Efficacy of Pulsed Low-Intensity Ultrasound in Wound Healing: A Single-Case Design
The presence of pressure ulcers in elderly patients is a widespread problem that results in considerable healthcare costs.1-3 For example, the incidence of Stage II and worse pressure ulcers has been reported to be more than 30% in nursing home residents.1 Therapeutic ultrasound (US) is one of several treatment methods used to enhance healing of pressure ulcers, including sharp debridement, wound cleansing, wound dressings, and electrical stimulation.3
Under in vivo and in vitro conditions in animal studies, ultrasound has been shown to have beneficial effects on factors associated with tissue healing. These effects include promotion of histamine release4 and mast cell degranulation,4,5 angiogenesis,6 increases in intracellular calcium,7 increases in collagen deposition and wound-breaking (or tensile) strength,8-10 and reduction of wound size.5
The limited clinical research that has been conducted on human subjects seems to be split as to the efficacy of US on both venous ulcers and pressure ulcers in the elderly,11-16 and data regarding full-thickness pressure ulcers in the elderly are particularly scarce.
The purpose of this study was to assess the efficacy of pulsed low-intensity ultrasound (LIUS) for healing a Stage III pressure ulcer in a geriatric patient. A Stage III ulcer was chosen for study because it is a full-thickness skin loss,3 which would enable the authors to note healing by measuring the surface area as it closed. A Stage II ulcer was not chosen because the authors believed the wound edges would be less discernible for reliable wound tracing. Pressure ulcers at Stage IV were not chosen because the damage in each wound extends to the muscle, bone, or supporting structures3; thus, requiring additional measurements such as wound volume.
Studies have reported mixed results on the efficacy of pulsed LIUS for wound healing in humans. Comparisons are difficult because of differences in US parameters; standard care characteristics; other treatment protocol variables; differences in type, size, and location of ulcers; and ambiguous descriptions of ulcer status or stage.
Dyson and Sucklin16 reported on two different trials involving patients with chronic varicose ulcers of the lower limb (reportedly chosen because of their superficial nature) that had never been treated with US and had not responded favorably to other types of treatment. All patients received the same form of standard care. One trial separated the patients into groups receiving pulsed US or placebo US; for patients with bilateral ulcers, one received pulsed US while the other received placebo US. In the other trial, an ulcer was treated with placebo US for the first month and pulsed US for the second month. Ultrasound characteristics were: 3 MHz, 1.0 W/cm2, pulsed 20%, 7.07 cm2 ERA, 5 minutes for ulcers of 2.5cm2 or less with 1 minute added for each additional 0.5 cm2 up to 10 minutes, three times a week. In the first trial, after the first 4 weeks, the pulsed US group had a 25% greater reduction in wound surface area than the placebo US group (statistically significant; P <0.05). In the second trial, a significantly greater reduction occurred in wound surface area after the period of pulsed US (34% average) compared to the period of placebo US (14% average); P <0.05.
Callam et al14 studied patients with chronic varicose ulcers of the lower extremity, most of which were superficial ulcers. The treatment group received pulsed LIUS (1 MHz frequency, 0.5 W/cm2 intensity, pulsed 20% duty cycle, 1 min/effective radiating area [ERA], once a week) in conjunction with standard care. The control group received the standard care alone. The authors reported a 20% greater reduction in wound surface area for the treatment group compared to the control group after 4 weeks (P <0.05) and the difference between groups was maintained for the entire 12 weeks of the study. In addition, approximately 20% more of the patients in the treatment group than in the control group healed within 12 weeks. However, Callam et al did not use a placebo US group and did not report the ERA of the face of the US transducer.
Eriksson et al13 compared two groups of patients receiving standard care for venous ulcers, with one group receiving additional treatment with US (1 MHz, 1.0 W/cm2, 10 minutes, twice a week for 8 weeks) and the other group receiving additional treatment with placebo US. The authors did not report whether the US was pulsed or continuous. They used a 2.8 cm diameter sound head for the wounds they classified as superficial (<1 cm in depth), and a 1.2 cm diameter sound head for the deep wounds; less than 20% of the wounds were classified as deep. They found approximately 10% more healed ulcers in the US group than in the placebo group throughout the course of the study, but this difference was not found to be statistically significant (P >0.05). In addition, they found an approximately 7% greater reduction in wound surface area for the US group compared to the placebo group over the course of the study, but this also was not statistically significant. The US intensity used by Eriksson et al possibly was too high to heal the wounds faster.
Lundeberg et al15 studied patients with venous leg ulcers, divided into pulsed LIUS and placebo US groups. Most of the ulcers were classified as superficial (<1 cm in depth). The US characteristics were: 1 MHz, 0.5 W/cm2, pulsed 10%, 10 minutes (1 min/ERA), 3 days/week for the first 4 weeks, twice a week for the next 4 weeks, and once a week for the last 4 weeks. Beginning with the fourth week, the pulsed LIUS group averaged approximately 8% more healed ulcers and a 5% greater reduction in wound surface area than the placebo group, but no significant difference (P >0.05) was noted between groups.
Two double-blind studies investigated the use of US for healing pressure ulcers in elderly patients, comparing US groups to placebo US groups. McDiarmid et al12 used US with the following treatment parameters: 3 MHz, 0.8 W/cm2, pulsed 20%, ERA 5.2 cm2; treatments of 5 minutes for ulcers of ≤3cm2, with 1 minute added for each additional 0.5 cm2 up to 10 minutes, three times a week. However, only superficial pressure ulcers (not extending beyond the dermis) were treated. Patients appeared to have been treated for 70 days, but nine of the 27 total patients (excluding losses to death or discharge) did not heal completely (three treated with US, six with placebo US). The authors reported that of the wounds that healed completely in each group, the median number of days until complete healing was 32 for the US group and 36 for the placebo group; the difference was not statistically significant (P = 0.8). They compared uninfected wounds to infected wounds for wound surface area (only for the initial 4 weeks of the study) and reported no significant difference between groups for uninfected wounds (P <0.9), but the US group had significantly smaller infected wounds (P <0.2) than the placebo group. The US group's infected wounds appeared to have decreased to less than 70% of their initial size on the average, while the placebo group's infected wounds increased slightly by the end of 4 weeks. One methodological issue may have affected the results of this study: Although the average age of the patients was reported to be 80 years, the sample included patients who were 18 years of age and no report of the distribution of age groups between the groups was mentioned.
A double-blind study by ter Riet et al11 included elderly patients with pressure ulcers of grades II, III, and IV (only 16 of 88 patients who began the study had a grade IV). Pulsed LIUS and placebo US groups were treated once a day, 5 days a week, for 12 weeks, over the wound bed and the wound margins. The US parameters were: 3 MHz, 0.5 W/cm2, pulsed 20%, larger wounds treated longer than smaller wounds. High- and low-dose ascorbic acid was added to each of the US interventions, along with standard care. The authors reported no significant differences between the pulsed LIUS and placebo US groups on all outcome variables, except that small wounds (≤1cm2) healed faster with pulsed LIUS (P = 0.05). The large variability in wound and patient characteristics may have influenced their findings, in addition to other sources of measurement error discussed by the authors. Additionally, some research sites in this study used chlorhexidine, which is cytotoxic to fibroblasts, to clean the wounds before US treatment. This may have adversely affected healing and obscured the possible beneficial effects of ultrasound.
Most of the studies previously discussed found some beneficial effect of ultrasound compared to a control or a placebo group, but they reported that their sample sizes provided insufficient power to detect a statistically significant difference between groups at the effect sizes that were manifested.12,13,15
Approval for this study was obtained from the Institutional Review Board of Samuel Merritt College, Oakland, Calif., and from Hillside Care Center, San Rafael, Calif. The patient was a 75-year-old woman in a skilled nursing facility with a Stage III pressure ulcer over the coccyx. Informed consent was obtained for her participation in this study. Her medical history included: right cerebrovascular accident, pulmonary embolism, deep vein thrombosis, depressive disorder, diabetes mellitus, hysterectomy, inferior vena cava filter, low back pain, and three spinal fusions. She was had fecal incontinence. Approximately 7 weeks before the study period, she was diagnosed with a Stage I pressure ulcer over the coccyx. The ulcer progressed to Stage III 1 to 2 weeks before the start of the study, at which time she was first seen by a physical therapist for this problem.
Just before the study period, the wound was described in the patient's medical record as being approximately 60% red, 25% yellow, and 15% dark brown; with minimal to moderate drainage that was slightly creamy and green tinged; and with some necrotic tissue and excoriation of the surrounding tissue. Wound measurement documentation during the week before the study period indicated little change in the status of this Stage III ulcer. The patient did not receive treatment for the ulcer using any other physical agent (eg, electrical stimulation, diathermy, therapeutic ultraviolet radiation) before or during the course of this investigation. The patient also was treated for a smaller Stage III pressure ulcer on her heel during the same time period in a similar manner (but without placebo US), but this was not a part of the study.
Pulsed LIUS was compared to placebo US in a baseline-AB design. In this design, no intervention or change in intervention may occur during an initial baseline period; different interventions are applied or the existing intervention regimen is changed in different ways in each of the following periods A and B. In this study, after the baseline period of 1 week, each US treatment was administered 5 days a week for two consecutive weeks. Throughout the baseline and US treatment periods, the patient additionally received the standard wound care treatment program at the facility (including wound dressings, low-air-loss bed, pressure relief, turning and positioning, nutritional supplementation, pulsed irrigation and sharp debridement, and topical systemic antibiotics as needed). Two clinicians participated in the treatments and data collection. Wound surface area measurements were taken at the end of each treatment by tracing the wound onto a transparency (MediRule II Transparency Measuring device by Briggs, Des Moines, Iowa). A software technician scanned all of the transparencies into a computer software program (AutoCAD, Release 14, by Autodesk Inc., Cupertino, Calif.) to determine the wound surface area. These measurements were used to indicate the rate of wound closure (ie, healing).
Instrumentation included two "25 Multi-Hz" US devices manufactured by Rich-Mar Corporation (Inola, Okla.), one of which was engineered to be a placebo device so it did not deliver US, but instead displayed treatment variables as if it were delivering US. The US characteristics used were: 3-MHz frequency, 0.3-W/cm2 intensity (Spatial Average Temporal Peak - SATP), pulsed at a 20% duty cycle, 5-cm2 ERA of the transducer surface, beam nonuniformity ratio (BNR) 5.5:1 or less, 4-minute treatment duration.
The patient was positioned on her side. An aqueous gel was used as a coupling medium and was applied to the wound perimeter. The sound head was moved in small overlapping circular movements at a rate of approximately 10 cm/second.
After taking wound-tracing measurements for 1 week (baseline period), pulsed LIUS and placebo US treatments were given for two consecutive weeks each. Treatments were given and wound-tracing measurements were taken almost every weekday during the study.
This was a double-blind study. Neither the clinicians performing the treatments and measurements nor the patient knew when the placebo US or the pulsed LIUS was being delivered. The software technician who calculated the wound surface area from the tracings also did not know the relationship among the tracings and the three periods of the study.
Each transparency was coded (based on the output of a random number generator) and issued in separate envelopes for each specific day. The code allowed the researchers to analyze all of the tracings at the end of the 5-week study period without knowing the order of the tracings or the measurement period (and, therefore, which treatment was delivered during the day of measurement). After all the tracings were analyzed for the area measurements, the code was referenced to a master list to determine the order of the tracings, enabling comparison of the treatment/control data.
The first 2-week treatment period was called Period A and the second was called Period B, with the US devices used in the respective periods and arbitrarily labeled A or B.
Intrarater and interrater reliability for the two clinicians, and intrarater reliability for the software technician, were assessed with ANOVAs. Each of the wound surface area measurements was plotted on a graph (see Figure 1) for analysis. Using the split-middle technique,17 a celeration line was calculated and drawn to identify the linear trend and slope for the data points obtained during a given period.
The slope (with its direction/trend) represents the rate of wound closure during that period. Comparisons were made among the baseline, pulsed LIUS, and placebo US periods for trend and slope. A binomial test17 was used to determine if the celeration line (ie, its slope) of a period was statistically different from that of the subsequent adjacent period; this test requires the extension of the earlier period's celeration line into the subsequent adjacent period (see Figure 2).
The authors' research hypothesis was the null hypothesis. The alpha level for this two-tailed probability was set at 0.05.
Intrarater reliability of each of the clinicians was 0.999. Interrater reliability between the clinicians was 0.997. Intrarater reliability for the software technician was 0.997. When data collection was concluded, the researchers identified that the device delivering pulsed LIUS was used during Period A and the device delivering placebo US was used during Period B.
The wound surface area measurements are plotted over the time course of the study (by day of measurement, see Figure 1). Within each period of the study (ie, baseline, Period A, Period B), the celeration line is identified. The limited data points during the baseline period did not permit use of the split-middle technique, so a best-fit line was drawn connecting the data points to create the celeration line. The celeration lines for Periods A and B were created according to the split-middle technique. The celeration line for Period A was adjusted to meet the criterion of equal data points above and below the line while maintaining the same slope. No such adjustment of the celeration line was necessary for Period B.
All three celeration lines exhibited downward trends, indicating a reduction in wound surface area during each period (ie, healing progressed).
The slopes of the celeration lines (see Table 1) represent the wound surface area closure rates, used here to indicate rates of healing, during the respective periods of the study. Therefore, the wound closure rate was 33.1% slower for Period A compared to the baseline period and 62.9% slower for Period B compared to Period A. In addition, wound closure rate was 75.2% slower for period B compared to the baseline period.
The binomial test showed that the celeration lines of the three periods were significantly different from each other (P = 0.002; see Table 2, which is based on Figure 2). The rate of wound closure was fastest during the first week when the patient received the standard of care alone. The rate of closure became progressively slower during the next 2 weeks -- the period during which the patient received pulsed LIUS treatments in addition to the standard of care. The rate slowed even further during the final 2 weeks when the patient received the placebo US in addition to the standard of care.
The use of pulsed LIUS did not improve the healing rate of this patient's pressure ulcer. The rate of healing over the course of the entire study appeared to be relatively rapid during the baseline period and slowed progressively by comparison over the successive periods of treatment during which pulsed LIUS and placebo US, respectively, were added to the standard care. One might conclude from this observation that standard care alone was the most effective for healing, that the addition of pulsed LIUS was less effective than the standard care alone, and that the addition of placebo US was less effective than the addition of pulsed LIUS. One might also conclude that both US applications slowed healing in this patient. However, the authors believe that the decrease in the slope (or rate of healing or closure) over the course of the study is what would be expected with normal wound maturation. The patient in this study appeared to have an uncomplicated course of healing that was not influenced by pulsed LIUS or placebo US. One limitation of the baseline-AB design is that one cannot determine if the decrease in rate of wound closure from one period to the next would have been different had the order of treatment application been different or if no treatment at all had been provided beyond standard care.
Methodological factors influencing results. A few methodological factors involving compliance with the intended protocol may have influenced the results. The protocol designed for treating the patient included an intensity of 0.5 W/cm2, based on prior studies5,8,14,19 that had been successful with this level. The authors' original protocol also called for the treatment duration to be determined by using the following formula: (treatment area/ERA) x 3 minutes11; therefore, as the wound healed, the treatment duration was to have decreased proportionally. However, the treating clinicians substituted an intensity of 0.3 W/cm2 and an unchanging treatment duration of 4 minutes. The choice of ultrasound application only to the periphery of the wound rather than over the wound was according to the clinicians' desire, based on their facility's existing protocol. In addition, measurements were not taken for four of the treatments, two during the baseline period and one each during periods A and B. This might have affected the slope of the wound surface area (ie, healing rate) particularly during the baseline period, which could only have had a total of five data points. In the authors' original plan, a baseline-ABAB design (2 weeks each period) had been considered, but the clinicians rejected the longer period and the fast healing rate of the patient's wound precluded a longer study period. With a 2-week baseline period, the authors would have eliminated this patient for the study had her healing rate continued unchanged.
Other factors that may have affected the results involve the parts of the treatment program that did not include US. Evidently, poor compliance was noted with the turning and positioning plan immediately before the study's baseline period; this was rectified during the baseline period and may have contributed to the marked improvement in healing during this period. In addition, other modifications including the dressings used during that period also may have augmented the healing response during the baseline compared to the initial measurement of the study.
Rate of healing. In one study, Stage III pressure ulcers were reorted to heal at a rate of 10 mm2/day using a low-air-loss mattress and standard care alone.18 In this study, the patient's Stage III pressure ulcer healed markedly faster during the periods of standard care alone or standard care combined with pulsed LIUS; the average rate over the course of the entire study was also faster at 21.6 mm2/day. In addition, the overall time until complete wound closure in this study was 35 days; only 20% of patients with Stage III pressure ulcers were reported by Ferrell18 to completely heal in as little as 90 days using low-air-loss mattresses and standard care alone.
In general, the overall pattern of healing rates in studies comparing US and non-US groups showed faster healing initially that slowed as time progressed.13-15 This behavior occurred in the authors' study, but was compressed into a shorter time frame because of the overall faster healing rates. The main similarity regarding healing rates among the studies of Eriksson et al,13 Callam et al,14 and Lundeberg et al15 is that healing with US was fastest during the initial period (ie, 2 to 4 weeks) of the study, but not appreciably faster than standard care or placebo US beyond this period.
The Stage III pressure ulcer of the patient in the authors' study fully healed in 5 weeks, while those in the other studies were at 40% or greater of their initial size, on the average, at similar points or longer in treatment. Therefore, the healing rates for the patient in the authors' study were markedly faster than those found in the studies previously described,12-15 despite the fact that almost all of the wounds in these studies were classified as superficial. In addition, ter Riet et al11 reported mean wound surface area reduction rates of approximately 30 mm2 or less per week. During literature review, the rates for trial two of the Dyson and Suckling study16 were calculated to be 40 mm2 or less per week (5.7 mm2/day); these rates are also much slower than those for the patient in the current study. Interesting to note: The pressure ulcer on the heel of the patient in the current study took approximately the same time to heal as the one over the coccyx and the pattern (rate) of healing seemed similar.
Generalizing the results of one single-case design experiment would be inappropriate. It is interesting to note how the rate of healing (for individual periods and overall) in the authors' study under typical clinical conditions compares to what has been reported in the literature under similar conditions. Although several possible reasons for differences can be offered, the most important message may be to be aware of the rate of healing while treating a patient to determine if alternative interventions or altering the treatment variables may be appropriate or necessary. Although the rate of healing in this study was comparatively fast, it does not mean that no improvement is possible.
Reducing the surface area of a Stage III pressure ulcer in a geriatric patient was faster with a standard wound care program alone during the initial week of treatment than when pulsed LIUS was added to the program during the next 2 weeks. In addition, the reduction was faster during the weeks of pulsed LIUS than the subsequent 2 weeks when placebo US was added to the standard care. Therefore, what effect, if any, the pulsed LIUS had on healing is unclear. The standard care was probably sufficient for healing the pressure ulcer in this patient. The slowing of the rate of healing after pulsed LIUS was added to the standard care was probably due to normal wound maturation effects in an uncomplicated course of healing rather than to a deleterious effect of the pulsed LIUS. Perhaps a patient with a larger wound or a more compromised health status would have provided a better test of the effects of pulsed LIUS.
The rates of wound surface area reduction in this patient were much faster than those for other studies assessing the effects of pulsed US, placebo US, or standard care regimens on healing of skin ulcers. Noting the rate of healing during the first week after initiation of a standard wound care program before deciding if other treatment interventions may be necessary may be clinically useful. Further research is necessary to fully evaluate the usefulness of pulsed LIUS for healing Stage III pressure ulcers in geriatric patients.
The authors are grateful to Donna Silsbee, PT, and Jay Corniea, PTA, for participating in treatment of the patient and in data collection; to Hillside Care Center (San Rafael, Calif.) for permitting the study to be conducted; to Rich-Mar Corporation for loaning the US devices; to Craig Mortonson for analyzing the wound tracings; to Dr. Andrea Taylor for her review and consultation; to Dr. Fred Feuchter for his assistance with the software used for creating the figures; and to Dr. Beth McManis for her assistance with the reliability assessments. This study was presented at the Annual Conference and Exposition of the American Physical Therapy Association, Washington, DC, June 6, 1999.
1. Brandeis GH, Morris JN, Nash DJ, Lipsitz LA. The epidemiology and natural history of pressure ulcers in elderly nursing home residents. JAMA. 1990;264:2905-2909 .
2. Powell JW. Increasing acuity of nursing home patients and the prevalence of pressure ulcers: a ten year comparison. Decubitus. 1989;2:56-58 .
3. Bergstrom N, Bennett MA, Carlson CE, et al. Clinical Practice Guideline Number 15: Treatment of Pressure Ulcers. Rockville, Md: US Department of Health and Human Services. Public Health Service. Agency for Health Care Policy and Research; 1994. AHCPR Publication 95-0652.
4. Fyfe MC, Chahl LA. Mast cell degranulation and increased vascular permeability induced by 'therapeutic' ultrasound in the rat ankle joint. British Journal of Experimental Pathology. 1984;65:671-676.
5. Byl NN, McKenzie AL, West JM, Whitney JD, Hunt TK, Scheuenstuhl HA. Low-dose ultrasound effects on wound healing: a controlled study with Yucatan pigs. Arch Phys Med Rehabil. 1992;73:656-664.
6. Young SR, Dyson M. The effect of therapeutic ultrasound on angiogenesis. Ultrasound Med Biol. 1990;16:261-269.
7. Mortimer AJ, Dyson M. The effect of therapeutic ultrasound on calcium uptake in fibroblasts. Ultrasound Med Biol. 1988;14:499-506.
8. Byl NN, McKenzie A, Wong T, West J, Hunt TK. Incisional wound healing: a controlled study of low and high dose ultrasound. J Orthop Sports Phys Ther. 1993;18:619-628.
9. Brossman E, Giernat L, Slusarczyl-Zalobna A, Torzecki Z. Histological effects on collagen production in late vitro growth phase of human fibroblasts. Rheumatologica. 1981;19:177-181.
10. Enwemeka CS. The effects of therapeutic ultrasound on tendon healing: a biomechanical study. Am J Phys Med Rehabil. 1989;68:283-287.
11. ter Riet G, Kessels AGH, Knipschild P. A randomized clinical trial of ultrasound in the treatment of pressure ulcers. Phys Ther. 1996;76:1301-1312.
12. McDiarmid T, Burns PN, Lewith GT, Machin D. Ultrasound and the treatment of pressure sores. Physiotherapy. 1985;71:66-70.
13. Eriksson SV, Lundeberg T, Malm M. A placebo controlled trial of ultrasound therapy in chronic leg ulceration. Scand J Rehabil Med. 1991;23:211-213.
14. Callam MJ, Harper DR, Dale JJ, Ruckley CV, Prescott RJ. A controlled trial of weekly ultrasound therapy in chronic leg ulceration. Lancet. 1987;July 25:204-206.
15. Lundeberg T, Nordström F, Brodda-Jansen G, Eriksson SV, Kjartansson J, Samuelson UE. Pulsed ultrasound does not improve healing of venous ulcers. Scand J Rehabil Med. 1990;22:195-197.
16. Dyson M, Suckling J. Stimulation of tissue repair by ultrasound: a survey of mechanisms involved. Physiotherapy. 1978;64:105-108.
17. Portney LG, Watkins MP. Foundations of Clinical Research: Applications and Practice. Norwalk, Conn.: Appleton and Lange; 1993.
18. Ferrell BA. Assessment of healing. Clin Geriatr Med. 1997;13:575-587.
19. Dyson M. Mechanisms involved in therapeutic ultrasound. Physiotherapy. 1987;73:116-119.