic-resistant and other bacterial infections frequently found in patients with tracheostomies. Results also suggest that using an antimicrobial drain sponge dressing may help control MRSA and P. aeruginosa in an institutional setting. A simple substitution from a regular drainage sponge to one impregnated with PHMB fits into existing clinical protocols and may improve infection control outcomes. Additional research is warranted.


    The authors wish to thank Jack Runner, Vice President, North Coast Clinical Laboratory Sandusky, Ohio, for his valuable contribution towards this study.