Skip to main content

Effect of Ultraviolet Light C on Bacterial Colonization in Chronic Wounds

Empirical Studies

Effect of Ultraviolet Light C on Bacterial Colonization in Chronic Wounds

Index: Ostomy Wound Manage. 2005;51(10):32-45.

    Methicillin-resistant Staphylococcus aureus (MRSA) is a strain of antibiotic-resistant bacteria commonly found in wounds. The prevalence of MRSA increased from less than 3% to rates as high as 40% between the early 1980s and 1990s in many hospitals in the US and Europe.1,2

In Canada, hospitals and long-term care facilities have recorded cases of MRSA,3,4 and community-acquired cases also have surfaced.5 Methicillin-resistant S. aureus is spread primarily by direct or indirect person-to-person contact. In hospitals, hand, environment, and equipment contamination are the acknowledged vehicles for the spread of MRSA from one individual to another.6

    The development of antibiotic-resistant bacteria in chronic wounds has resulted in the search for new antimicrobial therapies.7 Antibiotic resistance is most often recognized in vulnerable groups such as the very young or old8; the consequences of antibiotic resistance may include increased length of hospitalization, costs of diagnostic testing, and risk of morbidity and mortality to the individual.9,10

    Ultraviolet light (UVL) occupies the electromagnetic spectrum between X-rays and visible light. The wavelengths of UVL are divided into three bands: UVA (320 nm to 400 nm), UVB (290 nm to 320 nm), and UVC (200 nm to 290 nm). Previous studies have suggested that UVL, a combination of UVA, UVB, and UVC wavelengths, triggers cellular actions and physiological effects required for treating chronic wounds by stimulating cell proliferation,11 increasing epidermal thickening,12 enhancing blood flow in the cutaneous capillaries,13 facilitating wound debridement,14 and killing bacteria.15,16 Ultraviolet light A and B, the longer and mid-range wavelengths, respectively, have minimal bactericidal properties. They are used primarily to treat dermatological conditions such as psoriasis and dermatitis.17 Ultraviolet light C is used primarily in wound healing.

    Relatively few studies have examined the efficacy of UVL treatment in chronic wounds. The results of a randomized controlled study by Wills et al18 suggested that subjects with superficial pressure ulcers treated with a combination of UVA, B, and C healed faster than control subjects receiving only standardized wound care. Crous and Malherbe19 compared the effects of UVL and laser irradiation on the healing of chronic ulcers and observed that the effectiveness of both modalities in the treatment of chronic ulcers was clinically significant. Nussbaum et al14 reported that a combined therapy of ultrasound and UVC had a greater effect than laser therapy on wound healing. However, in both studies UVC was given in combination with either other types of UVL (UVA and UVB18) or together with ultrasound14; therefore, from this previous study, determining the effects of UVC alone on healing of chronic wounds is not possible.

    Ultraviolet light C has been reported in previous in vitro and in vivo studies to have the ability to kill bacteria, including antibiotic-resistant bacteria such as MRSA, in laboratory cells and in animal tissue.20-23 Conner-Kerr et al20 studied the effects of UVC on laboratory cells of MRSA and reported a MRSA kill rate of 99.9% at 5 seconds and 100% at 90 seconds. In a follow-up in vivo study, Conner-Kerr et al21 examined the effects of UVC on experimental wounds placed in mice artificially inoculated with MRSA. Ultraviolet light C was effective in lowering MRSA without adversely effecting healthy wound tissue. However, clinical studies to demonstrate that UVC can affect antibiotic-resistant bacteria that have colonized chronic wounds are not available.

    A technique often used to assess wound bioburden in chronic wounds is the semi-quantitative swab. Clinically, the semi-quantitative swab is the preferred method for bacterial determination because it is economically feasible, non-invasive, and easily administered. The semi-quantitative swab technique, correlated with the gold standard quantitative tissue biopsy, is a valid method for determining bacterial growth in the superficial layers of the wound bed.24,25 Nonetheless, no published reports have documented whether repeated assessments of bacteria burden in chronic wounds using semi-quantitative swabs are reliable and reproducible.

    Hence, the objectives of this prospective study were to: 1) establish the test-retest reliability of the semi-quantitative swab technique; and 2) determine whether a single exposure of UVC for 180 seconds per wound site has an effect on the relative amount of bacteria present in superficial layers of chronic pressure ulcers and leg wounds colonized with bacteria, including MRSA.


    Subject recruitment. Study participants were recruited from an outpatient wound management clinic, residential and inpatient care floors of a local hospital, and area nursing homes. To determine whether a single 180-second exposure of UVC has an effect on the relative amount of bacteria in chronic superficial wounds, 22 participants were recruited into a one-group, pre-test, post-test UVC treatment study. Of the 22 subjects participating in the UVC treatment study, 18 were concurrently enrolled in the semi-quantitative swab test-retest reliability study.

    Ethics approval for research involving human subjects was obtained from appropriate institutional review boards. The purpose, method, risks, and benefits of the UVC treatment and semi-quantitative swab procedure were explained verbally and in a letter of information to the subjects and/or their substitute decision-makers. Informed consent was obtained from all study participants.

    Inclusion and exclusion criteria. During the initial assessment, information including patient demographics, co-existing medical conditions, duration of ulcer, and history of oral and/or topical antimicrobial use was obtained using medical charts and/or through conducting a standard interview with the subjects or their substitute decision-makers. This information was used to screen each subject for the appropriate inclusion and exclusion criteria.

    Inclusion criteria. Patients with chronic ulcers that were present longer than 3 months despite good conventional wound care, superficial, less than 1 cm in depth, with no signs of undermining or sinus tracts were recruited into this study. Only individuals with wounds critically colonized with bacteria, as defined by Sibbald and colleagues,26 were recruited into this study. This included patients who had both a positive wound culture of at least one type of bacteria and two or more clinical signs of wound infection,27,28 including pain/tenderness, moderate exudate, foul odor, friable granulation tissue, slough in the wound bed, peri-ulcer skin erythema, elevated local temperature, and swelling.

    Exclusion criteria. Subjects with signs of systemic infection, cellulitis, septicemia, or osteomyelitis were excluded from the study. Additional exclusion criteria addressed individuals with medical conditions for which UVC treatment is contraindicated (ie, light allergies, photosensitivity, local tuberculosis, previous radiation treatment to the area, skin cancer [current or previous], HIV-positive status, recent skin graft application, acute eczema, or herpes simplex). These contraindications were identified in a confidential manner using a self-report questionnaire.

    Semi-quantitative swab technique. In this study, a standardized protocol for administering the semi-quantitative swab was used to determine the wound bioburden. This protocol involved cleansing the wound with a non-bacteriostatic saline solution29,30 then rotating the sterile cotton-tipped swab side to side across the wound bed from one edge to the other in a “Z” pattern31,32 (see Figure 1). Sufficient pressure was applied to the swab tip to cause the tissue fluid to be expressed.29 The swab then was stored in a sterile tube containing transport media and coded in a manner that concealed the subject’s identity and blinded the laboratory technician as to when the swab was obtained. The swab was transported to a facility licensed for microbiology testing and accredited by the Canadian Association of Physicians. Within 24 hours of receiving the samples, the semi-quantitative swab was inoculated and streaked onto four quadrants on various culture plates. After a 24-hour incubation period, the culture plates were thoroughly analyzed using a standardized medical laboratory procedure. An independent assessor, blinded to the subject’s identity and whether the swab was taken before or after UVC treatment, reported the semi-quantitative swab results as type and relative amount of bacteria.

    Specifically, the relative amount of bacterial growth present in the four quadrants of the culture plate was reported on an ordinal scale from 0 to 4, where zero (0) = no growth, 1 = scant growth, 2 = light growth, 3 = moderate growth, and 4 = heavy growth. It also was determined whether localized bacterial colonies were responsive to various antibiotic treatments or if they were antibiotic-resistant strains of bacteria.

    To establish the test-retest reliability of the semi-quantitative swab, swab results were analyzed to determine if two repeated measurements taken either pre-UVC or post-UVC from the same wound by the same trained investigator were reproducible. Nine subjects had repeated semi-quantitative swabs taken before UVC treatment and in nine subjects the repeat swabs were taken after UVC treatment. The time elapsed between repeated swabs taken to determine test-retest reliability was approximately 5 minutes.

    To determine whether UVC had an effect on the relative amount of bacteria present in chronic wounds, semi-quantitative swabs were taken before and immediately after UVC treatment (approximate time elapse of 5 minutes), and the results were compared.

    UVC treatment protocol. A 254-nm cold quartz generator (low-pressure argon gas/liquid mercury lamp, provided by Medfaxx Inc., Raleigh, NC) was used in this study. The UVC generator was pre-warmed for 5 minutes before placement over the ulcer. While the UVC lamp was warmed, the wound bed was cleansed with sterile saline, a thick layer of petroleum jelly was applied to the surrounding peri-ulcer skin and any islands of healthy granulation tissue, and the wound edges were covered with a drape. To shield their eyes from UVC irradiation, the investigator and subject wore protective goggles. The skin of the therapist also was protected from unwanted exposure to ultraviolet light rays.

    Using an application technique described previously in the literature,14,33 UVC was applied at a distance of 1 inch and perpendicular to the wound using pre-measured disposable spacers (see Figure 2). The UVC generator was applied for a treatment time of 180 seconds per wound site. This exposure time was selected based on the results from previous in vitro studies.34,20 Furthermore, a UVC treatment time of 180 seconds corresponds to an erythemal dosage level of E4, which has been suggested for the treatment of infected ulcers.14 All products applied to the subject were sterilized or discarded after a single use. Equipment that had to be reused was decontaminated using appropriate protocols.

    Data analysis.
    Objective 1: Test-retest reliability of semi-quantitative swab. Laboratory notes were consulted to determine the degree of agreement between the initial and the repeat swab results. Because results were reported in a non-parametric ranking scale from 0 to 4, the test-retest reliability of the semi-quantitative swab technique was determined using Cohen’s kappa35 (SPSS 10.0 for Windows 2000). According to Landis and Koch,36 a kappa value greater than 0.75 is sufficient to consider the technique reliable.

    For each of the 18 subjects recruited into the test-retest reliability study, only one wound per subject and one type of bacteria per wound were entered into analysis. For wounds in which more than one type of bacteria was isolated, only the predominant type of bacteria was used in the analysis. The predominant type of bacteria was selected according to a pre-determined priority list of bacteria based on bacterial virulence and prevalence in chronic wounds.37,22 Methicillin-resistant S. aureus was considered first, followed by S. aureus, P. aeruginosa, and other types of bacteria such as Streptococcus group B and G.

    Objective 2: Effect of UVC on bacteria. To determine if UVC had an effect on bacteria, semi-quantitative swab results obtained before and immediately after UVC treatment were compared. The Wilcoxon Signed-Rank Test (SPSS 10.0 for Windows 2000) was chosen for statistical analysis because the study involved one comparison within a matched sample and results were reported in an ordinal ranking scale.38 For each of the 22 subjects recruited into the pre-test/post-test UVC treatment study, only one wound per subject and one type of bacteria per wound were entered into analysis. For wounds containing more than one type of bacteria, a similar pre-determined priority list of bacteria (noted above) was used to determine the bacteria to be monitored for change.

    A subanalysis also was performed to monitor for changes observed following UVC treatment for each individual type of bacteria, including MRSA, S. aureus, P. aeruginosa, and Streptococcus group B and G. Descriptive statistics were used to report the change in semi-quantitative swab results obtained before and after UVC treatment.

    In addition, the percentage of wounds that changed from having heavy growth (4) to each of the following categories post UVC treatment — moderate (3), light (2), scant (1), and no growth (0) — was reported. The percentage of wounds treated with UVC that did not exhibit a change in the relative amount of bacteria post UVC treatment also was recorded. Descriptive statistics were used to report the bacterial characteristics for wounds that displayed no change in the relative amount of colonization following UVC treatment.


    Description of sample population. At the time of recruitment, 32 persons with chronic wounds were identified as having signs of clinical infection and were subsequently screened for a positive bacterial culture. Of the 32 potential subjects, 22 with clinical signs of infection also had positive swab cultures. The sample population recruited into this study was relatively elderly (mean age = 71.8 years) and included an equal number of men and women. At the time of study recruitment, their chronic wounds were present on average for more than 5 months.

    Four subjects had more than one wound present, from which one wound was randomly selected for study. All common types of chronic wounds were represented, including pressure ulcers, venous leg ulcers, and diabetic foot wounds. The majority of the ulcers studied were located on the leg and attributable to venous and/or arterial insufficiency (see Table 1). Many of the 22 study participants also were receiving concurrent therapy of oral and/or topical antimicrobials (see Table 1).

    The wounds were colonized with several different types of bacteria, including MRSA. Other than a few wounds that had multiple types of bacteria, the majority of wounds in the study contained a heavy growth of one type of bacteria, termed predominant bacteria (see Table 2).

    Test-retest reliability of the semi-quantitative swab. Of the 22 subjects recruited into the UVC treatment study, 18 were enrolled in the test-retest reliability study. Cohen’s kappa statistic indicated excellent test-retest reliability (0.92) for the semi-quantitative swab technique — in fact, identical results were obtained across the two test occasions for 17 of the 18 subjects.

    Effect of UVC on bacteria. A statistically significant reduction of predominant bacteria was noted following a single UVC treatment (P <0.0001, n = 22) (see Figure 3). Furthermore, significant reductions of MRSA (P <0.05), S. aureus (P <0.01), and other types of bacteria (combination of P. aeruginosa and Streptococcus group B and G, P <0.05) were noted (see Figure 3).

    The reduction in relative amount of MRSA colonization produced by a single exposure to UVC tended to be less than UVC-induced reductions observed in other types of bacteria. The majority of the wounds colonized with MRSA initially contained a heavy growth of bacteria — UVC treatment reduced the relative amount of bacteria by only one level. This meant that a moderate growth of MRSA remained in the wound after a single UVC treatment. For the 11 wounds initially colonized with S. aureus, only one still had a heavy growth of Staphylococcus following UVC treatment. Of the wounds initially colonized with a heavy growth of P. aeruginosa, 50% contained no bacterial growth after UVC exposure.

    The relative amount of bacteria did not increase in any of the subject’s wounds following UVC treatment. However, although a single 180-second UVC treatment reduced the relative amount of bacteria in chronic wounds, it seldom was sufficient to totally eliminate them. The majority of the 22 wounds treated with UVC exhibited a decrease in predominant bacteria of greater than one level; the growth level of bacteria remaining in wounds post treatment was light or scant. In six wounds, at least one type of bacteria isolated did not show a reduction in the relative amount of bacteria following UVC treatment. All of the wounds that failed to respond to UVC contained either S. aureus or MRSA. Additionally, two or three of the instances (67%) where the relative amount of bacteria was not reduced by UVC treatment involved wounds that contained more than one type of bacteria. Two subjects had wounds containing MRSA that were not affected by UVC treatment and both had failed to respond to other antimicrobial therapy, including either oral antibiotics or topical antimicrobial medication.


    The semi-quantitative swab obtained using a systematic technique was found to be a reliable method for measuring wound bioburden. The semi-quantitative swab technique had excellent test-retest reliability (kappa = 0.92). This finding strongly suggests that the semi-quantitative swab technique provides consistent and reproducible assessment of the type and relative amount of bacteria present on superficial layers of chronic wounds.

    A quick, inexpensive, and relatively non-invasive method of assessing wound bioburden, the semi-quantitative swab is often the procedure of choice for determining bacterial burden in the clinical setting. However, in comparison to tissue biopsy, which evaluates for the presence of bacteria deep within the tissue, the semi-quantitative swab determines bacterial presence only in the superficial layers of the wound bed. Previous studies have suggested a correlation between the results of semi-quantitative swab and quantitative biopsy. Furthermore, these studies have reported the semi-quantitative swab to be a valid method for determining bacterial burden.24,25

    Results of this study also suggest that a single 180-second treatment of UVC was able to kill bacteria, including antibiotic-resistant bacteria such as MRSA, present in all types of chronic superficial wounds, including pressure, diabetic, venous, and arterial ulcers. However, findings from this study suggest that the response to UVC may be dependent on the type and relative amount of bacteria present initially in the wound bed.

    In the present study, wounds that did not respond to UVC were more heavily colonized with multiple types of bacteria. This finding is not consistent with the results reported in the in vitro study by Sullivan et al,23 who found that a shorter treatment time for UVC was required when more than one type of bacteria was present in a culture plate. Sullivan et al proposed that in mixed culture conditions, when more than one type of bacteria is present, competition between different types of flora “is important for keeping pathogenic organisms at bay in the wound.” This competition may occur within the physical limitations artificially created when placing bacteria within a culture plate; however, similar competitive interactions may not occur in chronic wounds where no finite limitations to the substrate and oxygen supply exist. Differences observed in the present clinical study compared to in vitro studies performed by Sullivan et al could be explained by these changes to the competitive environment or by many of the other differences that exist between in vitro cell culture studies using cultures of commercially acquired strains of bacteria and naturally occurring flora obtained from swabbing chronic wounds.

    Results presented here also suggest that a longer UVC exposure time is required to completely eradicate multiple types of bacteria found in chronic wounds. Conner-Kerr et al20 reported complete elimination of bacteria in animal wounds within 5 to 90 seconds of UVC treatment. It was noted in this present study that a single 180-second exposure of UVC was able to reduce the relative amount of bacteria but could not eradicate all bacteria for wounds initially colonized with heavy amounts of bacteria. Longer UVC treatment times and repeated exposures to UVC may be required clinically because in situ bacteria invade host tissues and, therefore, are located at different depths beneath the wound surface. The three-dimensional heterogeneous wound environment is different from the uni-dimensional nature of cultured bacteria in vitro. Repeated UVC treatments are probably required because the shorter wavelength rays of UVC can penetrate only superficial layers of tissue — a single UVC treatment would not eliminate bacteria located in deeper compartments of the wound.

    Results of this study also suggest that UVC has a differential effect for specific types of bacteria. The change in relative amount of bacteria following UVC treatment tended to be less for MRSA and Streptococcus group B and G than for P. aeruginosa. Such findings are comparable to previous studies by Conner-Kerr et al21 and Sullivan et al,23 who reported an in vitro kill rate of 99.9% for P. aeruginosa at 3 seconds, compared to longer treatment times for MRSA and Streptococcus. Furthermore, Sullivan et al reported that shorter exposure times of UVC were detrimental to procaryotic and simple unicellular eucaryotic-like Pseudomonas organisms while sparing more complex multicellular organisms such as human cells. These researchers suggested that differences in sensitivity of micro-organisms to UVC might be explained by the differences in cell wall structure, replication rate, and multicellular versus unicellular arrangements.

    The primary mechanism by which UVC has been postulated to cause bacterial cell destruction involves induced structural changes to DNA.39 Sullivan et al23 suggested that the faster replication rate and DNA synthesis of bacteria renders them more sensitive than mammalian cells to UVC. Furthermore, these researchers noted that in procaryotes (bacteria), the genetic material is located freely in the cytoplasm; whereas, in eucaryotic cells, the genetic material is surrounded by both the plasma membrane and the nuclear membrane of the nucleus. It has been proposed that because bacteria do not have an additional physical barrier to protect their genetic material from UV irradiation, they are more vulnerable to the effects of UVC than mammalian cells.23 Conner-Kerr et al21 observed that UVC was able to selectively kill antibiotic-resistant bacteria such as MRSA present in the wound tissue of rats without detrimentally affecting healthy granulating tissue.

    This study was not designed as a randomized controlled trial (RCT) in which subjects are randomly allocated to intervention and placebo-treated groups. Rather, one group was selected for a pre-test, post-test study design to assess the effects of UVC on bacteria colonization of chronic wounds. Other factors might explain the observed reduction in bacteria following UVC treatment — for example, exposing the wound to the surrounding environment during the removal of the dressings could have influenced the treatment outcome. However, this is unlikely, given the short time frame between swabs (approximately 5 minutes). Furthermore, inherent changes within the subjects were unlikely due to the short period of the intervention.

    In this study, a 180-second UVC treatment time, which corresponds to an erythemal dosage level of E4, was used. This exposure time was not associated with the occurrence of blistering or burns post treatment. Future studies may examine UVC effects on bacteria using shorter or longer exposure times. In addition, future pre-clinical and clinical studies are needed to determine whether eliminating bacteria from chronic wounds using multiple UVC treatments is associated with accelerated wound healing.


    The semi-quantitative swab has been shown to be a reliable method for determining bacterial burden within the superficial layers of the wound bed and has demonstrated excellent test-retest reliability. Furthermore, a single UVC treatment of 180 seconds significantly reduced the relative amount of bacteria in several different types of chronic wounds, including pressure ulcers, venous ulcers, diabetic ulcers, and arterial ulcers. Future research involving well-designed, randomized, controlled clinical trials is needed to confirm that successive treatments of UVC can eliminate MRSA growth in chronic wounds. Properly designed clinical studies also are required to determine if UVC treatment that reduces the relative amount of bacteria present in the wounds ultimately results in accelerated wound closure.