Noncontact Normothermic/Nonthermal Wound Therapy and Noncontact Fluorescence Imaging of Wounds

Number: 0372

Policy

Aetna considers Warm-Up Active Wound Therapy, also known as noncontact normothermic wound therapy (NNWT) and warming therapy, experimental and investigational because there is insufficient evidence in the peer-reviewed medical literature that Warm-Up Active Wound Therapy/NNWT is effective for the treatment of wounds.

Aetna considers noncontact, nonthermal, low-frequency ultrasound therapy experimental and investigational for the treatment of wounds and all other indications (e.g., bacterial infections, deep tissue pressure injury, and femoral artery thrombosis) because its effectiveness has not been established.

Aetna considers the MolecuLight, a hand-held device for point-of-care fluorescence imaging of wounds, experimental and investigational because its effectiveness has not been established.

Background

Warm-Up Active Wound Therapy (Augustine Medical, Inc., Eden Prairie, MN), also known as noncontact normothermic wound therapy (NNWT) uses a non-contact radiant-heat bandage to treat chronic venous ulcers when conventional wound-healing therapy has failed.  The device consists of a noncontact, domed wound cover into which a flexible infrared heating card is inserted.  A battery pack powers the device and warms the wound to a pre-determined temperature.  The inside of the wound cover contains a foam ring, which acts as a wick to drain away exudate.

In a Decision Memorandum, the Center for Medicare and Medicaid Services (CMS) reviewed the evidence of the effectiveness of noncontact normothermic wound therapy.  The CMS concluded that “the medical literature does not support a finding that NNWT heals any wound type better than conventional treatment.”  CMS concluded, therefore, that there is insufficient evidence in the peer-reviewed medical literature to consider this device as reasonable and necessary for the treatment of wounds.

An assessment of treatments for chronic pressure ulcers by the Ontario Ministry of Health and Long-Term Care (2009) concluded that "thermal dressings such as noncontact normothermic dressings or radiant heat dressings were associated with greater improvement in stage III and IV pressure ulcers; however, this did not translate into more wound closure. There is no evidence at present to conclude that thermal dressings will result in more complete healing in stage III or IV pressure ulcers."

Several recent studies have evaluated the effectiveness of NNWT for the treatment of chronic wounds.  However, there are drawbacks from these studies -- small sample sizes and lack of long-term follow-up.  McCulloch and Knight (2002) examined the effect of a noncontact, radiant warming device in the treatment of neuropathic foot wounds secondary to diabetes.  Patients (n = 36) were assigned to management with off-loading and warming (treatment) or off-loading therapy only (control) for a period of 8 weeks or until healing.  Wounds of subjects in the treatment group healed at a rate of 0.019 +/- 0.019 cm2/day compared with that of 0.008 +/- 0.009 cm2/day in the control group (p = 0.049).  The difference between treatment and control groups barely reached statistical significance.

The authors of a small (13 patients) preliminary study on Warm-Up® Active Wound Therapy concluded that Warm-Up® Active Wound Therapy is a safe treatment modality for chronic venous stasis ulcers; however, further investigation using a larger prospective study is needed to demonstrate effectiveness

Kloth and associates (2002) studied the effect of NNWT versus standard wound care on patients (n = 40) with 43 stage III and IV pressure ulcers.  A sterile noncontact wound dressing was applied to 21 wounds for 24 hours per day, 7 days per week.  Each day after the wound was irrigated and the noncontact dressing was changed, a heating element in the dressing was activated for 3 1-hour periods for 12 weeks or until wound closure.  Twenty-two control wounds were treated with standard, moisture-retentive dressings 24 hours per day, 7 days per week for 12 weeks or until wound closure.  The healing rate for the treatment group was significantly greater than that for the control group (0.52 cm2 per week and 0.23 cm2 per week, respectively; p < 0.02).  However, the difference in the incidence of closure among wounds that completed the entire 12-week protocol between treatment and control groups were not significant (11 of 14 or 78.5 % for the treatment group and 8 of 16 or 50 % for the control group).

In a pilot study, Karr (2003) studied the use of NNWT in the treatment of wounds associated with osteomyelitis.  This study consisted of 2 arms:
  1. the control arm (11 patients with 11 ulcers) received standard wound care, and
  2. the treatment arm (5 patients with 6 ulcers) received NNWT.
Standard wound care resulted in complete ulcer healing at an average of 127 days, while NNWT resulted in complete ulcer healing at an average of 59 days, or 54 % faster than in the control arm.  However, the mean wound healing times between the 2 groups were not significantly different (p < 0.33).  Moreover, the median wound healing time for the 2 groups were quite similar (70 days for the control group and 68 days for the treatment group).  The authors concluded that a larger prospective study that evaluates NNWT for ulcers associated with osteomyelitis is warranted.

In a prospective, randomized, controlled study, Alvarez and colleagues (2003) compared diabetic foot ulcer healing in patients being treated with either NNWT applied for 1 hour 3 times daily until healing or 12 weeks, or standard care (saline-moistened gauze applied once-daily).  Surgical debridement and adequate foot off-loading was provided to both groups.  Evaluations were performed weekly and consisted of acetate tracings, wound assessment, and serial photography.  A total of 20 patients completed the study and both treatment groups were distributed evenly (n = 10).  Ulcers treated with NNWT had a greater mean percent wound closure than control-treated ulcers at each evaluation point (weeks 1 to 12).  After 12 weeks, 70 % of the wounds treated with NNWT were healed compared with 40 % for the control group.  However, the differences were not significant (p < 0.069).  The authors concluded that further study in a greater patient population is needed to assess the effectiveness of NNWT in treating neuropathic foot ulcers.

In a randomized controlled study, Thomas et al (2005) examined the effectiveness of radiant heat bandage on the healing of stage 3 or stage 4 pressure ulcers.  A total of 41 subjects with a stage 3 or stage 4 truncal pressure ulcer greater than 1.0 cm2 were recruited from outpatient clinics, long-term care nursing homes, and a rehabilitation center.  The experimental group was randomized to a radiant-heat dressing device and the control group was randomized to a hydrocolloid dressing, with or without a calcium alginate filler.  Subjects were followed until healed or for 12 weeks.  Eight subjects (57 %) in the experimental group had complete healing of their pressure ulcer compared with 7 subjects (44 %) with complete healing in the control group (p = 0.46).  The authors noted that although a 13 % difference in healing rate between the 2 arms of the study was found, this difference was not statistically significant.

In a single center randomized study with 49 patients, Alvarez et al (2006) reported that NNWT improves the healing of diabetic neuropathic foot ulcers.  Moreover, these researchers stated that further study in a greater patient population is needed to fully assess the effectiveness of this device and to provide additional information on whether local warmth can reduce the incidence of infection.

Serena et al (2009) examined if noncontact, nonthermal, low-frequency ultrasound (LFU) therapy is effective in controlling wound bacterial colony counts in a series of 4 related experiments.  First, ultrasound penetration in both wounded and intact skin was assessed in-vitro.  Compared to sham, noncontact ultrasound penetrated farther into both wounded (3.0 to 3.5 mm versus 0.35 to 0.50 mm) and intact (2.0 to 2.5 mm versus 0.05 to 0.07 mm, respectively) pig skin.  Second, using an in-vitro model to stain and count live/dead bacteria, 0 % of sham-treated and 33 % of Pseudomonas aeruginosa, 40 % of Escherichia coli and 27 % of Enterococcus faecalis were dead after 1 ultrasound application.  Minimal effects on methicillin-resistant Staphylococcus aureus (MRSA) and S. aureus were observed.  Third, using an in-vivo model, after 1 week, while differences between different bacterial species were observed, overall bacterial quantity decreased with ultrasound treatment (from 7.2 +/- 0.79 to 6.7 +/- 0.91 colony forming units [CFU] per gram of tissue [CFU/g]) and silver anti-microbial dressings (from 7.2 +/- 0.79 to 5.7 +/- 0.6 CFU/g) but increased to 8.6 +/- 0.15 CFU/g for sham and 8.6 +/- 0.06 CFU/g for water-moistened gauze.  Fourth, 11 patients (average age of 60 years) with pressure ulcers containing bacterial counts greater than 10(5) CFU/g of tissue received 2 weeks of noncontact ultrasound therapy.  The quantities of 7 bacterial organisms were reduced substantially from baseline to 2 weeks post-treatment.  None of the wounds exhibited signs of a clinical infection during the treatment period and no adverse events were observed.  Taken together, these 4 studies indicated that noncontact ultrasound can be used to reduce bacterial quantity.  The authors concluded that controlled clinical studies are needed to determine the effectiveness of this treatment and to further elucidate its effects on various Gram-negative and Gram-positive bacteria.

In an in-vitro study, Conner-Kerr and colleagues (2010) examined the effects of LFU delivered at 35 kHz on bacterial viability, cell wall structure, and colony characteristics, including antibiotic resistance on vegetative forms of MRSA.  They concluded that studies to elucidate the observed effects of LFU on MRSA and evaluate its effect in-vivo are needed.  Furthermore, in a Cochrane review on therapeutic ultrasound for venous leg ulcers, Cullum et al (2010) concluded that the studies evaluating ultrasound for venous leg ulcers are small, poor-quality and heterogeneous.  There is no reliable evidence that ultrasound hastens healing of venous ulcers.  There is a small amount of weak evidence of increased healing with ultrasound, but this requires confirmation in larger, high-quality randomized controlled trials.  There is no evidence of a benefit associated with LFU.

Voigt and colleagues (2011) examined if LFU used as an adjunctive therapy improves the outcomes of complete healing and reduction of size of chronic lower limb wounds.  PubMed, Cochrane/CENTRAL, technical assessment, relevant wound-related journals, and clinical guidelines were searched along with contacting manufacturers and authors of relevant randomized controlled trials (RCTs) were completed.  Searches focused on the use of LFU in RCTs.  Data were collected via a data collection form and was adjudicated independently via coauthors.  Meta-analyses and heterogeneity checks were performed using Mantel-Haenszel and inverse variance (fixed and random effects) statistical methods on studies with similar outcomes (complete healing and percent wound area reduction) over similar time periods.  Single study results were reported via the statistical methods used in the study; 8 RCTs were identified.  Results demonstrated that early healing (at less than or equal to 5 months) in patients with venous stasis and diabetic foot ulcers was favorably influenced by both high- and low-intensity ultrasound delivered at a low frequency -- either via contact or noncontact techniques.  However, the authors noted that the quality of the data may be suspect, especially for low-frequency, low-intensity noncontact ultrasound because of significant biases.

Madhok et al (2013) stated that debridement is a crucial component of wound management.  Traditionally, several types of wound debridement techniques have been used in clinical practice such as autolytic, enzymatic, bio-debridement, mechanical, conservative sharp and surgical.  Various factors determine the method of choice for debridement for a particular wound such as suitability to the patient, the type of wound, its anatomical location and the extent of debridement required.  Recently developed products are beginning to challenge traditional techniques that are currently used in wound bed preparation.  These investigators reviewed the current evidence behind the use of these newer techniques in clinical practice.  They noted that there is some evidence to suggest that LFU therapy may improve healing rates in patients with venous ulcers and diabetic foot ulcers.

Low-Frequency Ultrasound Therapy Femoral Artery Thrombosis

Zhu and colleagues (2016) studied the thrombolytic effect of LFU combined with targeted urokinase-containing microbubble contrast agents on treatment of thrombosis in rabbit femoral artery; and determined the optimal combination of parameters for achieving thrombolysis in this model.  A biotinylated-avidin method was used to prepare microbubble contrast agents carrying urokinase and Arg-Gly-Asp-Ser (RGDS) peptides.  Following femoral artery thrombosis in New Zealand white rabbits, microbubble contrast agents were injected intravenously, and ultrasonic exposure was applied.  A 3 × 2 × 2 factorial table was applied to categorize the experimental animals based on different levels of combination of ultrasonic frequencies (Factor A: 1.6-MHz, 2.2-MHz, 2.8-MHz), doses of urokinase (Factor B: 90,000 IU/kg, 180,000 IU/kg) and ultrasound exposure time (Factor C: 30-min, 60-min).  A total of 72 experimental animals were randomly divided into 12 groups (n = 6 per group).  Doppler techniques were used to assess blood flow in the distal end of the thrombotic femoral artery during the 120 minutes thrombolysis experiment.  The rate of re-canalization following thrombolysis was calculated, and thrombolytic effectiveness was evaluated and compared.  The thrombolytic re-canalization rate for all experimental subjects after thrombolytic therapy was 68.1 %.  The optimal parameters for thrombolysis were determined to be
  1. an ultrasound frequency of 2.2-MHz, and
  2. a 90,000 IU/kg dose of urokinase.  Ultrasound exposure time (30 minutes versus 60 minutes) had no significant effect on the thrombolytic effects.

The combination of local LFU radiation, targeted microbubbles, and thrombolytic urokinase induced thrombolysis of femoral artery thrombosis in a rabbit model.  The ultrasonic frequency of 2.2-MHz and urokinase dose of 90,000 IU/kg induced optimal thrombolytic effects, while the application of either 30 minutes or 60 minutes of ultrasound exposure had similar effects.

Low-Frequency Ultrasound Debridement in Chronic Wound Healing

Chang and colleagues (2017) stated that ultrasound debridement is a promising technology that functions to disperse bacterial biofilms and stimulate wound healing.  These researchers focused on LFU (20 to 60 kHz) and summarized the findings of 25 recent studies examining ultrasound efficacy.  Ultrasound debridement appears to be most effective when used 3 times a week and has the potential to decrease exudate and slough, decrease patient pain, disperse biofilms, and increase healing in wounds of various etiology.  The authors concluded that although current studies are generally of smaller size, the results are promising and they recommended the testing of LFU therapy in clinical practice on a larger scale.

Low-Frequency Ultrasound for the Treatment of Bacterial Infections

Cai and colleagues (2017) noted that single anti-microbial therapy has been unable to resist the global spread of bacterial resistance.  These researchers reviewed literatures of available in-vitro and in-vivo studies and the results showed that (LFU has a promising synergistic bactericidal effect with antibiotics against both planktonic and biofilm bacteria.  It also can facilitate the release of antibiotics from medical implants.  The authors stated that as a non-invasive and targeted therapy, LFU has great potential in treating bacterial infections.  However, more in-depth and detailed studies are still needed before LFU is officially applied as a combination therapy in the field of anti-infective treatment.  Moreover, these investigators noted that there is still a long way to go before clinical application of combination therapy of LFU with antibiotics.  First of all, the current studies involved a narrow range of susceptible pathogens.  There are very few studies on the most threatening MDR bacteria.  Secondly, frequency, intensity, and pulse cycle varied a lot at present.  The promising frequency and intensity from in-vitro studies are likely to cause local damage in in-vivo studies.  Therefore, LFU parameters appropriate for clinical application need to be further explored.  Thirdly, 1 study indicated that LFU treatment reduced the interface shear strength and initial stability of vancomycin-loaded acrylic bone cement-stem.  So the impact of LFU on the physical properties of the implant materials requires a comprehensive examination.  At last, because bacteria will partially be removed from the biofilm surface when LFU is applied, whether it will bring the risk of spreading the pathogens and forming systemic bloodstream infection also requires more careful evaluation.

Low-Frequency Ultrasound for the Treatment of Deep Tissue Pressure Injury

Honaker and associates (2016) stated that the optimal treatment for deep tissue pressure injuries (DTPI) has not been determined.  Deep tissue pressure injuries represent a more ominous early stage pressure injury that may evolve into full thickness ulceration despite implementing the standard of care for pressure injury.  In a longitudinal, prospective, historical case control study, these researchers examined the effectiveness of noncontact LFU (NLFU) plus standard of care (treatment group) in comparison to standard of care (control group) in reducing DTPI severity, total surface area, and final pressure injury stage.  The Honaker Suspected Deep Tissue Injury Severity Scale (range of 3 to 18 [more severe]) was used to determine DTPI severity at enrollment (time 1) and discharge (time 2).  A total of 60 subjects (treatment = 30; control= 30) were enrolled in the study.  In comparison to the control group mean DTPI total surface area change at Time 2 (0.3 cm2 ), the treatment group had a greater decrease (8.8 cm2 ) that was significant (t = 2.41, p = 0.014, r2  = 0.10).  In regards to the Honaker Suspected Deep Tissue Injury Severity Scale scores, the treatment group had a significantly lower score (7.6) in comparison to the control group (11.9) at time 2, with a mean difference of 4.6 (t = 6.146, p = 0.0001, r2  = 0.39).  When considering the final pressure ulcer stage at time 2, the control group were mostly composed of unstageable pressure ulcer (57 %) and DTPI severity (27 %).  In contrast, the treatment group final pressure ulcer stages were less severe and were mostly composed of stage 2 pressure injury (50 %) and DTPI severity (23 %) were the most common at time 2.  The authors concluded that the results of this study showed that DTPI severity treated with NLFU within 5 days of onset and in conjunction with standard of care may improve outcomes as compared to standard of care only.

In a retrospective, descriptive study, Wagner-Cox and colleagues (2017) examined the effect of NLFU on DTPI, both hospital-acquired and those present on admission (POA).  Medical records from 44 adult patients with a DTPI treated with NLFU were reviewed; 22 had a hospital-acquired DTPI (HADTPI) and 22 had DTPI POA.  Age of subjects was 71.3 ± 16.3 years (mean ± SD); 52 % were men.  Data were collected from the medical records including demographic as well as relevant clinical characteristics, DTPI measurements, and DTPI evolution/resolution.  Data were summarized and examined using descriptive statistics (e.g., frequencies and percentages and means and standard deviations).  Differences between groups were examined using paired t-tests or the Mann-Whitney U test and the Chi-square test as appropriate.  In addition, the heel DTPI subgroup (n = 8) was examined separately due to the small sample size.  All patients with HADTPI and DTPI POA treated with NLFU exhibited a statistically significant decrease in injury size from initiation to discontinuation of NLFU therapy (24.6 cm versus 14.4 cm, p = 0.02).  No statistically significant difference in wound resolution was found between HADTPI versus DTPI POA (27 % versus 18 %, p = 0.47).  Mean size of both HADTPI and DTPI POA decreased significantly from 15.9 to 13.4 cm (p = 0.045) by NLFU therapy.  Wounds were classified as resolved at completion of treatment in 23 % (10 out of 44) of all treated patients.  Of all patients with the potential to be resolved (not discharged early or expired), 63 % (10 out of 16) had wounds classified as resolved.  The authors concluded that the findings of this study suggested that NLFU is a viable and promising therapeutic option for both HADTPI and DTPI POA.  Moreover, they stated that future studies are needed to confirm these results and to examine efficacy and feasibility of DTPI across care settings.

The MolecuLight Device

The MolecuLight is a wound imaging device that can visualize fluorescent bacteria and measure wound surface area in real-time.  However, there is currently insufficient evidence to support its use in identification and management of wounds with bacterial burden.

Blumenthal and Jeffery (2018) noted that the MolecuLight i:X Imaging Device is a portable, non-invasive, real-time camera used to visualize the bacterial load in a wound.  It uses violet light illumination and a dual band-pass optical filter to capture the fluorescence of endogenous structures in the tissue matrix and harmful bacteria.  The MolecuLight i:X captures images of wounds and highlights potentially detrimental levels of bacteria.  This is an initial evaluation of using the MolecuLight i:X camera in the management of burns to demonstrate the following: the ability of the device to guide clinicians in their management of the burn (i.e., detect, identify, and specify swabbing locations).  Burn wounds were photographed under standard light and violet light illumination to compare presentations of obvious infection signs and symptoms.  Microbiology swab samples were obtained to correlate any bacterial presence to the images.  The fluorescence images were used to guide swabs to where the bacteria were congregating.  A total of 20 patients were imaged; 4 patients did not have bacterial contamination based on their images and swab results; 16 patients showed growth of Staphylococcus aureus, Pseudomonas aeruginosa, or other bacteria; 9 of the patients, by definition, had infections.  These findings were correlated with the typical signs and symptoms of infection, the fluorescence images, and the microbiology results.  The efficacy of the MolecuLight i:X was evident due to the microbiology results correlating to the images.  The authors stated that further research is being done to test the device in terms of being an early intervention tool.  These researchers stated that with these early results and guidance of swab samples, the MolecuLight i:X may be able to detect bacterial load before an infection and subsequent graft failure, thereby shortening lengths of hospital stay and improving overall healing.  These preliminary findings from a small (n = 20) pilot study need to be validated by well-designed studies.

In a pilot study, Pijpe and colleagues (2019) compared the detection of bacteria in burn wounds between an bacterial fluorescence imaging device MolecuLight i:X, (Canada), and standard microbiological swabs.  Wounds were swabbed 3 times on one occasion; once with a standard swab, once with a high-fluorescent area swab, indicating a bacterial load of greater than 104 colony-forming units (CFU)/g, and a finally with a non-fluorescent (nF) area swab.  Proportion agreement of the microbiological results was calculated and the accuracy of the device to detect relevant bacteria was assessed.  A total of 14 patients with 20 wounds participated in the study.  Median post-burn day at sampling time was 21 days.  Of the 20 wounds, 9 had a positive swab result in either of the 3 swabs, and 11 showed a high-fluorescent area.  Overall, positive and negative proportion agreement between standard swab and high-fluorescent swab sample results were 100 %.  Sensitivity, specificity, positive and negative predictive values (PPV and NPV) of presence of high-fluorescence were 78 %, 64 %, 64 %, and 78 %, respectively.  For Pseudomonas aeruginosa detection, these results were 100 %, 70 %, 44 % and 100 %, respectively.  The authors concluded that the diagnostic accuracy of the bacterial fluorescence imaging device to detect relevant bacteria in burn wounds was moderate and the reliability was equal to standard swabbing.  Moreover ,these investigators stated that further research in larger sample sizes and on the relevance of minimal bacterial load and its potential to help with Pseudomonas aeruginosa management is needed.

Hurley and associates (2019) noted that sub-surface bacterial burden can be missed during standard wound examination protocols.  The real-time bacterial fluorescence imaging device, MolecuLight i:X, visualizes the presence of potentially harmful levels of bacteria through endogenous auto-fluorescence, without the need for contrast agents or contact with the patient.  The intended use of the imaging device is to assist with the management of patients with wounds by enabling real-time visualization of potentially harmful bacteria.  In a prospective, single-center, observational study, these researchers examined the accuracy of the wound imaging device at detecting pathogenic bacteria in wounds.  This trial was conducted in an out-patient plastic surgery wound care clinic.  Patients had their wounds photographed under white and auto-fluorescent light with the imaging device.  Auto-fluorescent images were compared with the microbiological swab results.  A total of 33 patients and 43 swabs were included, of which 95.3 % (n = 41) were positive for bacteria growth.  Staphylococcus aureus was the most common bacterial species identified.  The imaging device had a sensitivity of 100 % and specificity of 78 % at identifying pathological bacteria presence in wounds on fluorescent light imaging.  The PPV was 95.4 %; the NPV was 100 %.  It demonstrated a sensitivity and specificity of 100 % at detecting the presence of Pseudomonas spp.  The authors concluded that the imaging device used could be a safe, effective, accurate and easy-to-use auto-fluorescent device to improve the assessment of wounds in the out-patient clinic setting.  In conjunction with best clinical practice, the device can be used to guide clinicians use of antibiotics and specialized dressings.  Moreover, these investigators stated that further research should be directed to its application in other environments, including pre-operative and peri-operative applications as a surgical assessment tool.

The authors stated that this study had several drawbacks.  Blood and highly vascularized tissue are demonstrated as black on the fluorescent light photographs.  Often, these researchers encountered wounds with minimal active bleeding, which rendered the device incompatible.  This was overcome with copious irrigation at the bedside with limited success.  Thus, they considered active bleeding or visible vascularized tissue as a relative contraindication to use of the device.  Dressings containing silver, a potent anti-microbial, also rendered the photograph black.  This was a major drawback when applied in the authors’ out-patient burns clinic, as the majority of these patients had various silver-based dressings applied for their anti-microbial properties.  Darkness was needed for the device to produce accurate and quality auto-fluorescent images.  This was overcome by the use of the imaging device accessory product DarkDrape, which is made of high density polyethylene with an adjustable draw-string to ensure appropriate lighting conditions are met precisely.  The accessory device is single-use only, which is not practical in everyday clinic use.

Rennie and colleagues (2019) stated that the persistent presence of pathogenic bacteria is one of the main obstacles to wound healing.  Detection of wound bacteria relies on sampling methods, which delay confirmation by several days.  However, a novel hand-held fluorescence imaging device has recently enabled real-time detection of bacteria in wounds based on their intrinsic fluorescence characteristics, which differ from those of background tissues.  This device illuminates the wound with violet (405 nm) light, causing tissues and bacteria to produce endogenous, characteristic fluorescence signals that are filtered and displayed on the device screen in real-time.  The resulting images allow for rapid assessment and documentation of the presence, location, and extent of fluorescent bacteria at moderate-to-heavy loads.  This information has been shown to assist in wound assessment and guide patient-specific treatment plans.  However, proper image interpretation is essential to assessing this information.  To properly identify regions of bacterial fluorescence, users must understand: fluorescence signals from tissues (e.g., wound tissues, tendon, bone) and fluids (e.g., blood, pus); fluorescence signals from bacteria (red or cyan); the rationale for varying hues of both tissue and bacterial fluorescence; image artifacts that can occur; and some potentially confounding signals from non-biological materials (e.g., fluorescent cleansing solutions.  The authors concluded that numerous publications on the device have discussed its high sensitivity for bacterial detection, benefits of use during wound assessment, and the various wound treatments that fluorescence images can guide.  Moreover, these researchers stated that ongoing and future studies with the device will evaluate the clear potential for fluorescence-guided wound care to influence wound area reduction rates and wound healing.

Serena and colleagues (2019) noted that clinical evaluation of signs and symptoms (CSS) of infection is imperative to the diagnostic process.  However, patients with heavily colonized and infected wounds are often asymptomatic, leading to poor diagnostic accuracy.  Point-of-care (POC) fluorescence imaging rapidly provides information on the presence and location of bacteria. This clinical trial (#NCT03540004) aimed to evaluate diagnostic accuracy when bacterial fluorescence imaging was used in combination with CSS for identifying wounds with moderate-to-heavy bacterial loads.  Wounds were assessed by study clinicians using NERDS and STONEES CSS criteria to determine the presence or absence of moderate-to-heavy bacterial loads, after which the clinician prescribed and reported a detailed treatment plan.  Only then were fluorescence images of the wound acquired, bacterial fluorescence determined to be present or absent and treatment plan adjusted if necessary.  These researchers examined 17 venous leg ulcers (VLUs) / 2 diabetic foot ulcers (DFUs).  Compared with CSS alone, use of bacterial fluorescence imaging in combination with CSS significantly improved sensitivity (22 % versus 72 %) and accuracy (26 % versus 74 %) for identifying wounds with moderate-to-heavy bacterial loads (greater than or equal to 104 CFU/g, p = 0.002).  Clinicians reported added value of fluorescence images in greater than 90 % of study wounds, including identification of wounds incorrectly diagnosed by CSS (47 % of study wounds) and treatment plan modifications guided by fluorescence (73 % of study wounds).  Modifications included image-guided cleaning, treatment selection, debridement and anti-microbial stewardship.  The authors concluded that findings from this pilot study suggested that when used in combination with CSS, bacterial fluorescence may improve the diagnostic accuracy of identifying patients with wounds containing moderate-to-heavy bacterial loads; and guide more timely and appropriate treatment decisions at the POC.

Farhan and Jeffery (2020) stated that pediatric burn injuries are vulnerable to severe complications, most often infection, making prompt and precise diagnosis of bacterial bioburden vital to preventing detrimental consequences and optimizing patients' outcomes.  Currently, burn wounds are assessed for infection via examining the CSS of infection, which can be confirmed by swab culture analysis.  While the former approach is subjective and experience-dependent, the latter technique is susceptible to missing sub-surface, biofilm-associated colonization, and any peripheral bacterial burden, and also delays confirmation by up to 5 days.  The MolecuLight i:X is a hand-held, non-contact fluorescence imaging device, which can reveal real-time information regarding clinically significant levels of bacteria and their bio-distribution in surface and sub-surface burn wound tissues.  These investigators conducted a single-center, observational study to examine the device’s efficacy in identifying critical bacterial levels in pediatric burn wounds and to test the children's compliance and the overall feasibility of the device integration into the current diagnostic practice.  A total of 10 patients with 16 wounds were recruited and assessed for the presence or absence of CSS of infection and the presence or absence of bacterial fluorescence on images, with swabs taken to confirm findings.  Results demonstrated the device's ability to visualize clinically significant bacterial burden and to localize distribution of pathogens.  All clinicians agreed on the high compliance with the device and high feasibility of incorporating the device into routine wound assessments.  The results of this study may pave the way toward including bacterial fluorescence imaging into the standard diagnostic algorithm for pediatric burn population.

Chew and associates (2020) stated that early diagnosis of wound infections are crucial as they have been shown to increase patient morbidity and mortality.  These researchers examined the use of MolecuLight i:X to identify infections in acute open wounds in hand trauma.  Data were collected from patients who attended the hand trauma unit over a 4-week period before having surgery.  Wounds were inspected for clinical signs of infection and auto-fluorescence images were taken using the MolecuLight i:X device.  Wound swabs were taken and results interpreted according to report by microbiologist.  Auto-fluorescence images were interpreted by a clinician blinded to the microbiology results.  A total of 31 patients were included and data collected from 35 wounds; 3 wounds (8.6 %) showed positive clinical signs of infection, 3 (8.6 %) were positive on auto-fluorescence imaging and 2 (5.7 %) of wound swab samples were positive for significant infection.  Auto-fluorescence imaging correlated with clinical signs and wound swab results for 34 wounds (97.1 %).  In 1 case, the clinical assessment and auto-fluorescence imaging showed positive signs of infection but the wound swabs were negative.  The authors concluded that auto-fluorescence imaging in acute open wounds may be useful to provide real-time confirmation of bacterial infection and thus guide management.

Hill and Woo (2020) noted that the UPPER/LOWER infection checklists look for signs and symptoms of local/superficial infection (UPPER) and deep infection (LOWER) to help clinicians in identifying and distinguishing between these infection levels, facilitating appropriate treatment.  The presence of 3 or more UPPER or LOWER criteria is indicative of infection.  In a prospective, multi-site, observational study, these researchers examined the use of incorporating real-time bacterial fluorescence imaging into the UPPER/LOWER checklists to enhance identification of infection in wounds.  They evaluated 43 chronic wounds (1 wound per patient).  Infection was identified in 27 wounds (62.8 %) according to the UPPER/LOWER checklist criteria; 3 wounds were positive for both UPPER and LOWER infection, 1 wound was positive for LOWER infection only, and 23 wounds were positive for UPPER infection only.  Fluorescence images were taken to detect wounds with high bacterial loads (greater than 104 CFU/g), indicated by the presence of red or cyan fluorescence.  Red or cyan fluorescence from bacteria was observed in 88 % of wounds (n = 38); all wounds positive for UPPER/LOWER were also positive for bacterial fluorescence.  In 18 (41.9 %) of the 43 wounds, fluorescence information added a 3rd check to the UPPER/LOWER threshold, turning a negative diagnosis into a positive diagnosis of infection.  Bacterial load was detected in 22/27 wounds swabbed, 17 of which exhibited heavy growth; in all wounds with detectable bacterial load, fluorescence signal was observed (PPV = 100 %, NPV = 83 %).  Using microbiology as ground truth, inclusion of fluorescence information as an additional item in the checklists increased the sensitivity of the UPPER/LOWER checklist from 82 % to 95 %.  The authors concluded that the findings of this study suggested that the UPPER/LOWER checklist and fluorescence imaging work in a complementary manner to identify wounds with high bacterial burden at the POC.

The authors stated that this study had several drawbacks.  Both clinicians performing the evaluations were experts and familiar with the mnemonics and fluorescence imaging.  Validation of the content of the mnemonics is needed to determine reliability of results among non-experts.  Microbiology culture analysis was not available for all study wounds; therefore, the diagnostic accuracy measures reported in this study described 27 of 43 study wounds.  The fluorescence imaging device could detect bacteria in wounds up to a maximum depth of 1.5 mm and did not provide real-time information on the bacterial species present or non-bacterial components (i.e., fungi) that may be present; wound sampling was needed to obtain this information.  However, the high PPV of fluorescence reported in this trial, and in other studies, indicated that sampling may not always be needed.  The single visit nature of this observational study prevented follow-up visits in most cases to examine if the treatment selections based on checklist classification and fluorescence information were appropriate.  As outcomes data were not available for all patients to validate treatment plan changes, additional studies examining the impact of fluorescence-guided treatment selection are needed.  However, in patients that were followed over multiple visits (e.g., case 6), reduction of UPPER/LOWER symptoms and bacterial fluorescence was observed at follow-up.  Moreover, these researchers stated that due to the nature of the patient population, there was a low proportion of true negative study wounds (i.e., wounds with bacterial loads less than 104 CFU/g); therefore, specificity and NPV results should be interpreted with caution.

Le and co-workers (2021) stated that high bacterial load contributes to chronicity of wounds and is diagnosed based on assessment of clinical signs and symptoms (CSS) of infection, but these characteristics are poor predictors of bacterial burden; POC fluorescence imaging (FL) MolecuLight i:X could improve identification of wounds with high bacterial burden (greater than 104 CFU/g).  FL detects bacteria, whether planktonic or in biofilm, but does not distinguish between the 2.  In a prospective, controlled, multi-center study, these researchers compared diagnostic accuracy of FL to CSS during routine wound assessment.  Post-assessment, clinicians were surveyed to examine the impact of FL on treatment plan.  This trial was carried out by 20 clinicians from 14 outpatient advanced wound care centers in the U.S.  Wounds underwent assessment for CSS followed by FL.  Biopsies were collected to confirm total bacterial load.  A total of 350 patients completed the study (138 diabetic foot ulcers, 106 venous leg ulcers, 60 surgical sites, 22 pressure ulcers, and 24 others); 287/350 wounds (82 %) had bacterial loads greater than 104 CFU/g, and CSS missed detection of 85 % of these wounds.  FL significantly increased detection of bacteria (greater than 104 CFU/g) by 4-fold, and this was consistent across wound types (p < 0.001).  Specificity of CSS+FL remained comparably high to CSS (p = 1.0).  FL information modified treatment plans (69 % of wounds), influenced wound bed preparation (85 %), and improved overall patient care (90 %) as reported by study clinicians.  The authors concluded that this novel non-contact, hand-held FL device provided immediate, objective information on presence, location, and load of bacteria at POC; and the use of FL facilitated adherence to clinical guidelines recommending prompt detection and removal of bacterial burden to reduce wound infection and facilitate healing.

The authors stated that this study had several drawbacks.  First, due to the imprecision of soft tissue biopsy trimming, the biopsies were cut to a greater depth than the 1.5-mm excitation limit of the imaging device; therefore, it was possible that the biopsy may have detected slightly more anaerobic bacteria than the device was able to.  Second, the conditions of culture analysis were unfavorable for fastidious bacteria and may have resulted in under-reporting the diversity of bacteria species present in the wound.  This study focused primarily on high bacterial loads as a contributor to delayed wound healing; however, additional systemic factors that were not reported in this study, including vascular insufficiency and protease activity, must also be considered.  Clinicians had limited experience using FL in a clinical context before the study, which may have contributed to lower sensitivity to detect bacteria at loads of greater than 104 CFU/g than previously observed.  In prior FL studies, sensitivity estimates ranging from 72 % to 100 % were reported, likely due to more clinician experience using the device.  As with other diagnostic imaging modalities, these investigators anticipated that the performance measures reported should be improved with increased experience.  This single time-point study meant that effectiveness of changes in treatment plan based on FL could not be measured.  Longitudinal RCTs examining wound healing may further elucidate the impact of POC diagnostic imaging of bacteria.  Evidence from small longitudinal observational studies showed accelerated wound area reduction with use of FL.  Due to the limited (1.5 mm) depth of excitation as well as inability to detect non-porphyrin-producing bacteria, including species from the Streptococcus, Enterococcus, and Finegoldia generas (which account for an estimated 12 % of the most prevalent wound pathogens and rarely occur mono-microbially), it is recommended that FL be used in combination with CSS.

Raizman and associates (2021) noted that pseudomonas aeruginosa (PA) is a common bacterial pathogen in chronic wounds known for its propensity to form biofilms and evade conventional treatment methods.  Early detection of PA in wounds is critical to the mitigation of more severe wound outcomes.  Point-of-care bacterial fluorescence imaging has been used to illuminate wounds with safe, violet light, triggering the production of cyan fluorescence from PA.  A prospective, single-blind clinical study was carried out to determine the PPV of cyan fluorescence for the detection of PA in wounds.  Bacterial fluorescence using the MolecuLight i:X imaging device revealed cyan fluorescence signal in 28 chronic wounds, including venous leg ulcers, surgical wounds, diabetic foot ulcers and other wound types.  To correlate the cyan signal to the presence of PA, wound regions positive for cyan fluorescence were sampled via curettage.  A semi-quantitative culture analysis of curettage samples confirmed the presence of PA in 26/28 wounds, resulting in a PPV of 92.9 %.  The bacterial load of PA from cyan-positive regions ranged from light to heavy.  Less than 20 % of wounds that were positive for PA exhibited the classic symptoms of PA infection.  The authors concluded that the findings of this study suggested that cyan detected on fluorescence images can be used to reliably predict bacteria, specifically PA at the POC.

The authors stated that this study had several drawbacks.  This trial was designed to test the PPV of cyan on fluorescence images; therefore, it did not provide information on NPV, sensitivity or specificity of the images.  These have been evaluated by other studies, which demonstrated high sensitivity and specificity of the images for detecting high bacterial load.  Future work that examines the presence of PA from wound regions positive or negative for cyan fluorescence may help to clarify the specificity of cyan for PA.  Furthermore, the semi-quantitative culture based microbiological confirmation used has inherent limitations, as this method could under-estimate bacterial loads and each semi-quantitative category was associated with a wide range of CFU/g counts.  For example, light growth has been shown to range from 103 to 106 CFU/g in wound samples.  Under-estimating bacterial loads may have resulted in a slightly lower reported PPV in this study, as scant growth of PA was detected in 1 clearly cyan-positive wound and was considered a false positive in this analysis.  Additional studies utilizing gold standard quantitative or molecular culture-based methods to analyze wound biopsies may help to clarify the bacterial loads detected from regions of cyan fluorescence.  The fluorescence imaging procedure also has inherent limitations, namely the need for darkness during imaging and a limited depth of excitation (approximately 1.5 mm) for detection of subsurface bacteria; however, as PA tends to be a surface or immediately sub-surface pathogen, this posed less of a limitation for the detection of PA than it may be for other pathogens.

Lopez and colleagues (2021) stated that wound biofilms must be identified to target disruption and bacterial eradication but are challenging to detect with standard clinical assessment.  These researchers examined if bacterial fluorescence imaging could detect porphyrin-producing bacteria within a biofilm using well-established in-vivo models.  Mouse wounds were inoculated on Day 0 with planktonic bacteria (n = 39, porphyrin-producing and non-porphyrin-producing species, 107 colony forming units (CFU)/wound) or with polymicrobial biofilms (n = 16, 3 biofilms per mouse, each with 1:1:1 parts Staphylococcus aureus/Escherichia coli/Enterobacter cloacae, 107 CFU/biofilm) that were grown in-vitro.  Mouse wounds inoculated with biofilm underwent fluorescence imaging up to Day 4 or 5.  Wounds were then excised and sent for microbiological analysis.  Bacteria-matrix interaction was examined with scanning electron microscopy (SEM) and histopathology.  A total of 48 hours after inoculation with planktonic bacteria or biofilm, red fluorescence was readily detected in wounds; red fluorescence intensified up to Day 4.  Red fluorescence from biofilms persisted in excised wound tissue post-wash.  SEM and histopathology confirmed bacteria-matrix interaction.  The authors concluded that this pre-clinical study was the first to demonstrate the fluorescence detection of bacterial biofilm in-vivo using a POC wound imaging device.  These findings have implications for clinicians targeting biofilm and may facilitate improved visualization and removal of biofilms.

Table: CPT Codes / HCPCS Codes / ICD-10 Codes
Code Code Description

Information in the [brackets] below has been added for clarification purposes.   Codes requiring a 7th character are represented by "+":

CPT codes not covered for indications listed in the CPB:

97610 Low frequency, non-contact, non-thermal ultrasound, including topical application(s), when performed, wound assessment, and instruction(s) for ongoing care, per day

HCPCS codes not covered for indications listed in the CPB:

A6000 Non-contact wound warming wound cover for use with the non-contact wound warming device and warming card
E0231 Non-contact wound warming device (temperature control unit, AC adaptor and power cord) for use with warming card and wound cover
E0232 Warming card for use with the non-contact wound warming device and non-contact wound warming cover

ICD-10 codes not covered for indications listed in the CPB (not all-inclusive):

A00.0 - A09 Intestinal infectious diseases
A20.0 - A28.9 Certain zoonotic bacterial diseases
A30.0 - A49.9 Other bacterial diseases
E10.40 - E10.49
E11.40 - E11.49
Diabetes with neurological complications
E10.51 - E10.59
E11.51 - E11.59
Diabetes with peripheral circulatory complications
E10.610 - E10.69
E11.610 - E11.69
Diabetes with other specified complications
I70.231 - I70.25 Atherosclerosis of the lower extremities with ulceration
I70.261 - I70.269 Atherosclerosis of the lower extremities with gangrene
I73.9 Peripheral vascular disease, unspecified
I74.3 Embolism and thrombosis of arteries of the lower extremities
I83.001 - I83.029 Varicose veins of lower extremities with ulcer
L02.01, L02.11, L02.211 - L02.219
L02.31, L02.411 - L02.419
L02.511- L02.519, L02.611 - L02.619
L02.811 - L02.818, L02.91
L03.011 - L03.91
Other cellulitis and abscess
L05.01 - L05.02 Pilonidal cyst or sinus with abscess
L05.91 - L05.92 Pilonidal cyst or sinus without mention of abscess
L89.000 - L89.95 Pressure ulcer
Numerous options Open wound
T81.31x+ - T81.32x+ Disruption of external or internal operation (surgical) wound, not elsewhere classified
T81.40xA - T81.49xS Infection following a procedure
T81.89x+ Other complications of procedures, not elsewhere classified [non-healing surgical wound]

The above policy is based on the following references:

  1. Alvarez O, Patel M, Rogers R, Booker J. Effect of non-contact normothermic wound therapy on the healing of diabetic neuropathic foot ulcers. J Tissue Viability. 2006;16(1):8-11.
  2. Alvarez OM, Rogers RS, Booker JG, Patel M. Effect of noncontact normothermic wound therapy on the healing of neuropathic (diabetic) foot ulcers: An interim analysis of 20 patients. J Foot Ankle Surg. 2003;42(1):30-35.
  3. Blumenthal E, Jeffery SLA. The use of the MolecuLight i:X in managing burns: A pilot study. J Burn Care Res. 2018;39(1):154-161.
  4. Cai Y, Wang J, Liu X, et al. A review of the combination therapy of low frequency ultrasound with antibiotics. Biomed Res Int. 2017;2017:2317846.
  5. Canadian Coordinating Office for Health Technology Assessment (CCOHTA). Warm-up therapy. Emerging Device List. No. 8. Ottawa, ON: CCOHTA; July 2001.
  6. Center for Medicare and Medicaid Services (CMS). Noncontact Normothermic Wound Therapy (NNWT). Coverage Issues Manual Section 60-25. Baltimore, MD: CMS; 2002.
  7. Center for Medicare & Medicaid Services (CMS). Decision Memo for Warm-Up Wound Therapy® a/k/a Noncontact Normothermic Wound Therapy (#CAG-00114N). Baltimore, MD: CMS; January 14, 2002 (amended February 14, 2002).
  8. Chang YR, Perry J, Cross K. Low-frequency ultrasound debridement in chronic wound healing: A systematic review of current evidence. Plast Surg (Oakv). 2017;25(1):21-26.
  9. Chew BJW, Griffin M, Butler PE, Mosahebi A. The use of MolecuLight i:X device in acute hand trauma. J Plast Reconstr Aesthet Surg. 2020;73(7):1357-1404.
  10. Conner-Kerr T, Alston G, Stovall A, et al. The effects of low-frequency ultrasound (35 kHz) on methicillin-resistant staphylococcus aureus (MRSA) in vitro. Ostomy Wound Manage. 2010;56(5):32-42.
  11. Cullum NA, Al-Kurdi D, Bell-Syer SE. Therapeutic ultrasound for venous leg ulcers. Cochrane Database Syst Rev. 2010;(6):CD001180.
  12. Farhan N, Jeffery S. Utility of MolecuLight i:X for managing bacterial burden in pediatric burns. J Burn Care Res. 2020;41(2):328-338.
  13. Hill R, Woo K. A prospective multisite observational study incorporating bacterial fluorescence information into the upper/lower wound infection checklists. Wounds. 2020;32(11):299-308.
  14. Honaker JS, Forston MR, Davis EA, et al. The effect of adjunctive noncontact low frequency ultrasound on deep tissue pressure injury. Wound Repair Regen. 2016;24(6):1081-1088.
  15. Hurley CM, McClusky P, Sugrue RM, et al. Efficacy of a bacterial fluorescence imaging device in an outpatient wound care clinic: A pilot study. J Wound Care. 2019;28(7):438-443.
  16. Karr JC. External thermoregulation of wounds associated with lower-extremity osteomyelitis. A pilot study. J Am Podiatr Med Assoc. 2003;93(1):18-22.
  17. Kloth LC, Berman JE, Nett M, et al. A randomized controlled clinical trial to evaluate the effects of noncontact normothermic wound therapy on chronic full-thickness pressure ulcers. Adv Skin Wound Care. 2002;15(6):270-276.
  18. Le L, Baer M, Briggs P, et al. Diagnostic accuracy of point-of-care fluorescence imaging for the detection of bacterial burden in wounds: Results from the 350-patient fluorescence imaging assessment and guidance trial. Adv Wound Care (New Rochelle). 2021;10(3):123-136.
  19. Lopez AJ, Jones LM, Reynolds L, et al. Detection of bacterial fluorescence from in vivo wound biofilms using a point-of-care fluorescence imaging device. Int Wound J. 2021 Feb 9 [Online ahead of print].
  20. Macario A, Dexter F. Is noncontact normothermic wound therapy cost effective for the treatment of stages 3 and 4 pressure ulcer? Wounds. 2002;14(3):93-106.
  21. Madhok BM, Vowden K, Vowden P. New techniques for wound debridement. Int Wound J. 2013;10(3):247-251.
  22. McCulloch J, Knight CA. Noncontact normothermic wound therapy and offloading in the treatment of neuropathic foot ulcers in patients with diabetes. Ostomy Wound Manage. 2002;48(3):38-44.
  23. Ontario Ministry of Health and Long-Term Care. Management of chronic pressure ulcers. An evidence-based analysis. Ontario Health Technology Assessment Series. Toronto, ON: Ontario Ministry of Health and Long-Term Care; 2009;9(3).
  24. Pijpe A, Ozdemir Y, Sinnige JC, et al. Detection of bacteria in burn wounds with a novel handheld autofluorescence wound imaging device: A pilot study. J Wound Care. 2019;28(8):548-554.
  25. Raizman R, Little W, Smith AC. Rapid diagnosis of pseudomonas aeruginosa in wounds with point-of-care fluorescence imaging. Diagnostics (Basel). 2021;11(2):280.
  26. Rennie MY, Dunham D, Lindvere-Teene L, et al. Understanding real-time fluorescence signals from bacteria and wound tissues observed with the MolecuLight i:XTM. Diagnostics (Basel). 2019;9(1):22.
  27. Robinson C, Santilli SM. Warm-up active wound therapy: A novel approach to the management of chronic venous stasis ulcers. J Vasc Nurs. 1998;16(2):38-42.
  28. Santilli SM, Valusek PA, Robinson C. Use of a noncontact radiant heat bandage for the treatment of chronic venous stasis ulcers. Adv Wound Care. 1999;12(2):89-93.
  29. Serena T, Lee SK, Lam K, et al. The impact of noncontact, nonthermal, low-frequency ultrasound on bacterial counts in experimental and chronic wounds. Ostomy Wound Manage. 2009;55(1):22-30.
  30. Serena TE, Harrell K, Serena L, Yaakov RA. Real-time bacterial fluorescence imaging accurately identifies wounds with moderate-to-heavy bacterial burden. J Wound Care. 2019;28(6):346-357.
  31. Thomas DR, Diebold MR, Eggemeyer LM. A controlled, randomized, comparative study of a radiant heat bandage on the healing of stage 3-4 pressure ulcers: A pilot study. J Am Med Dir Assoc. 2005;6(1):46-49.
  32. Voigt J, Wendelken M, Driver V, Alvarez OM. Low-frequency ultrasound (20-40 kHz) as an adjunctive therapy for chronic wound healing: A systematic review of the literature and meta-analysis of eight randomized controlled trials. Int J Low Extrem Wounds. 2011;10(4):190-199.
  33. Wagner-Cox P, Duhame HM, Jamison CR, et al. Use of noncontact low-frequency ultrasound in deep tissue pressure injury: A retrospective analysis. J Wound Ostomy Continence Nurs. 2017;44(4):336-342.
  34. Whitney JD, Salvadalena G, Higa L, et al. Treatment of pressure ulcers with noncontact normothermic wound therapy: Healing and warming effects. J Wound Ostomy Continence Nurs. 2001;28(5):244-252.
  35. Zhu Y, Guan L, Mu Y. Combined low-frequency ultrasound and urokinase-containing microbubbles in treatment of femoral artery thrombosis in a rabbit model. PLoS One. 2016;11(12):e0168909.