Skip to main content

The Effects of Ultraviolet C Irradiation in the Treatment of Chronic Wounds: A Retrospective, Descriptive Study

Empirical Studies

The Effects of Ultraviolet C Irradiation in the Treatment of Chronic Wounds: A Retrospective, Descriptive Study

Index: Wound Management & Prevention 2019;65(7):16–22 doi: 10.25270/wmp.2019.7.1622


Although there is evidence supporting the bactericidal effects of ultraviolet C (UVC) in chronic wounds, clinical studies examining the effects of UVC on wound healing are limited. Purpose: The objective of this single-center, retrospective, descriptive study was to evaluate the effect of UVC as an adjunct modality to standard wound care. Methods: Data from January 1, 2015, through August 30, 2015, from all patients receiving UVC treatment and wound etiology-appropriate standard care provided twice weekly at a wound care center and lymphedema clinic were extracted from the electronic health records. Demographic (age, gender) and clinical wound data (duration, type, initial and final dimensions, number of UVC treatments, time to healing, healed versus not healed, and adverse events) were extracted, deidentified, and analyzed using descriptive and chi-squared analysis and analysis of variance. Wound type categories with small sample sizes were combined for analysis. Results: Among the 127 patients studied (64 women [50.4%], average age 69 ± 15.0 years; and 63 men [49.6 %], average age 61 ± 15.8 years; P = .03), 224 wounds received UVC treatment. Initial average wound area was 24.1 ± 121.8 cm2, and the most common type of wound was venous (58, 25.9%), followed by neuropathic/diabetic foot ulcer and traumatic wounds (43 each, 19.2% each). Mean wound duration was 57.7 ± 63 days; overall average patient age was 64.8 ± 15.9 years. Average number of treatments of 11.9 ± 13.4, and average time to healing was 45.2 ± 44.4 days. A total of 127 patients with 161 wounds (71.9%) healed after an average of 45 (range 4–260) days, including 36 venous leg ulcers (62.1%) and 33 neuropathic wounds/diabetic foot ulcers (76.7%). No adverse events were recorded. Conclusion: UVC may be an effective adjuvant treatment modality when used with standard care in the treatment of chronic wounds. Prospective studies are needed to ascertain the effectiveness and efficacy of UVC in the treatment of chronic wounds.



Nonhealing, chronic wounds (ie, wounds that have failed to proceed through an orderly and timely reparative process) present significant clinical challenges to health care professionals. In the United States, chronic wounds affect approximately 6.5 million patients (2% of the US population)1; among these, venous leg ulcers  (VLUs, wounds due to chronic venous insufficiency) affect between 500 000 and 2 million people annually.2 Diabetes affects more than 29.1 million people (approximately 9.3% of the US population), and it is estimated that up to 25% of all people with diabetes will develop a diabetic foot ulcer (DFU) during their lifetime.3,4 These staggering numbers demonstrate a clinical need to provide effective treatment interventions in order to accelerate healing in chronic wounds. VLUs and DFUs represent the majority of chronic wounds treated each year in wound centers.2,4 

Ultraviolet (UV) light has been used for many years as a treatment modality for chronic wounds and a variety of skin conditions. UV energy encompasses wavelengths between 180 nm and 400 nm with 3 distinct bands: UVA (400–315 nm), UVB (315–280 nm), and UVC (280–100 nm).5 Although experimental data6 from as early as 1945 suggest the positive effects of both UVA and UVB in wound healing, chronic, prolonged exposure to high-dose UVB has been linked to carcinoma formation.7,8 In contrast, UVC has been shown to be safe to use in a clinical setting,5 with evidence-based research demonstrating bactericidal effects.9 

A quantitative analysis by Conner-Kerr et al9 examined the effectiveness of UVC radiation on antibiotic-resistant strains of Staphylococcus aureus and Enterococcus faecalis in vitro. Between 2 and 5 replications of each organism at 108 organisms/mL were prepared and plated on a sheep blood agar medium. Organisms on agar plates were treated with UVC (254 nm, 15.54 mW/cm2 output), resulting in a 99.9% kill rate with irradiation times of 5, 8, 15, 30, 45, and 60 seconds and a 100% percent kill rate at irradiation times of 90 and 120 seconds for a methicillin-resistant strain of S aureus (MRSA). In addition, 99.9% kill rates for vancomycin-resistant E faecalis were identified at irradiation times of 5, 8, 15, and 30 seconds.9

Thai et al10 presented a case series describing 3 individuals with chronic wounds that were locally positive for MRSA. In this study,10 UVC (using a 254-nm cold quartz UVC generator) was applied at a perpendicular distance of 1 inch from the wound base. Wounds were treated for 180 seconds; semiquantitative swab results taken before and following 7 UVC treatments showed a reduction in bacteria growth. In 2 of the cases, heavy growth of MRSA was identified before UVC treatment, and after 7 UVC treatments, light growth was noted. In the third case, light growth of MRSA was identified before UVC treatment and occasional growth (scant) of MRSA was noted after 7 UVC treatments. Two (2) of the 3 wounds were healed following 1 week of UVC treatment, and the third wound showed evidence of healing. 

A prospective, 1-group, pre-post treatment study by Thai et al11 found UVC was effective in reducing bacteria in chronic wounds (pressure ulcers [PUs], VLUs, and DFUs). Participants included individuals (N = 22) with chronic ulcers exhibiting at least 2 clinical signs of infection (pain/tenderness, moderate exudate, foul odor, friable granulation tissue, slough in the wound bed, periulcer skin erythema, elevated local temperature, and swelling) and critically colonized as defined by Sibbald et al,12 as well as patients who had both a positive wound culture of at least 1 type of bacteria. These individuals received a single 180-second treatment using an UVC light lamp (wavelength = 254 nm) placed 1 inch from the wound bed. Semiquantitative swabs taken immediately before and after UVC treatment demonstrated excellent test-retest reliability of the semiquantitative swab technique used to evaluate the type and amount of bacteria present in chronic wounds (Cohen’s kappa = 0.92). A statistically significant (P <.0001) reduction in the relative amount of Pseudomonas aeruginosa, S aureus, and MRSA following a single treatment of UVC was observed. The greatest reduction in semiquantitative swab scores following UVC treatment were observed for wounds colonized with P aeruginosa and wounds colonized with only 1 species of bacteria. Significant (P <.05) reductions in the relative amount of bacteria also were observed in 12 ulcers in which MRSA was present, but the effects of UVC were not as robust for MRSA (only 1 level reduction) after the standard 180-second treatment time. These data indicate multiple treatments of UVC for 180 seconds may be required to eradicate organisms that are sensitive to antibiotics.

Although evidence is strong for the bactericidal effects of UVC in chronic wounds in vitro, clinical studies examining the effects of UVC, specifically the effect on wound healing, are limited. In a 1995 review, Kloth13 summarized modalities that have been shown to have a positive effect on nonhealing, slowly healing, or regressing wounds as well as the effects of UVC on wound healing including hyperplasia, reepithelialization or desquamation of the leading edge of periulcer epidermal cells, granulation tissue formation, and sloughing of necrotic tissue. The literature review by Rennekampff et al14 noted UVC also may stimulate and restore normal melanocyte numbers and distribution in reepithelialized wounds while preventing hypopigmentation. However, few studies have examined the healing response on chronic wounds of various etiologies treated with UVC. The purpose of this retrospective, descriptive study was to evaluate the outcomes of UVC as an adjunct modality when used with standard care for chronic wounds.


With the approval by the Medical Executive Committee of Northern Hospital of Surry County (NHSC), Mt. Airy, North Carolina, a retrospective chart review was performed of data from all outpatients who received UVC treatment in addition to standard care and whose treatment was completed (defined as patients who were discharged from the wound center regardless of reason for discharge) within an 8-month assessment period (January 1, 2015, through August 30, 2015) at The Wound Care Center and Lymphedema Clinic of NHSC. Extracted variables included patient age and gender, wound duration (how long the patient had the wound before initiation of treatment) and type (etiology), initial and final wound dimensions (calculated based on length and width of each wound extracted from the electronic health record), number of UVC treatments, time to healing (defined by the primary investigator as 100% reepithelialization and calculated from start to end date of treatment), final wound status (healed or unhealed), and occurrence of treatment-related adverse events. All patient data were deidentified during abstraction (following Health Insurance Portability and Accountability Act guidelines) from Meditech (Medical Information Technology, Inc, Westwood, MA) into an Excel (Microsoft Corp, Redmond, WA) file by the primary investigator.

Standard care per specific wound etiologies. All wounds were provided standard care appropriate to their etiology each time UVC treatment was performed.

VLUs. Wounds were cleaned with pulsatile lavage at 6 psi with concurrent suction and 0.9% normal saline as an irrigation solution or a saline wash, with a primary wound dressing such as carboxymethylcellulose (CMC)/alginate selected to maintain a moist wound healing environment, covered as necessary with a secondary dressing, and provided multilayer compression.

Neuropathic/DFUs. Wounds received sharp, selective debridement as required, cleansed with pulsatile lavage at 6 psi with concurrent suction and 0.9% normal saline as an irrigation solution or a saline wash, and covered with a primary wound dressing to facilitate a moist wound healing environment and a gauze/absorbent pad as a secondary dressing. Total contact casts, orthopedic wedge shoes, and heel offloading shoes were provided as needed. 

Pressure ulcers/injuries (PU/Is) and all other wound types. Wounds were cleaned with pulsatile lavage at 6 psi with concurrent suction and 0.9% normal saline as an irrigation solution or a saline wash with a primary wound dressing selected to maintain a moist wound healing environment and covered with an appropriate secondary dressing. Support surfaces and pressure redistribution devices were utilized for all PU/Is.

Dressing selection. Primary wound dressings (applied directly to the wound bed) were selected based on recommendations for clinical assessment of superficial infection and treatment of chronic wounds according to the mnemonic, NERDS, as recommended by Sibbald et al12 in which N = nonhealing wounds, E = exudative wounds, R = red and bleeding wound surface granulation tissue, D = debris (yellow or black necrotic tissue) on the wound surface, and S = smell or unpleasant odor from the wound. If 2 or more of the signs of infection were present, a primary dressing such as CMC/alginate dressing containing silver was applied. Otherwise, a CMC/alginate dressing without silver was applied to the wound. 

Secondary dressings were selected based on wound size and amount of exudate. These included absorbent pads, cotton gauze, and a self-adhering wrap. Multilayer compression was applied to patients with VLUs. 

Procedure for UVC application. Before dressing application, UVC was applied at a distance of 1 inch perpendicular to the wound base for 90 seconds for each wound after cleansing. A model V-254 UVC lamp (Medfaxx Inc, Raleigh, NC) was used for each treatment (see Figure 1). This model lamp has an output intensity of 4 mW/cm2 with a therapeutic dose of 120 to 360 mJ/cm2.15 In preparation for treatment, the periwound can be protected with draping materials as well as applying petrolatum jelly. However, it has been argued that protecting the adjacent periwound area eliminates a potential source of epithelial cells and wound healing factors that have been known5 to be stimulated with treatment. Therefore, in the authors’ clinic, the periwound was not protected. The periwound was assessed after each treatment for any adverse reaction to UVC. Although manufacturer information15 states shortwave UV is generally considered harmless to the average person, patients were assessed before and after each treatment for photosensitivity. Care also was taken to ensure eyes were protected from the light beams, even though eye protection is not needed with this model of UVC lamp because very little UVC is produced at a distance of 3 to 4 inches away from the filter, certainly not enough to create a basal cell response (as reported by Medfaxx, Inc).16 Although the lamp did not contact the wound or patient’s skin, the lens was cleaned with saline after each treatment and the wand handle was disinfected with a germicidal wipe. UVC is contraindicated in the presence of melanoma or a history of melanoma, history of invasive squamous cell carcinomas, human immunodeficiency virus, fever, deep x-ray therapy, and light sensitive diseases such as, but not limited to, porphyria and lupus erythematosus.16

Patients received treatment as described twice per week until the wounds were healed as determined by the primary investigator or the patient was discharged. All initial wound assessments (including initial measurements) and all final wound assessments were completed by the primary investigator; 2 investigators entered the data into the patient records. UVC was added to treatments on January 1, 2015, when the instrumentation became available at the authors’ facility.

Statistical analysis. Initially, wound types classified according to the etiology of the wound resulted in 15 different categories. However, several of these categories had <10 wounds (most had <5), which limited subsequent analysis. Nonparametric analysis showed no significant differences in data among the smaller groups. For a more robust analysis, several of the smaller groups were combined into 1 group, which then allowed for a more complete analysis. The wound etiologies that were combined included cat scratch, dog bite, MRSA infections, burn, abscess, wounds of mixed etiology (multiple contributing etiologies), and surgical wounds. Analyses included descriptive statistics for all dependent variables, and means are presented with 1 standard deviation. A chi-squared analysis was used to determine the association between gender and wound category. If a patient had more than 1 wound, each wound was analyzed separately/independently. Analysis of variance (ANOVA) was used to determine differences between wound classification groups for time to healing and number of treatments. Differences were considered significant if P ≤.05. All analyses were performed using SPSS, version 23 (IBM, Armonk, NY).


A total of 127 patient files, representing 224 wounds, were analyzed. Gender was distributed equally (64 women [50.4%], 63 men [49.6 %]), with a significant difference in ages between genders (average age of women 69 ± 15.0 [range 25–93] years, average age of men 61 ± 15.8 [range 11–92] years; P = .03). Combining groups with smaller numbers for statistical analysis (abscess = 2, arterial = 2, burn = 3, cat scratch = 1, dog bite = 2, hematoma = 1, MRSA = 9, surgical = 7) resulted in 7 groups by wound type (see Table 1). The most common type of wound was venous (58, 25.9%), followed by neuropathic/DFU and traumatic wounds (43 each, 19.2% each), representing 64% of the total number of wounds. Following the combination of groups, the smallest group was postoperative surgical dehiscence wounds (10, 4.5%). Other than age differences between the genders, no significant differences were noted. 

An overall healing rate of 71.9% (161 of 224 wounds) was achieved for all wound types. A “not healed” status was assigned to all patients whose final wound status could not be confirmed. Patients in the “not healed” category included patients whose care was initiated by the author’s clinic but were managed by home health services or skilled nursing facilities and did not return to the author’s clinic for assessment; in addition, patients who were hospitalized for nonwound-related pathologies and outpatients who did not return to the author’s clinic to confirm wounds were 100% healed were considered “not healed.” Table 2 presents the percent of wounds healed by type, including VLUs (36, 62.1%), traumatic wounds (36, 83.7%), wounds from cellulitis (8, 72.7%), wounds of unknown etiology (10, 90.9%), dehiscence of surgical wounds (9, 90.0%), neuropathic wounds/DFUs (33, 76.7%), PU/Is (14, 66.7%), and combined groups (16, 59.3%). Chi-squared analysis (frequency) showed no significant difference in healing status by wound type.

Frequency of wounds per patient ranged from 1 to 7, with 102 (54%) having 1 to 2 wounds, 19 (8.5%) with 3 to 4 wounds, and 6 (3%) having >4 wounds. Persons with more than 1 wound were found among each wound etiology category. Data on patient age, gender, wound duration (chronicity), wound area, number of treatments, and time to healing are presented in Table 3. Average wound duration was 57.7 ± 63.0 (range 1–277) days; average patient age was 64.8 ± 15.9 (range 11–93) years. Average initial wound area was 24.1 ± 121.8 cm2 (range .01–1350 cm2), average number of treatments was 11.9 ± 13.4 (range 1–116), and an average time to healing was 45.2 ± 44.4 (range 4–260) days. Mean time to healing by wound type is presented in Figure 2. Initial exploration of the data distribution showed several individuals whose values were statistical outliers (defined as more than 3 standard deviations outside the mean). This variability can be seen in Table 2 and is illustrated in Figure 2 for time to healing by wound type. To maintain the integrity of the data, no outliers were removed. 

ANOVA showed no significant differences among wound category groups for number of treatments or wound duration. However, a significant difference was noted between the wound category groups for time to healing (P = .01). Post-hoc analyses were unable to distinguish any differences between specific subgroups. No adverse events were noted with regard to UVC utilization. 


This retrospective, observational study evaluated the effects of UVC as a therapeutic modality when used with standard care in the treatment of chronic wounds of various etiologies (VLUs neuropathic/DFUs, traumatic, wounds due to cellulitis, PU/I, postoperative wound dehiscence, etiologies unknown, and combined groups of wound etiologies). Overall, 71.9% of all wound types (91 patients, with some patients also having wounds that were unhealed) healed with an average of 11.8 treatments and an average time to healing of 45.2 days. Average wound chronicity was 57.7 days, with an average wound area of 24.1 cm2 for all wound types. Although the average wound chronicity in this study was close to 58 days, wounds with shorter durations (1 day for women and 4 days for men; see Table 2) also were included in the data analysis. These wounds, which were not present initially but formed after initiation of treatment, were identified during the patient’s course of treatment.  

An extensive literature review17 of publications with a moderate or strong rating for internal validity showed individuals who received standard care for VLUs required an average of 168 days (24 weeks) to heal, approximately 15% never heal, and wounds recur once or multiple times in 15% to 71% of cases. According to a systematic review18 and a prospective randomized trial,19 40% to 70% of VLUs heal over 3 months and 50% to 80% heal over 6 months. A descriptive study20 showed 61% of VLUs healed using human skin allografts in 84 days (12 weeks), a literature review21 showed 67.6% healed using a cryopreserved placental membrane in 84 days (12 weeks), and a prospective, randomized controlled study22 showed 57% healed with the use of a bilayered living skin construct in 84 days (12 weeks). In the current study, 62.1% of VLUs healed, which is similar to previously published rates, but the mean time to healing (32.6 days) was shorter than the published healing rate of 84 days (12 weeks). 

According to a prospective study,23 neuropathic/DFUs provided standard care that included debridement and offloading had an average healing time of 77.7 days (95% CI: 62-93). More recently, a prospective study24 reported 59.3% of neuropathic/DFUs healed at 84 days (12 weeks), 70.5% healed at 140 days (20 weeks), and 86.6% healed at 364 days (52 weeks). In addition, a comparative analysis25 of a bioengineered living cellular construct versus a dehydrated human amniotic membrane allograft for the treatment of neuropathic/DFUs found 48% and 72% wounds healed at day 84 and day 168 (weeks 12 and 24), respectively, for the bioengineered living cellular construct versus 28% and 47% for dehydrated human amniotic membrane allograft, respectively. In the current study, 76.7% of neuropathic/DFUs wounds healed in a mean time to healing of 56 days. 

In the current study, healing rates are slightly lower because data on significant outliers were not deleted (ie, percent of wounds healed without outliers would be even shorter). Suggested quality measures from the U.S. Wound Registry26 for 2016 include that achieving 30% closure of VLUs and DFUs at 28 days (4 weeks) of care can serve as a predictor of wound healing. In a comparative study by Fife et al,27 real-world healing rates were identified for VLUs, DFUs, and PUs from the U.S. Wound Registry. The dataset, analyzed from September 28, 2001, through December 1, 2016, which included 99 588 VLUs, 77 891 DFUs, and 77 891 PU/Is, showed at 12 weeks ~30% of DFUs/PU/Is were healed and ~45% of VLUs were healed. 



Limitations of this study include smaller sample sizes of certain wound etiologies that were combined for statistical analysis and the inherent limitations of a retrospective study. The researchers attempted to reduce study bias by including all patients who received UVC treatment and whose treatment was completed within the 8-month assessment period. In addition, initial or last available measures were included for all patients. The presence of statistical outliers, while not a factor in the analyses, can skew the averages for the population healing times. Another limitation was that wound status variables other than wound size were not abstracted nor were comorbid conditions considered. Also, several patients had more than 1 wound; hence, they were not independent variables, which can skew the results.


UVC is a therapeutic modality that has been shown to be effective at reducing bacteria in chronic wounds of various etiologies. The purpose of this retrospective, descriptive study was to evaluate the outcomes of UVC as an adjunct modality when used with standard care for chronic wounds. In this study involving 127 patients with 224 wounds, 71.9% of wounds healed after an average of 45 days (range 4–260 days) including 36 VLUs (62.1%) and 33 neuropathic wounds/DFUs (76.7%). The latter data compare favorably with previously published healing rates of VLUs and DFUs. No adverse events were recorded. Although the results of this study are promising, prospective effectiveness and efficacy studies are needed to confirm the results of this descriptive study. 


Dr. Yarboro is Director of Wound Care Services, Northern Hospital of Surry County, Mount Airy, NC. Dr. Millar is Director of the Doctor of Physical Therapy program, Winston-Salem State University, Winston-Salem, NC. Mr. Smith is a physical therapist assistant, Northern Hospital of Surry County.


Please address correspondence to: Douglas Yarboro, DPT, MBA, CWS, ICLM, Northern Hospital of Surry County, 314 S. South Street, Suite 200, Mount Airy, NC 27030; email: