Assessment of Wound Bioburden Development in a Rat Acute Wound Model: Quantitative Swab Versus Tissue Biopsy

Disclosure: Research funded by East Carolina University Research/Creative Activity Grant #99-53 Introduction One of the greatest controversies in wound management is the debate over the usefulness of swab cultures for predicting the presence of wound infection. The swab culture method has received criticism because it is thought to only estimate wound surface microbial numbers and not intra-tissue numbers. Because of these limitations, the Centers for Disease Control (CDC) recommends the use of either wound fluid aspiration or tissue biopsy to determine the levels of bacteria in the wound bed.[1] This recommendation is also supported by the Treatment of Pressure Ulcers Guideline Panel 1994 publication.[2] In recent years, studies have shown that quantitative swab cultures have high sensitivity (ranging from approximately 87–100%) and good specificity (85–94%) and accuracy (approximately 90–99%) (with the exception of pressure ulcers) of diagnosing wound infection when compared with either aspirated or tissue biopsy specimens.[3–9] These studies included investigations of both acute and chronic types of wounds; however, much of the research focused on chronic wounds. Based on these studies, the evidence supports the fact that a quantitative swab method of collection and culturing is an acceptable method for diagnosing wound infections in situations where aspiration and/or biopsy are not feasible or desired. This is particularly significant because a broader group of healthcare providers can utilize swabs to collect specimens and swab methods are less invasive and less painful to most patients. A recent survey indicates that approximately 54 percent of wound care specialists in the United States (US) collect only swab cultures, while approximately 42 percent collect both swab and biopsy specimens.[10] Research studies should focus now on the best type of swab/transport systems to be used, on the actual method of processing in the laboratory, and most importantly, on investigating the best method of swab collection and standardization of practices. The wound culturing survey of US wound care specialists indicates that there is great variation in the exact methods used for collection (whether clinicians cleanse or do not cleanse wounds prior to collection, the actual site within the wound that is swabbed, whether a premoistened swab or dry swab is used, the exact swab technique that is used, types of tests ordered, etc.).[10] These findings have been confirmed by two surveys conducted in the United Kingdom on select groups of wound care clinicians.[11,12] Currently, there is no single accepted method for the performance of swab cultures. Some methods of swab collection include the 10-point diagonal and the one-point rotation methods as well as sampling a 1cm2 area of the wound bed.[13–15] To address the continuing controversy of the best culturing methods, we compared the one-point quantitative swab with a tissue culturing method in a rat acute wound model. We selected the one-point method in this model because of the relatively small size of wounds that can be induced in rats and to better avoid contamination with peri-wound flora. This method also allows tissue sampling from the same site that was swabbed. Samples were collected by each method to determine types and numbers of bacteria present in the center of the wound bed at three different times: baseline, two days post-wounding, and four or 12 days post-wounding. Methods Animals. Animals were maintained according to the guidelines of the Animal Care and Use in Research and Instruction of East Carolina University and the National Institutes of Health Guide for the Care and Use of Animals. Fourteen adult, female, Sprague-Dawley rats (250g) were allowed to acclimate in the animal housing facility of East Carolina University for one week prior to the experiment. Animals were housed singly in a controlled lighting (light cycle: 12 hours on, 12 hours off) and temperature (22 degrees C) facility. Animals were maintained on a normal diet and fed ad libitum. On the day of surgery, the rats were randomly assigned to either the four-day (n=6–7) or 12-day (n=7) survival group. Anesthesia. The baseline tissue samples were collected after the animals were anesthetized using ketamine/xylazine (ketamine 100mg/mL, xylazine 10mg/mL with a dosage of 0.1cc/100g of body weight) given by intramuscular injection. The Day 2 post-surgery samples were collected by bell jar induction using five-percent sofluorane with a controlled ventilation system. Samples were collected on Day 4 or Day 12 after euthanasia. Surgery. Prior to surgery, the upper cervical and thoracic region of the dorsal surface of each rat was shaved. The surgical area was cleaned with betadine scrub followed by a normal saline wash. Using aseptic technique, full-thickness acute surgical wounds were induced in the mid-scapular region of each rat using an 8mm biopsy punch (Miltex Instrument Company, Inc., Bethpage, New York). Subsequently, sterile forceps were used to remove the full-thickness skin flap. Sterile gauze was then utilized to cleanse the wounds of any coagulated blood and to control bleeding. Swab cultures. The Culturette II Collection and Transport aerobic swab system (Becton Dickinson & Company, Sparks, Maryland) was utilized for collection and transport of swab cultures. Manual pressure was applied to the wound with enough force to produce a small amount of exudate. The swab cultures were taken from the center of each surgical wound by rotating the swab three times clockwise. A brief study was conducted to determine the volume of fluid (saline) that could be collected with Culturette II™ swab system. Ten swabs, one each from a paired set, from a single lot number were tested. Measurements were made to the nearest 1mL. Tissue biopsy cultures. Tissue samples were taken from the center of the surgical wound using sterile surgical scissors. All tissue samples were placed in preweighed sterile microcentrifuge tubes containing 1mL sterile saline. Tissue samples were stored on ice prior to transport to lab for weighing. After weighing the tissue samples, each sample was prepared for quantitative microbiological analysis (colony counting). Moist wound dressings. After baseline swab and tissue samples were taken from the surgical wound of each rat, the wounds were dressed with 0.1cc Saf-Gel hydrogel (ConvaTec, Princeton, New Jersey) using a sterile 1cc tuberculin syringe. The wounds were subsequently covered with sterile gauze to prevent drying of the wound bed. The bandages were kept in place with coflex bandaging and surgical tape. The hydrogel and gauze dressings were changed daily for either 4 or 12 days prior to euthanasia. After mechanically debriding the wounds with forceps, they were cleansed with sterile saline-soaked gauze. The hydrogel and gauze bandages were then reapplied. Colony counts. Tissue samples were placed into separate sterile disposable tissue grinders (SAGE Products, Inc., Crystal Lake, Illinois) and homogenized into a fine suspension according to manufacturers directions. Each tissue suspension was then serially diluted in sterile saline and 0.1mL aliquots plated to trypticase soy agar with five-percent sheep blood (SBA) plates (Smith River Biologicals, Ferrum, Virginia). Swab samples were placed into a tube containing 1mL sterile saline, vortexed thoroughly, serially diluted, and plated in the same manner as the tissue samples. The SBA plates were incubated overnight at 35?C. Colony counts were performed, and the number per swab or per gram of tissue was calculated. Results The average colony counts obtained by the one-point swab culturing technique on Day 4 was 2.2x105, whereas the average tissue count was 3.7x107/g for biopsy samples collected (Table 1). By Day 12, samples collected by swabbing contained an average of 4.7 x105 colonies, and biopsied tissues contained 5.7x107 colonies/g (Table 1). Statistical analysis (Table 1) indicated that the two methods of collection detected similar levels of colonization on Day 4 (t=1.27/critical t=1.86) but that significantly different levels of colonization occurred by Day 12 (t=1.97/critical t=1.75). Comparison of the colony counts obtained from specimens collected after two and four or two and 12 days post-wounding allowed an examination of the levels or trend of colonization on the wound surface (swab culturing) versus the wound bed (biopsy tissue). The average colony counts obtained by swab culturing greatly increased from Day 2 to Day 4 (6.7x102 to 2.2x105) and to Day 12 (2.8x103 to 4.7x105) indicating a longer time (Day 4) to stabilization of numbers of microorganisms on the surface (Table 2). The average colony counts obtained from tissue biopsy samples did not significantly increase from Day 2 to Day 4 or to Day 12. The average tissue counts for Day 2 to Day 4 were 3.4x107 to 3.7x107, and the tissue counts for Day 2 to Day 12 were 3.7x108 to a slight drop of 5.7x107 (Table 2), indicating earlier (Day 2) stabilization in the wound bed. Even though there was stabilization of actual bacterial numbers in the wound, the quantitative one-point swab method underestimated the number of bacteria in the wound bed by a two-log difference on Day 4 and Day 12. The final colony counts obtained by the one-point swab culturing technique and by tissue biopsy sampling were reproducible. The average colony counts obtained by culturing in four-day old wounds was 2.2x105 and was 4.7x105 (Table 1) in 12-day old wounds with a percentage similarity of microbial numbers of 46.8 percent (Table 3). The average colony counts obtained by tissue biopsy was 3.7x107/g in four-day old wounds and 5.7x107/g (Table 1) in 12-day old wounds, which yields a percentage similarity of 64.9 percent in bacterial numbers (Table 3). Discussion The use of swab culture techniques for detecting wound infections remains controversial. Both the CDC and the guidelines provided by the Treatment of Pressure Ulcers Guideline Panel in 1994 do not recommend swab culturing for use in establishing a diagnosis of chronic wound infection without distinguishing between the potential merits of quantitative or semiquantitative methods.[1,2] The CDC actually rcommends the use of either tissue biopsy or fluid aspiration to determine bacterial content in wounds. However, as noted in the introduction and reviewed by Stotts,[3] swab culturing techniques do compare favorably with aspirated and biopsy-collected wound specimens for diagnosing wound infection. The data obtained in this study in a rat acute surgical wound model indicates that the one-point swab culture yielded colony counts lower (105) than the tissue biopsy samples (107) in this same model (Table 1). This finding of underestimation by quantitative swab culturing when compared with biopsy is consistent with results of several studies.[16–20] Half of these studies obtained a one-log underestimation by swab and half demonstrated a two-log underestimation, including the current study (Table 4). The current study was performed in a rat wound model, and at least a part of the explanation of the two-log difference obtained could be related to the small size of the wound area sampled. The one-point rotation swab collection technique was performed by placing the swab in the center of wound and rotating three complete turns in one area in order to reduce the likelihood of contamination with periwound flora. This procedure likely limited the amount of sample collected (particularly the number of tissue cells, if not the volume of tissue fluid) and potentially the number of bacteria isolated. Biopsies were then collected immediately after swabbing, and tissue sample was collected directly underneath the swabbed area. Additionally, the biopsied tissues were not dipped in alcohol and flamed before homogenization because of the small size of the tissues collected (average weight 0.0211g for Day 4 and 0.0132g for Day 12) (Table 5). Some of the factors that could have contributed to the differences found in the studies cited (Table 4) include type of wound cultured, actual tissue sampled (eschar, non-necrotic wound bed, etc.), experimental model (excisional versus incisional; human versus animal), and tissue preparation (alcohol flamed or not). Additionally, this discrepancy in numbers could be related to the volume of wound fluid that was collected by the swab system utilized in each study. A brief testing of the Culturette II swab system used in this study indicates that approximately 175µl volume of sample was absorbed per swab (Table 6). No actual volume measures were given in the other studies. The results of the studies cited (Table 4) are not congruent with the findings of Levine and colleagues,[14] which found that quantitative swab cultures overestimate biopsy cultures by one log in burn patients. These consistent findings of underestimation of bacterial numbers by quantitative swab culturing were obtained using differing quantitative swab culturing techniques, including the one-point rotation method, the Brentano wet culturing technique, the 1cm2 area for five seconds, and a modified swab technique using an eight-layer gauze (Table 4).[13,14,17,21] Some of the inconsistency with Levine’s results can be attributable to differences in the collection techniques including differing volumes of wound fluid collected, exposure time of the collection device to the wound bed, and composition of swab material. Levine,[14] Bill, et al.,[18] and Steer, et al.,[19] utilized the 1cm2 collection area. However, Levine, et al., used a cotton-tipped swab (overestimated bacteria in tissue by one log), while Bill, et al., (underestimated bacteria in tissue in 21% of specimens) and Steer, et al., (underestimated bacteria in tissue by two logs) both utilized alginate swabs. The underestimation of bacterial numbers found in this current study was obtained utilizing rayon-tipped swabs. It is thought that acute and chronic wounds differ in the environment in which microbial flora exist. However, quantitative bacteriologic findings from both acute and chronic types of wounds (in animals and humans) indicate that a similar relationship exists between surface microbial numbers and intra-tissue bacterial numbers. As discussed above, multiple studies in both acute and chronic wounds demonstrate an underestimation of bacterial numbers by surface culturing. This is contrary to currently accepted teachings established by Levine’s studies in 1976 and needs to be reevaluated. Utilizing an acute wound model, we were able to evaluate the bacterial counts in newly induced wounds versus wounds of longer duration. Statistical analysis (t-test) was performed to determine if the two collection techniques used found significant differences in bacterial levels on Days 4 and 12. The t-test results showed that the two methods detected similar levels of colonization on Day 4 but were significantly different by Day 12 (Table 1). These results suggest that swab and biopsy sampling may be more closely correlated in early stages of colonization. The possibility exists that the two regions (surface vs. wound bed) of the wound were colonizing at two different growth rates. The data obtained in this study indicates that colonization within the wound bed, as measured by tissue biopsy, becomes stable by Day 2. Levels of colonization in animals that were wounded and maintained for four days were 104 by Day 2 and remained at the same logarithmic power on Day 4 (Table 2). The level of colonization in animals wounded and maintained for 12 days was 108 on Day 2 and was 107 by Day 12 (Table 2). Although this slight drop was noted, it appears that colonization was very close to stabilized within the tissue by Day 2 after wounding. Similarly, a previous study performed by Price, et al.,[22] demonstrated that bacterial levels were stable Day 1 to Day 3, and this current study demonstrates stability out to Day 12 in an acute rat model. Wound surface colonization levels determined by the one-point swab collection technique reveal a different trend in stabilization of microbial numbers. In animals four days post-wound induction, the levels of colonization were 102 from swab samples at Day 2, which increased to 105 by Day 4 (Table 2). A similar increase was detected in animals 12 days post-wounding with colonization levels rising from 103 on Day 2 to 105 by Day 12 (Table 2). Based on these results, the levels of colonization within the tissue bed stabilize earlier than on the wound surface, which seems reasonable due to the exposure of the surface to a variable environment. Our data indicates that in a rat wound model, stabilization of microbial flora (both in number and types of bacteria detected) occurs as early as Day 2 in the wound bed and by Day 4 on the wound surface. The types of bacteria isolated (Coagulase-negative Staphylococcus species and Corynebacteria species) remained constant throughout the study regardless of the method of culture. Because of this early stabilization, the rat acute wound model would allow more rapid testing of various wound treatment modalities. However, further studies need to be conducted to determine if similar patterns are present in animals in a chronic disease state. The results in this study lead to the question of whether the two techniques are actually different or if their difference is primarily due to the fact that they are sampling colonization levels of two very different environments within the same wound. Because of the difference in the inherent amount of sample that can be collected using a swab (composed primarily of tissue fluid with a lower number of tissue cells) versus the amount obtained by biopsy (composed of a large number of tissue cells and some tissue fluid), there will always be differences in the actual numbers of microorganisms recovered by each method. The major question is whether the wound surface environment has predictive value for the underlying tissue’s risk for infection. Therefore, given that our data support stability of colonization by Day 4 on the wound surface and within the wound bed, the question becomes: Is the sampling by the two culturing techniques once stabilization of microorganisms has occurred reproducible (i.e., Once there are stabilized numbers of bacteria in the wound, does the swab collection and biopsy collection methods yield reproducible numbers thereafter when specimens are collected)? This was addressed by examining the logarithmic values for swab and tissue biopsy on Days 4 and 12 (after colonization is stabilized in both wound environments). Both methods appear to be stable and repeatable in their level of detection, with the swab technique detecting 105 on Days 4 and 12 and biopsy detecting 107 on Days 4 and 12 (Tables 1 and 3). Logarithmically, both techniques appear to be very stable measures of colonization over time. However, each technique’s results varied upon examination of actual numbers rather than logarithmic values when comparing Day 4 to Day 12. To further test the stability of these measures, we determined the percent similarity between Days 4 and 12 swab values as well as Days 4 and 12 biopsy values. The swab technique was approximately 47-percent stable, while the biopsy was approximately 65-percent stable resulting in an 18-percent difference between the two methods (Table 3). Considering that the majority of laboratory reports, which categorize colonization as few, moderate, or many, or on other interval scales (1+, 2+, etc.), it is not clear whether 18 percent is significant enough to influence the level at which colonization is reported. Even in laboratories that perform quantitative microbiological assays, there will be inherent errors in the dilution procedure (mixing, pipetting, etc.), so actual numbers of bacteria will differ to some degree. The trend in current research defines infection as a combination of host resistance, microorganism virulence, the type of wound, and the amount or dose of microorganisms present in the wound. If the one-point rotational swab technique or other quantitative methods of specimen collection provide stable detection, both in organism type and in relative numbers of microorganisms, and are capable of predicting microbial numbers within the wound bed, the inability to report precise numbers of microorganisms using the swab technique needs to be weighed against the trauma induced to the wound bed when performing a tissue biopsy. Before the controversy over the appropriateness of swab cultures can be fully resolved, the disagreement over the definition of wound infection must be settled. The appropriateness of one threshold value for determining wound infection or for evoking clinical concern over the potential for impaired wound healing must be questioned.[23] The usefulness of a single threshold value derived from studies on a limited number of microorganisms appears to be at odds with a common definition of wound infection. This definition states that infection is present when the host responds to actively replicating microorganisms. However, this definition is not adequate in all situations (i.e., in immune dysfunctional patients). The importance of considering microorganism virulence when determining whether a tissue is infected has been known for some time. Additionally, the presence of antibiotic resistance in a microorganism and the degree of resistance should also be an important factor in determining the acceptable threshold value for clinical infection. Therefore, the determination of the type or types of microorganisms present and antimicrobial susceptibility pattern may be more important than numbers. Actual numbers of microbes present is not the primary consideration in diagnosis of true wound infections. Opinion leaders in the treatment of burn wounds generally agree that quantitative cultures of biopsy specimens aid in identifying predominant microorganisms present, which is helpful in determining appropriate intervention strategies but do not in themselves adequately diagnose invasive infection.[24,25] Evolving thought suggests that changes in the wound appearance, systemic signs, and particularly histologic examination, should all be taken together to assess invasive infection. Given the limited data on determining a threshold value for clinical infection in a wound for various species or strains of bacteria and the issue of evolving antibiotic resistance, the argument can be made that no one threshold value is acceptable for defining clinical infection of a wound. As such, the argument over the utility of the swab culture becomes less critical especially if the definition of wound infection is modified to reflect a range change (in numbers and types of microorganisms) rather than a threshold value. The issue today is not whether quantitative swab cultures are effective in diagnosing wound infection (i.e., provide relative numbers and likely pathogen identification), but whether semi-quantitative methods are effective in diagnosing wound infection, because these quantitation procedures are more commonly used in routine microbiology laboratories. Several authors have addressed this question and have found that semiquantitative swab culturing methods show good sensitivity, specificity, and have acceptable predictive value for accessing infection.[4,5,9,26] While tissue biopsy cultures remain the most preferred method for diagnosing wound infection, mounting evidence indicates that swab cultures do provide a useful alternative that is accessible to a greater number of wound management clinicians. An inherent difference in the sample collected (primarily tissue fluid versus tissue cells) using swab and biopsy prodecudres will always exist, and multiple variables can contribute to this difference as discussed previously. The exact numbers obtained with each approach will vary even within a single study and certainly will vary when comparing multiple studies. However, averaging the data obtained still can allow for the assessment of trends in colonization and/or infection. In a true wound situation, particularly chronic wounds, many variables (underlying diseases, immune system competence, and available healthcare) relate to an individual patient’s ability to contend with increased bioburden levels. This supports the need for individual practitioners and/or group practices to standardize the collection techniques used (both swabbing and biopsy), so that comparisons can be made more easily from their own patients’ samples (consecutive samples from the same patient or from many different patients). The major issue then becomes whether the method being used allows for the isolation and identification of any potential pathogens within wound, regardless of exact numbers. Conclusion Quantitative swab culturing with rayon tipped-swabs detects similar types of microorganisms in acute surgical wound beds as compared to biopsy cultures. The one-point quantitative swab culture method utilized in this study underestimated bacterial numbers by a factor of two logs. However, the difference based on actual numbers of bacteria (18%) detected by the two methods is unlikely to influence reports provided by the majority of clinical laboratories using a semiquantitative processing technique and interval scale reporting system. The findings in this study are similar to other studies that have assessed both methods of culturing in acute and chronic wounds.