ADVERTISEMENT
Quantitation of Bacteria in Clean, Nonhealing, Chronic Wounds
Address correspondence to: George T. Rodeheaver, PhD Plastic Surgery Research University of Virginia Health System Box 801351 Charlottesville, VA 22908-1351 Phone: 434-924-2126 E-mail: grt3s@virginia.edu
Abstract: Quantitative swabs were obtained from 30 clean, chronic wounds on 30 different patients during one visit. The number of organisms and the predominant organism were determined. All samples were processed under both aerobic and anaerobic conditions. Nineteen (63%) of the 30 clean wounds had bacterial levels that were ≥ 105 cfu/cm2. There was no correlation between ≥ 105 cfu/cm2 and delayed wound healing. The most frequently isolated predominant organism was Staphylococcus aureus. In these clean, chronic wounds, an obligate anaerobic organism was identified as predominant or co-predominant in only 2 (6.7%) of 30 wounds. The influence of bacteria on wound healing is complex and controversial. It is accepted that all open, chronic wounds are colonized with bacteria. Most physicians believe that if the wound does not display traditional signs of infection, then the bacteria are not interfering with the healing process. As new information is being presented, many physicians are starting to believe that high levels of bacteria may inhibit healing in the absence of traditional signs of infection.1,2 The level of bacteria that inhibits wound healing but does not display the standard clinical signs of infection has been termed “critical colonization.”3 In this situation, additional criteria are required to diagnose covert infection.4 Gardner et al5 recently assessed the validity of these additional criteria, which include serous exudate, foul odor, discolored or friable granulation tissue, and delayed healing or wound deterioration. They also performed quantitative biopsies and used Robson’s6 definition of greater than 100,000 (> 105) organisms/g of tissue as being infected. Eleven of the 36 (31%) wounds in their study were infected; 91% of those wounds contained necrotic tissue. For these 11 infected wounds not displaying traditional signs of infection, 80% demonstrated delayed healing and friable granulation tissue. Unfortunately, quantitative tissue biopsy procedures are not available to most clinicians. A procedure that is more readily available is the quantitative swab technique.7 Using this technique, Bill et al,8 quantitated the level of bacteria in 38 clean, nonhealing, chronic wounds that showed no classical clinical signs of infection. Tissue biopsies showed that 74% of these nonhealing, clean wounds contained > 105 organisms/g of tissue. The quantitative swab technique detected 79% of these infected wounds. With this quantitative information, the wound care plan was altered to focus on reducing the level of bacteria in these wounds, which resulted in healing. The use of quantitative bacteriology to direct a wound care program is limited by the technical difficulties and expertise required to process the samples. It would be significantly beneficial to wound care providers if a simple diagnostic test were available to document if a clean, nonhealing wound contained > 105 bacteria/g of tissue. In the development of such a diagnostic test, it is necessary to know whether anaerobic organisms play a significant role in the number of bacteria present in clean, nonhealing wounds. The purpose of this pilot study was to quantitate the number of aerobic and obligate anaerobic organisms in a small number of clean, nonhealing wounds.
Materials and Methods
This study was conducted under a protocol approved by the University of Virginia’s Institutional Review Board for Health Sciences Research. The study was a prospective, nonrandomized, observational, clinical trial involving informed patients with chronic wounds. Experimental design. The primary end point was the enumeration of total microbial load and the contribution of obligate anaerobes to the total microbial load. A secondary objective was to identify the predominant organisms colonizing the wound. Thirty patients with chronic wounds who met the inclusion criteria were enrolled in the study. The inclusion criteria were: age 18 years or older; informed and willing to participate; had a chronic wound of at least 1 cm in one dimension, and with a duration of at least 8 weeks; the wound had to be clean and free of any necrotic tissue or slough. At enrollment, the wound was assessed and a quantitative swab was obtained. The swab was processed within 20 minutes for microbial quantitation. The quantitation was performed aerobically and anaerobically. The total number of organisms present in the wound was reported as log10 of the colony forming units (cfu)/cm2. The predominant organism in each wound was identified. Wound assessment. Upon enrollment, the patient’s demographics and wound history were obtained. The wound history included type, location, duration, and current therapy. Then the study wound was assessed for pain, erythema, edema, temperature, exudate, odor, presence and type of granulation tissue, and recent history of wound healing. The wound was photographed, wound dimensions measured, and the wound margins traced on transparent film. Quantitative swab. The technique reported by Bill et al8 was used to obtain a quantitative swab. The wound surface was cleaned with sterile saline and gauze. A sterile alginate swab (Fisherbrand®, Fisher Health Care, Houston, Tex) was rotated within a 1-cm2 area in the center of the wound for 5 seconds and was applied with sufficient pressure to express fluid from the underlying tissue. The swab tip was broken off into a transport tube containing 10-mL of pre-reduced thioglycollate medium (Remel, Lenexa, Kan) and placed on ice for immediate transport to the microbiology lab for processing within 20 minutes. The tube containing the swab tip was vortexed for 1 minute to disperse the bacteria. From this bacterial suspension, 3 serial 1:10 dilutions were prepared in pre-reduced thioglycollate medium. For anaerobic quantitation, 100-µL aliquots from each tube were plated in duplicate on anaerobic blood agar CDC formulation plates (Remel, Lenexa, Kan) and placed in bags that formed an anaerobic environment (GasPak Pouch™ System, Becton-Dickinson, Sparks, Md). For aerobic quantitation, similar plating was conducted in duplicate on both blood agar (TSA with 5% sheep blood, Remel, Lenexa, Kan) and sheep blood agar containing colistin and naladixic acid (Remel, Lenexa, Kan). All plates were incubated at 35 ± 2˚C for 2 days (aerobes) or 5 days (anaerobes) without additional CO2. Following incubation, the dilution plates containing between 20–200 colonies were counted, and the number of cfu/cm2 calculated. A colony representing the primary morphology was isolated for identification. If two morphologies were of similar number they were considered co-primary, and both were isolated and identified. Organisms that grew under anaerobic conditions were quantitated, the primary morphology was isolated and re-streaked on blood agar (both aerobic and anaerobic) and reincubated under aerobic and anaerobic conditions to differentiate facultative organisms from obligate anaerobes. The predominant organism from each wound swab was identified using standard microbiological techniques. When required, the Vitek 2 identification system with both the GP card and the GN card was used (BioMerieux, Durham, NC). Obligate anaerobes were identified by 16S rDNA sequencing using the MicroSeq® Microbial Identification System (Applied Biosystems, Foster City, Calif).
Results
Patient and wound assessment. Of the 30 patients enrolled, 18 (60%) were male and 20 (67%) were Caucasian; mean age was 53 ± 20 years. The mean duration of the wounds was 305 ± 295 days (range, 63–1095 days). Each patient had one or more comorbid 
conditions, such as coronary artery disease (n = 15, 50%), diabetes (n = 14, 47%), paraplegia/quadriplegia (n = 10, 33%), peripheral vascular disease (n = 9, 30%), peripheral neuropathy (n = 8, 27%), or smoking (n = 7, 23%). The most frequently assessed chronic wound was the Stage 4 pressure ulcer (n = 13, 43%), followed by diabetic ulcer (n = 6, 20%), venous ulcer (n = 5, 17%), and various other types (n = 6, 20%). The pressure ulcers were predominantly sacral (n = 16, 54%), followed by ischial (n = 7, 23%), and heel (n = 7, 23%). All patients indicated that they complied with the wound care regimen regarding offloading, use of compression garments, and appropriate dressing changes. A variety of dressings were being used; negative-pressure wound therapy (n = 5, 17%, [V.A.C.®, KCI, San Antonio, Tex]), and wet-to-dry gauze dressings (n = 5, 17%) were used more often than other dressing types. None of the patients were receiving systemic antibiotics. Four patients were being treated with a topical antibiotic gel. Wound infection. None of the 30 wounds showed any classic, clinical signs of wound infection. However, each wound was assessed for subtle signs of critical colonization.5,6 The vast majority of wounds did not display subtle signs of critical colonization other than nonhealing and serous exudate (Table 1). Wound bacterial levels. Nineteen of the 30 (63%) wounds had bacterial levels that were > 105 cfu/cm2 (Table 2). The mean log10
bacterial level of the 30 wounds was 5.38 ± 1.20 cfu/cm2. In this study, there was no apparent correlation between the bacterial level and the change in wound size during the 4 weeks prior to the quantitative swab (Table 3). One of the 6 wounds that became larger over 4 weeks had a wound bacterial count of 102. Of the 3 wounds that became smaller, 1 had a bacterial count of 105. Six of 20 wounds that showed no change in size had bacterial counts < 105. Wound bacteria. Only the bacteria present on the dilution plate that were countable (20–200 cfu) were quantitated. The morphology that was selected as predominant had to be the majority of colonies present. The 
exact percentage of predominance or number of other morphologies was not relevant to this report. It is probable that normal skin flora may have played a significant role since these swabs were taken from open wounds. A slight majority (56.7%) of the wounds contained a single morphology that was predominant. The most frequently isolated predominant morphology was Staphylococcus aureus (40%) followed by Corynebacterium spp. (30%; Table 4). Streptococcus spp. was only predominant in 3 (10%) wounds. In one of these wounds, the predominant streptococci were obligate anaerobes identified as Anaerococcus sp (formerly Peptostreptococcus). This wound also contained facultative streptococci present at a
1 log lower level. In only one other wound was an obligate anaerobic organism detected. In that wound, the anaerobic organism and the facultative organism had a similar morphology and that morphology was the only morphology detected. The obligate anaerobe was identified as Peptoniphilus sp (formerly Peptostreptococcus), and the facultative organism was Corynebacterium sp. Thus, in only 2 of 30 wounds was an obligate anaerobic organism detected as the predominant (n = 1) or co-predominant (n = 1) organism.
Discussion
The chronic wounds enrolled in this study were clean and free of necrotic tissue, slough, and purulent exudate. Despite being clean, 26 (87%) of these wounds had not shown any signs of healing during the previous 4 weeks and 1 wound (3%) had broken down. None of these wounds was considered infected according to the classical signs. Quantitative swabs indicated that 19 (63%) of the wounds contained > 105 cfu/cm2. This could suggest that bacterial levels lower than 105 cfu/cm2 may inhibit wound healing. Six wounds (20%) that were not healing had bacterial levels of 103 (n = 3) or 104 (n = 3) cfu/cm2. Obviously, numerous other factors could account for nonhealing in these wounds. Obligate anaerobic organisms were detected as the predominant (n = 1) or co-predominant (n = 1) organism in 2 of the 30 (6.7%) clean wounds. In their review, Bowler et al9 reported that anaerobes constituted an average of 38% of the total number of microbial isolates in the 9 studies reported. In the majority of these studies, the wounds were infected or contained devitalized tissue; anaerobic organisms would be anticipated to be present in these types of wounds. Gjodsbol et al10 isolated anaerobic organisms from 18 of the 46 (39%) venous leg ulcers they studied, but did not describe the status of the ulcer when the sample was obtained. Sapico et al11 reported that when the wounds were clean, anaerobic organisms were isolated in only 1 of 25 wounds (4%). Gardner et al12 performed quantitative and qualitative assessments of bacteria in 66 chronic wounds and isolated an anaerobic organism from only 1 wound (1.5%), and in this case, the anaerobe was not the predominant organism. These results suggest that anaerobic organisms are rarely the predominant organisms in clean wounds. Bowler et al9 reported in their review that S aureus, P aeruginosa, and β-hemolytic streptococci were the primary causes of delayed healing and infection in both acute and chronic wounds. The frequency of isolating S aureus in chronic wounds can be as high as 94%.10 Sapico et al11 found that once the necrotic debris was removed from chronic wounds, the predominant organisms were S aureus and P aeruginosa. Gardner et al12 isolated S aureus in 34 of the 66 (52%) chronic wounds studied, and found that S aureus was the predominant (n = 10, 29%) organism of the wounds. The results of the present study are in agreement with these reports; S aureus was predominant in 40% of the wounds, with streptococci predominant in 10%, and P aeruginosa predominant in 3.3% of the wounds. The second most predominant organism in this study was Corynebacterium sp, which was predominant in 27% of the wounds. Corynebacterium is a normal resident of the skin and is usually not pathogenic.13 However, Corynebacterium in high numbers will cause skin infections and can be associated with delayed healing.14–17 In the present study, Corynebacterium was the predominant organism in 2 of the 6 wounds that had recently gotten larger just prior to study enrollment.
Conclusion
The results obtained in this pilot study were from 30 wounds. The results did indicate that the presence of obligate anaerobic organisms as the predominant organism in clean wounds is not frequent. The results appear to be in agreement with those of much larger studies, and thus are not controversial.
Acknowledgment
The authors thank the technologists at the University of Virginia Health System Clinical Microbiology Laboratory for their assistance with the identification of the organisms isolated.
From the Department of Plastic Surgery Research, University of Virginia Health System, Charlottesville, Virginia
Disclosure: This study was supported by a grant from 3M Health Care (St. Paul, Minn).
