A survey of therapeutic ultrasound use by physical therapists who are orthopaedic certified specialists.

For decades, ultrasound (US) has been a widely used and well-accepted physical therapy adjunct modality throughout the world, (1-7) particularly for the management of musculoskeletal conditions. (1,2,4-7) Ultrasound was first introduced as a therapeutic modality in the 1950s, when both animal and human studies demonstrated its ability to safely heat tissue several centimeters below the skin, particularly tissues that are high in collagen. (8,9) This finding, combined with studies demonstrating that collagenous structures yield better to a stretch when they are warmed (within therapeutic ranges), bolstered the popularity of US, particularly for conditions such as shortened tendons and muscles, joint capsule tightness, and scar tissue limitations. (10) In the late 1960s and 1970s, reports on the nonthermal therapeutic effects of US, primarily in the area of enhanced tissue healing, further bolstered its popularity. (11,12)

Although US has been a popular adjunct modality for decades, the lack of studies confirming its benefits has led contemporary scientists to question the traditional view of its therapeutic benefits. (3) Indeed, eleven (13-23) of fifteen (3,13-26) systematic reviews could not draw any definitive conclusions about the effectiveness (or lack of effectiveness) of US because of insufficient evidence. Several reviews (3,16,24,26,27) also report disagreement and confusion about the most efficacious treatment parameters for US.

Against this backdrop, anecdotal reports from clinicians suggest that US remains a popular modality. It has been nearly 20 years since the use of US in the United States has been assessed. (4) In that study, 79% of respondents reported using US at least once per week, and 45% more than 10 times per week. The study, however, did not attempt to identify specific conditions for which US was used. Has US remained a popular modality and, if so, for which conditions is it most often used?

The purpose of this descriptive study was to determine how frequently physical therapists with experience in orthopedics use US and to determine their opinions about the clinical importance of US in achieving beneficial outcomes when managing common musculoskeletal impairments (eg, pain, soft tissue inflammation, tissue extensibility limitations, tissue injury, soft tissue swelling, and scar tissue limitations). Three research questions were addressed:

(1) How often do orthopedic physical therapists use US as part of the management of selected impairments?

(2) Do orthopedic physical therapists believe US is clinically important in managing these impairments?

(3) What US treatment parameters do orthopedic physical therapists use for management of these impairments?

Method

Target Population

Given the range of settings in which physical therapists practice, we wanted to provide some assurance that those responding to the survey had recent experience in orthopedics and were likely to be familiar with the orthopedic physical therapy literature. Thus, we chose to limit our population to physical therapists who have achieved the designation of Orthopaedic Certified Specialist (OCS). To become an orthopaedic certified specialist, physical therapists must demonstrate a minimum of 2,000 hours of direct patient care in orthopedics over the previous 10 years, with 25% of those hours occurring in the last 3 years. (28) Physical therapists seeking certification as an OCS must also pass a written examination that assesses competency in the specialty practice of orthopedic physical therapy. An OCS must recertify every 10 years.

Although we have no direct assurance that physical therapists with the OCS designation are familiar with the current literature about US specifically, the nature of the certification process requires successful completion of a rigorous certification examination aimed at current best practice in orthopedic physical therapy. The likely career path of individuals who choose to pursue OCS certification makes it reasonable to assume that most individuals with the OCS certification are working with patients with orthopedic conditions.

Sampling Procedures

The American Physical Therapy Association's (APTA) List Rental Services Department randomly selected a sample of approximately 400 members of APTA who had an OCS certification. The List Rental Services Department randomly selected the Northeast and Mid-Atlantic region of the United States for the sample and provided mailing labels for all 457 APTA members with an OCS certification living in this region (Connecticut, District of Columbia, Delaware, Maine, Maryland, Massachusetts, New Hampshire, New Jersey, New York, Pennsylvania, Rhode Island, Vermont, Virginia, and West Virginia). In March 2003, following approval by the Marymount University Institutional Review Board for the Use of Human Subjects in Research, a survey questionnaire and cover letter explaining the study and informed consent procedures were mailed to all 457 physical therapists with an OCS designation. The return of a completed survey questionnaire implied the consent to participate in the study. Ten weeks after the first mailing, a second copy of the survey instrument was mailed to those who had not responded.

Respondents

Of the 457 survey questionnaires mailed, 213 were returned. Six of the 213 returned survey questionnaires indicated respondents did not want to participate in the study, yielding a 44.9% usable response rate. Respondents were 97 men and 104 women. Chi-square analyses revealed no significant differences in the proportion of survey respondents from each state compared with the OCS population in each state reported by the American Board of Physical Therapy Specialties. (28) The age of the respondents ranged from 26 to 70 years ([bar.X.bar]=42.7, SD=7.6). The number of years since initial licensure as a physical therapist ranged from 3 to 41 ([bar.X.bar]= 17, SD=7.4). The number of years of OCS certification ranged from 1 to 19 ([bar.X.bar]=5.5, SD=3.7).

Survey Instrument

We developed a 77-question survey instrument by first searching the existing research literature to identify impairments, treatment goals, and machine parameters commonly associated with the use of US. Based on this literature review, a list of potential survey questions was compiled and placed in draft survey format. This initial survey was presented to 10 experienced physical therapists. Feedback about the accuracy, comprehensiveness, and clarity of the survey's content was obtained from these initial reviewers and used to develop a more definitive survey for pilot testing of face and content validity.

The pilot survey instrument was examined sequentially by 3 separate groups of 5 physical therapists working in orthopedics. Each physical therapist completed the survey independently and then discussed his or her responses within a small group. The physical therapists described their interpretations of each question and explained why they answered each question as they did. For those questions for which there were discrepancies in interpretation among respondents, we sought suggestions for rewording. The research team considered the feedback from each group and made revisions based on this feedback. We then modified the survey instrument and presented it to the next small group. We continued to pilot test the instrument on small groups of clinicians until we achieved agreement on 90% of the items. When the research team reached a consensus on the instrument, a university professor with experience in survey research reviewed and approved the revised instrument.

The regularity of physical therapists' use of US for each of the following 6 impairments was assessed by the survey instrument: pain, soft tissue inflammation (eg, tendinitis, bursitis), decreased tissue extensibility, delayed tissue healing, soft tissue swelling (eg, edema, joint effusion), and scar tissue remodeling. Respondents were asked to "Indicate the percentage of patients for whom you would include US, with or without phonophoresis, as an adjunctive therapy for each of the following impairments." Respondents chose one of the following responses: (1) "with less than 10% of patients," (2) "for about 25% of my patients," (3) "for about 50% of my patients," (4) "for about 75% of my patients," and (5) "for more than 90% of my patients." The option was available to list a maximum of 2 "other" impairments for which they would use US, in addition to the 6 listed impairments.

To assess beliefs about the clinical importance of US for managing the same 6 impairments, respondents were asked "For those patients whom you do treat with US, how clinically important do you believe US is in achieving each of the following therapeutic goals?" Respondents chose one of the following responses: (1) "I would not use US," (2) "US is not important," (3) "US is minimally important," (4) "US is somewhat important," (5) "US is very important," or (6) "US is essential." Again, the option to list 2 additional "other" impairments was available.

Lastly, respondents indicated the treatment parameters (duty cycle, intensity, frequency) they would use to achieve therapeutic goals related to each of the previously named impairments. Respondents selected 1 of 4 duty cycles (10%, 20%, 50%, or 100%) and 1 of 2 frequencies (1 MHz or 3 MHz), and they indicated the intensity (in watts per square centimeter) they would use for superficial and deep tissue. Respondents had the option to select "N/A" if they did not use US for the impairment. Demographic information was gathered to identify respondent's age, sex, state of residence, year of graduation from a physical therapist education program, and year of certification as an orthopedic certified specialist.

Data Analysis

Data from completed survey questionnaires were entered into the Statistical Package for the Social Sciences (SPSS), version 11.0. * Descriptive statistics were used to characterize the shape, central tendency, and variability within the study sample. The frequency of each response was calculated and examined for trends. The relationship between the perceived clinical importance of US and the frequency of US use for each of the 6 impairments was analyzed using the Spearman rho correlation coefficient. We expected to find that respondents who believe US is clinically important would use it more frequently. Six separate correlations were performed. Therefore, the Bonferroni correction factor was used to calculate the level of significance, as protection against a type I error. Using this correction factor, the significance level was established at P<.008. (29)

The temporal average intensity (TAI) was calculated as the product of the duty cycle and average pulse intensity (in watts per square centimeter) that the respondent reported using for each of the impairments. This calculation reflects the average intensity of the US beam relative to the duty cycle. This study did not examine the total amount of US energy delivered in relationship to the treatment time or size of the sound head.

Results

Respondents who reported using US for at least 25% of their patients (choosing any 1 of 4 survey responses: "for about 25% of my patients," "for about 50% of my patients," "for about 75% of my patients," or "for more than 90% of my patients") were categorized as "likely to use US." Respondents who reported using US "with less than 10% of my patients" were categorized as "unlikely to use US." Table 1 displays aggregated frequency counts for the categories "likely to use US" and "unlikely to use US." In this study, 83.6% of the respondents indicated they were likely to use US to decrease soft tissue inflammation, 70.9% to increase tissue extensibility, 68.8% to enhance scar tissue remodeling, 52.5% to increase soft tissue healing, 49.3% to decrease pain, and 35.1% to decrease soft tissue swelling.

Additionally, respondents added 19 impairments under the category of "other." The most frequent "other" conditions were muscle spasm (4 respondents), calcium deposits (3 respondents), and hematoma (3 respondents). The remaining 9 "other" conditions were each listed by 1 respondent and represented either a physiological response (increase cell permeability, increase blood flow, stimulate wound healing) or a pathology/injury (sprain, strain, joint capsule inflammation, plantar fasciitis, bone healing, temporomandibular disorder). There were too few conditions in any one "other" category for further analysis of these responses.

Aggregated frequency counts summarize respondents' judgment about the clinical importance of US in managing each of the impairments addressed by the survey (Tab. 2). For each of the impairments, the responses "US is somewhat important," "US is very important," or "US is essential" were grouped together into the category "US is clinically important." The responses "US is not important" and "US is minimally important" were grouped together into the category "US is not clinically important." Table 2 also identifies the number of respondents who responded "I would not use US." In this study, 71.1% of all respondents identified US as clinically important for the management of tissue extensibility, 62.6% for soft tissue inflammation, 53% for remodeling scar tissue, 47% for tissue healing, 39.4% for pain control, and 27.9% for decreasing soft tissue swelling. For each of the 6 impairments, a statistically significant and moderately strong Spearman rho correlation was found between "frequency of use of US" and "perceived clinical importance of US" (Tab. 3).

Spearman rho correlations also were performed to determine any potential relationship between year of initial physical therapy licensure and perceived clinical importance of US for managing each of the 6 impairments. Very weak, but statistically significant, correlations were identified for pain management ([r.sub.s]=.24), tissue extensibility ([r.sub.s]=-.20), soft tissue inflammation ([r.sub.s]=.24), scar tissue management ([r.sub.s]=-.20), and soft tissue swelling ([r.sub.s]=-.20). Correlation coefficients (r values) less than .25 are generally interpreted as showing little or no relationship between the variables. (29)

Table 4 summarizes the frequency responses for preferred US parameters for each of the 6 impairments under review: duty cycle (10%, 20%, 50%, 100%), intensity, and frequency (1 MHz, 3 MHz). Respondents could choose "N/A" if they did not use US for the specific impairment. Less than 2.5% of respondents chose 10% pulsed US for any impairment; therefore, in Table 4, responses of "10%" and "20%" were combined to form the category "less than or equal to 20% duty cycle." Continuous US (100% "on" duty cycle) was chosen almost exclusively when the goal was to increase tissue extensibility (93.6%, n=160). Continuous US also was the most frequently chosen pulse mode when the goal was to decrease pain (75%; n=105) or to remodel scar tissue (81.8%, n= 130). A pulsed mode (either [less than or equal to] 20% or 50%) was chosen most frequently when the goal was to decrease soft tissue swelling (82.3%, n=79) and decrease soft tissue inflammation (70.5%, n=117).

Respondents identified the intensity setting they would use to achieve each therapeutic goal based on whether US was being applied to superficial or deep tissue. The intensities identified by respondents ranged from 0.10 W/[cm.sup.2] to 3.30 W/[cm.sup.2] (bimodal central tendency scores= 1.00 W/[cm.sup.2] and 1.50 W/[cm.sup.2]) for superficial tissues and from 0.40 W/[cm.sup.2] to 4.00 W/[cm.sup.2] (mode=1.50 W/[cm.sup.2]) for deep tissues. Table 5 displays the number of respondents who indicated they would use US to deliver medication (phonophoresis) for each of the 6 impairments reviewed.

From the parameters that reportedly would be used to achieve the 6 therapeutic goals, a TAI was calculated as the product of the duty cycle (expressed as a percentage) and the pulse average intensity. The mean TAI scores were highest for decreasing pain in deep tissues (1.58 W/[cm.sup.2]) and lowest for decreasing soft tissue swelling in superficial tissues (0.52 W/[cm.sup.2]). Respondents consistently used 3 MHz when treating superficial tissues and 1 MHz for deep tissues, regardless of the therapeutic goal.

Discussion

Use of Ultrasound

The frequency responses for likelihood of using US for each of the 6 impairments clustered into 3 general ranges. In the high range, at least two thirds of respondents indicated they would use US for soft tissue inflammation (83.6%), tissue extensibility (70.9%), and scar tissue remodeling (68.8%). Two impairments clustered in the middle range with frequency counts fairly evenly divided between respondents who would use US versus those who would not use US: pain management (49.3% would use US) and tissue healing (52.5% would use US). One impairment, soft tissue swelling, fell in the lower range, with only 35.1% of respondents indicating that they would use US for this condition.

About haft of the respondents (54.1%) reported they would use US to deliver medication via phonophoresis for soft tissue inflammation, 22.2% of respondents would use phonophoresis for pain management, and 19.8% of respondents would use phonophoresis for soft tissue swelling. We did not anticipate high use of phonophoresis and did not include any survey items to further delineate medication preferences.

In providing insights into changes in the use of US over time, the findings of this study can only be indirectly compared to the 1988 study by Robinson and Snyder-Mackler, (4) which is the next most recent study of US use in the United States. Differences in the characteristics of the respondents surveyed, the patient population addressed, and the specific questions asked to identify frequency of US use make direct comparisons impossible. Our study had a more narrow focus, targeting physical therapists who treated patients with orthopedic problems and asking respondents to estimate the percentage of patients they treat with US. In contrast, Robinson and Snyder-Mackler did not specify any particular group of conditions and quantified use by numbers of patients per week rather than percentage of patient caseload.

Similarities in outcomes, however, are noticeable and do not suggest any major change in the use of US over the last 19 years. In the study by Robinson and Mackler, (4) 45% of respondents indicated that they used US more than 10 times per week; 64% indicated that they used US 6 or more times per week. In our study, when scores across impairment categories were averaged together, approximately 60% of respondents indicated they were likely to use US, with a frequency count across specific impairment categories varying between 35.1% and 83.6%.

Over the past 10 years, researchers from Australia, (1) Canada, (5) and the Netherlands (6) have examined US use in their countries. The specific nature of the questions differed some what among the studies. In both Canada and Australia, however, clinicians indicated that they used US at least once a day. This included 93.7% of private practice physical therapists in Alberta, Canada, (5) and 84.7% of sports physical therapists in Australia. (1) The Dutch study (6) reported that physical therapists practicing in the Dutch primary health care system chose US 30% of the time as the primary intervention to address at least 1 of 4 treatment goals for their patients. Respondents could only identify 1 intervention as "primary" for each of 4 treatment goals. These studies from across the globe support the findings of our study. Ultrasound is a commonly used modality in physical therapist practice and is perceived to be important by clinicians for selected conditions.

Perceived Importance of US

We speculated that a high correlation between scores for the frequency of use of US and beliefs about the clinical importance of US existed. Indeed, statistically significant and moderately strong correlations were identified for each of the impairment categories. Respondents who believed US was clinically important were more likely to use US than those who did not believe US was important. We did not ask respondents to provide any justification for their opinion about the importance of US in the management of their patients. This is a notable question that was beyond the scope of this study. Future examination of this topic could provide insights to guide further research.

It would be appropriate to determine whether the impressions of the respondents were shaped primarily by their reflective judgments about the clinical outcomes they have observed in their patients, by their interpretation of the current literature, or by their personal comfort with this well-established and popular modality. Were their decisions primarily pragmatic based on availability and ease of use of US compared with alternative treatments? Do their patients request US? Is US typically reimbursable with adequate justification of need, whereas alternative therapies may not be? These are a few of the questions that could help qualify responses of those surveyed.

The respondents in our study had been licensed physical therapists for an average of 17 years (SD=7.4) and averaged 5 years as an OCS. We speculated that those who entered clinical practice at a time when the emphasis on evidence-based practice was less well established might consider US more clinically important when compared with more recent graduates. For each impairment category, therefore, a correlation was performed comparing the date of initial licensure as a physical therapist and respondent opinion of the importance of US. We found a very weak to absent relationship between these variables with [r.sub.s] values less than .25 for each category. We concluded, therefore, that, in this group of orthopaedic clinical specialists, the length of time since initial licensure as a physical therapist did not substantively influence respondent opinion of the importance of US in clinical practice.

Future Research Needs

There remains much controversy over the appropriateness of US as an adjunct modality in physical therapy. The biophysical basis of its use, particularly the nonthermal effects, has become controversial, (30) and there are very few controlled trials that have found US to be clinically effective. Most systematic reviews of the topic conclude that there are very few clinical trials to either support or refute the benefit of US. (13-23) If physical therapists base their decision to use US primarily on clear demonstration of effectiveness in the scientific literature, then they are likely to stop using US based on lack of supporting evidence. However, against this near void of clinically applicable research, we found experienced and advanced-practice clinicians continuing to use US regularly for specific impairments often encountered in orthopedic settings, and indicating that US has an important role in managing selected impairments. In 2001, Robertson and Baker (3) called for high-quality clinical trials to address this concern and help end this confusion. To date, very few clinical trials of this nature have been completed. Our study provides insights into how advanced practice clinicians are using US to manage musculoskeletal conditions. This information should help guide researchers in identifying key conditions to target for future clinical trials.

Strengths and Limitations of the Study

The strength of the design of this survey was its clarity, the ability for respondents to complete the survey quickly, and the variety of impairments addressed. This format also served as a limitation of the study. The survey was organized around impairments, not specific clinical scenarios. The use of clinical case scenarios (eg, describing a typical patient with lateral epicondylitis) would have provided greater detail from which to ground clinical decisions about US use and specific machine parameters. This, however, would have required us to present a detailed scenario for each of the impairments assessed and would have made the survey unacceptably long and tedious to complete.

This study sampled only physical therapists practicing in the Northeast and Mid-Atlantic region of the United States. Regional differences across the United States may be present and, by the nature of our sample, were not captured.

We also limited the number of US parameters to be considered. This study did not examine respondent opinions about the size of the treatment area, size of the sound head, or treatment time. Thus, estimates of the total amount of US energy delivered could not be calculated. We did not seek opinions about the number and frequency of treatment sessions. Each of these parameters has been identified in the literature as being potentially important to the effectiveness of US. (10,27,31) We limited the investigation to the most fundamental parameters in order to keep the survey simple and time efficient. The specific size of the treatment area, size of the sound head's effective radiating area, and treatment time are all interconnected concepts. Including these parameters in the survey would have provided more definitive information but would have required organizing the survey around patient case scenarios. Indeed, 4 respondents did not fully complete the parameters section, commenting that they needed more information about the patients in order to make a decision.

This study did not ask respondents to confirm actual use of US through documentation such as a review of prior treatment logs or by maintaining a prospective treatment log. Rather, respondents simply provided their best judgment at one particular point in time. There may be differences between their perception of their own behavior and their actual behavior. We also chose not to inquire about the number of treatment sessions respondents typically administered for each incident of care. We believed that the influence of insurance coverage would complicate the interpretation of responses. Without adding questions to the survey, it would be difficult to determine whether responses about the number of treatment sessions represented the physical therapist's view of the ideal number of treatment sessions or a less than ideal number of treatment sessions approved by the third-party payer. Although this is an important consideration, we believed answering this question was best left to future research.

Conclusion

Our study demonstrates that physical therapists with the OCS designation practicing in the Northeast and Mid-Atlantic region of the United States regularly use US and believe US is clinically important for managing selected musculoskeletal impairments. In this study, the 3 most common impairments that US was used to manage were soft tissue inflammation (83.6%), tissue extensibility (70.9%), and scar tissue remodeling (68.8%). Similarly, 71.1% of respondents identified US as clinically important for managing tissue extensibility, 62.6% for soft tissue inflammation, and 53% for remodeling scar tissue. More than half of respondents indicated they use US to deliver medication (phonophoresis) for soft tissue inflammation. These findings support the need to further investigate the underlying clinical decision-making factors that contribute to the use of US by physical therapists and to engage in research that helps answer critical questions about the clinical effectiveness of US.

Invited Commentary

Val J Robertson

Ultrasound offers many clinical users considerable potential. Many current uses of ultrasound (US) made by physical therapists, however, probably contribute little to patients' outcomes. In years to come, the findings of the survey by Wong et al (1) will provide an interesting reality check for physical therapists: When did we accept how little evidence there is that supports how physical therapists use US? The findings suggest that, instead of adjusting practice according to research evidence, many physical therapists persist with long-discredited ideas and practices that are not evidence based. It also is not apparent that they are adopting practices justified by evidence.

Wong et al (1) chronicled the self-reported uses of US by physical therapy experts for 6 impairments. The 205 contributing experts (44.9% effective return rate, with 457 physical therapists surveyed) were all orthopaedic certified specialists (OCS), and the specified impairments were "pain, soft tissue inflammation, decreased tissue extensibility, delayed tissue healing, soft tissue swelling, ... and scar tissue remodeling." The authors indicate that they chose common impairments and provided opportunities for OCS respondents to add "other" options. The most frequently reported "other" impairments that the respondents identified managing with ultrasound were muscle spasm, calcium deposits, and hematoma. Other uses of US that received only one mention each included to increase blood flow and enhance bone healing.

The article also provides a summary of the most frequently reported parameters used (Tab. 4). To summarize, a frequency of 3 MHz was used for superficial lesions and a frequency of 1 MHz was used for deep lesions (not defined). A typical intensity, apparently irrespective of frequency, was 1 to 1.5 W/[cm.sup.2]. The space-averaged, time-averaged intensity ([I.sub.SATA]) (which Wong et al (1) called the "temporal average intensity") varied from 0.02 to 3.3 W/[cm.sup.2] for superficial tissues and from 0.10 to 2.5 W/[cm.sup.2] for deep tissues. These findings suggest little understanding by the OCS respondents of the physical properties of US, including the effects of frequency on the absorption of US energy. (2(p262))

More than 50% of the OCS respondents reported US as clinically important (somewhat important, essential, or very important) for treating 3 of the 6 impairments: soft tissue inflammation, decreased tissue extensibility, and when scar tissue remodeling was needed. The absence of supporting peer-reviewed evidence for this finding suggests that practice possibly is unduly altered by reimbursement pressures or the need to meet patients' expectations of what a physical therapy treatment should include.

There are 2 ways of interpreting the findings of the survey by Wong et al. (1) One way is with some optimism that maybe the nonresponding experts (55.1% of those surveyed) recognized the contradiction in investigating uses of a modality to manage conditions for which the researchers recognized there is no evidence to support its effectiveness. The alternative is further confirmation that introducing evidence-based change to practice is very difficult. The present commentary will assume this latter interpretation: nearly 50% of experts in a relevant area reported uses that are difficult to justify on the basis of peer reviewed research evidence. With this in mind, I will focus on 3 issues: (1) how the findings of the survey by Wong et al contribute to practice development, (2) the effectiveness of therapeutic US, and (3) dosage issues.

Practice Development

There is an underlying contradiction in the study by Wong et al (1): as the authors report in the "Discussion" section, "there are very few controlled trials that have found US to be clinically effective." Despite this, the researchers investigated the use of US for 6 conditions for which the authors indicate there is little or no high-quality evidence that it improves the outcomes. (3-5) Why, when, and how is investigating a seemingly ineffective practice of using a modality such as US justified?

Investigating practices known as unlikely to be effective can be very informative. From a "big picture" perspective, results such as those reported by Wong et al (1) can provide readers with insights into whether practice is adopting uses supported by research.

Wong and colleagues' findings, (1) therefore, are quite disappointing. Their findings suggest that research published over the past 10 years has had little effect on clinical practices in physical therapy that involve using therapeutic US (as distinct from diagnostic US). The same practices appear to be continuing, apparently with few new options being adopted.

The questionnaire that Wong et al (1) used in their survey was not provided, but the data and description of the study suggest that the researchers missed an ideal opportunity to monitor the adoption of new uses of US by physical therapists, including for diagnostic imaging and for biofeedback during muscle retraining. It is difficult to accept that the study by Wong et al will contribute, as they suggested in the last sentence of their abstract, to helping "researchers prioritize needs for future research on the clinical effectiveness of US." Rather, the study suggests a need to investigate how to provide clinicians with more ways of accessing and integrating knowledge into practice. This need likely extends beyond US, electrical stimulation, and other modalities.

Effectiveness of Ultrasound

The second issue related to the study by Wong et al (1) is: When, and for what, is there evidence that applying US is effective? It is used frequently for imaging a growing list of tissues as the technology improves, for guiding needles during procedures, (6) for biofeedback, and for mixing compounds. Ignoring those and many other related uses, there also is clear evidence that US can promote healing in some types of tissues under some circumstances.

A number of articles published around the turn of the century attest to the value of using US to manage fractures and delayed union. (7-9) The outcomes and the possible mechanisms were well described in an article written for sports medicine clinicians. (10) That same article also canvassed other possible uses of US for managing soft tissue problems. They included promoting the repair of ligaments, tendons, cartilaginous tissues, and muscle. None of these conditions were included or mentioned by Wong et al. (1)

Laboratory researchers have investigated additional possibilities. A cursory search retrieved many current and relevant studies. For example, at the cellular level, pulsed US can promote proteoglycan synthesis in primary bovine chondrocytes. (11) This effect is more than just that induced by the heat produced by US. Such findings may result in future clinical uses in humans, or they may not be relevant in an in vivo context. In rabbits, osteochondral defects treated with US for 40 minutes a day improved faster than those treated for any lesser time. (12)

Surgically induced damage at the bone-to-tendon junction of the patella-patellar tendon complex of rabbits repaired better if treated with low-intensity pulsed US. (13) By week 4, those rabbits treated with US showed evidence of more advanced repairs, and by week 16, this difference was clear. At that stage, biomechanical testing demonstrated a higher failure load and greater strength in the bone-to-tendon junction in those rabbits treated with US than in the control group. It is unclear whether the control group was handled identically and sham ultrasound also was applied. This possible confound was managed in a study using rats and examining, in part, the effects of US on knee ligament healing. (14) Again, the findings showed that US applications produced, even after 2 weeks, an increased ultimate strength and stiffness in the US-treated ligaments.

These results from animal studies suggest that US holds considerable potential for clinical uses in physical therapy. Some caution is warranted despite the findings being consistent with evidence that low-intensity pulsed US promotes fracture union in humans: controlled in vivo studies in humans have not yet been reported, and the parameters and dosages that Wong et all reported physical therapists now use are very different.

Fracture repair is promoted by low-intensity pulsed US. The parameters usually used are a frequency of 1.5 MHz, a beam nonuniformity ratio of 2.16, an effective radiating area of 4 [cm.sup.2], and an [I.sub.SATA] of 0.03 W/[cm.sup.2].(2) In addition, the dosage is usually daily, 20-minute applications, using a stationary applicator. The parameters are produced by a system specifically designed and used to treat fractures (Exogen). The extent to which these parameters may be varied is not yet clear, but this dosage and system of usage raises more questions.

By definition, low-intensity pulsed US has an [I.sub.SATA] of less than 0.1 W/[cm.sup.2]. (15) Warden et al (15) investigated whether using a standard, conventional therapeutic US unit could produce the same outcome as that produced by the purpose-designed US equipment known to promote bone union. They used rats with surgically created femoral fractures. After 40 days, more healing had occurred in those rats treated with the active US than in those rats treated with the sham low-intensity pulsed US. The parameters used included a frequency of 1.0 MHz, pulsed so that the [I.sub.SATA] was 0.1 W/[cm.sup.2]; a beam nonuniformity ratio of less than 6; an effective radiating area effective radiating area of 5 [cm.sup.2]; and daily 20-minute applications 5 times a week with a stationary applicator. The machine output was tested and confirmed weekly during the study.

Questions remain: How effective is conventional equipment for managing human fractures and bone damage? To what extent can US improve the rate and extent of recovery of a range of types of tissue damage in humans? This includes ligament and cartilaginous damage, injuries for which there currently is little convincing evidence.

Another issue is, how realistic is it to expect US to improve the rate of healing of normal tissues? The conditions for which there is supporting evidence typically concern tissues that are slow to heal, where healing has been delayed for various reasons, or detailed biomechanical studies are done after sacrificing an animal. The initial stages of acute soft tissue healing normally take less than 2 weeks, and patients may not want to make return daily visits for additional weeks to get an unseen benefit. This also raises an interesting problem for clinical users of US: If a condition with a usual rate of initial repair of 14 days is speeded up by, say, 10% by using US, would the outcome measures used clinically detect this change? Similarly, with changes in the ability of ligaments to respond to lengthening, will it ever be possible clinically to demonstrate a faster rate of strength development in vivo?

Fractures are typically managed with US for 20-minute sessions once per day. Few physical therapy clinicians apply US this often and for this long to a dosage area. Given the parameters that Wong et al (1) reported as favored, this is a blessing. Skin burns of possibly considerable depth would be inevitable with either frequency of US, especially if applied with a stationary applicator.

Existing clinical and laboratory findings suggest that considerable change is needed to existing uses of US made by physical therapists. Disappointingly, only one respondent in the study by Wong et al seemed aware of the use of US for promoting bone healing (ie, from among 205 clinical experts).

Another issue with US is, when should this modality be preferred? For example, as a method of increasing tissue extensibility, when should it be preferred over an application of heat, of heat and stretching, or of stretching alone? The relevant component of US for increasing tissue extensibility is accepted as heat, but this can be provided by other means.

Shortwave diathermy (SWD) and microwave, for example, are known to be effective methods of deep heating. Even in young, uninjured people, applying SWD can increase the range of movement available in an adjacent joint. (16) The mechanism is not understood, but the effect is readily demonstrated, and SWD enables heating of a large tissue volume. By contrast, the temperature of only small volumes of tissue can be effectively heated using therapeutic US.

There is no evidence that US is effective for managing pain, and little is known about the comparative effectiveness of modalities including heat and cold for relieving different types of pain or how well they compare with different types of medication. Similarly, the relative benefits of different modalities and electrical stimulation (eg, pulsed current from a transcutaneous electrical nerve stimulation machine) are not known. Physical therapists need evidence-based guidelines for making choices from a range of specific modalities. Clinicians are best placed to investigate these types of questions, particularly experts who would usually treat a higher number of patients with relevant presenting problems.

In summary, in vitro and some in vivo research shows that US can improve some outcomes for patients. However, these uses were not those investigated by Wong et al, (1) nor did more than 0.4% (1/205) of the experts who responded indicate knowledge of the types of findings for which research-based evidence exists. This is despite a recent article (15) in this same journal reporting the possible relevance of standard therapeutic machines for managing bony damage in vitro. Neither do the self-reported types of uses made of US suggest that many of the therapists surveyed recognize that alternatives to US might be more effective. Pain and reduced tissue extensibility, for example, are possibly better managed by alternatives to US.

Dosages

The third issue to address in this commentary is dosage. Dosage problems have dogged studies of therapeutic US. One problem concerns the lack of stability of output produced by US equipment. This problem has been well documented in locations ranging from North Wales (17) to Texas (18) to New Zealand (19) and has been known for a long time (19,20) and their implications discussed often and in many places. (21) More recent evidence shows that, although the tested equipment met Food and Drug Administration regulations and was consistent with the manufacturer's advice, considerable differences could exist in the space-averaged intensity produced by transducers with the same frequency. (22) A serious omission in the survey by Wong et al (1) was a question asking clinicians how often they have their equipment tested for electrical safety and the output calibrated. The results of that question might have been very enlightening.

Ultrasound has been used clinically for nearly 70 years, (23) well prior to the 1950s as Wong et al (1) suggested. Despite that, no dose-response relationship has been identified. This makes it difficult to understand why therapists choose the dosages they use. The patterns of absorption of US at frequencies of 1 and 3 MHz are different. Dosages should reflect this difference, but they do not, suggesting that some users lack an understanding of the basic properties of US. To apply US at a frequency of 3 MHz and an intensity of approximately 1 W/[cm.sup.2] ([I.sub.SATA] range=0.023.3 W/[cm.sup.2]) over a superficial lesion seems excessive unless there is no underlying bone and the applicator is moved rapidly. Otherwise, the dosage does not appear to account for the high rate of absorption of 3-MHz US in a small volume of tissue if over a bone. (2) Even without bone directly underneath a superficial lesion, this is still a high dosage and not seemingly consistent with using US with a frequency of 1 MHz at only slightly different intensities, from 0.1 to 2.5 W/[cm.sup.2], over deep tissue. These assumptions may be incorrect, because Wong et al collapsed the 1- and 3-MHz US dosages but indicated that respondents preferred them for deep and superficial lesions, respectively. This leaves readers in a conundrum: Do respondents not know and understand the differences and modify dosages accordingly, or is this interpretation a product of how the results are presented? Hopefully, the higher intensities reported were for US applied at a frequency of 1 MHz, not 3 MHz, and a few survey responses skewed the findings unduly.

Variations in dosages reported in studies of US as used by physical therapists are always intriguing. Robertson and Baker, in an earlier systematic review of clinical uses of US, commented on this, saying, "the dosages of ultrasound used in the studies we reviewed varied considerably and for reasons that were not always clear." (5(p1348)) The findings of the survey by Wong et al imply that nothing has changed and that even experts use a very wide range of dosages. Admittedly, in this instance, users were asked to self-report, no checking was possible, and respondents had to imagine an ideal patient for each of the 6 conditions included.

Taken as a whole, the dosages reported by Wong et al (1) should concern the profession, given they suggest a limited understanding of the relevant biophysical properties of US. This is compounded by long-identified problems with US equipment. The question was not asked, so we can only assume that much of the equipment used is not necessarily in optimal condition and recently tested for electrical safety and output.

Summary

The study by Wong et al (1) raises concerns as to how well physical therapists integrate current research findings and use them to change practice. The results suggest that a considerable number of experts use practices that the researchers themselves reported as not supported by evidence. Whether this was intentional or a by-product of pressures such as for reimbursement and patient expectations is not known.

The near 50% nonresponse rate might suggest that many experts recognized that the questions asked by the researchers were only exploring uses known not to be effective rather than the effect of new findings on the profession. Perhaps, but the findings also suggest unjustifiable differences in dosages, another fallout from the lack of supporting peer-reviewed evidence and the continuing absence of a dose-response relationship.

I am still left with 2 major puzzles. First, why did Wong et al (1) investigate patterns of usage of a modality for which there is little evidence of effectiveness? If this was a test of integration of research into practice, then discussion of this issue and the findings could have been very interesting. Second, why claim to investigate the "conditions [for which US is] ... most often used" and then constrain respondents to the 6 categories with only an option of "other"? This was an ideal opportunity to identify what OCS therapists use US for and why.

The main contribution of the article by Wong et al (1) will remain a reality check for those concerned with the integration of peer-reviewed evidence into practice, and the findings provide cause for concern as this study was of clinical experts in physical therapy.

VJ Robertson, PT, PhD, is Professor, Teaching & Research Unit, Gosford Hospital, Bldg 14, Corner of Holden Rd and Racecourse Rd, Gosford, New South Wales 2250, Australia. Address all correspondence to Dr Robertson at: [e-mail]

DOI: 10.2522/ptj.20050392.ic

References

(1) Wong RA, Schumann B, Townsend R, Phelps CA. A survey of therapeutic ultrasound use by physical therapists who are orthopedic certified specialists. Phys Ther. 2007;87:986-994.

(2) Robertson VJ, Ward AR, Low J, Reed A. Electrotherapy Explained. 4th ed. London, United Kingdom: Elsevier Science Ltd; 2006.

(3) Speed C. Therapeutic ultrasound in soft tissue lesions. Rheumatology. 2001;40:1331-1336.

(4) Gursel K, Ulus Y, Bilgic A, et al. Adding ultrasound in the management of soft tissue disorders of the shoulder: a randomized placebo-controlled trial. Phys Ther. 2004;84:336-343.

(5) Robertson VJ, Baker KG. A review of therapeutic ultrasound: effectiveness studies. Phys Ther. 2001;81:1339-1350.

(6) Grassi W, Filippucci E, Busilacchi P. Musculoskeletal ultrasound. Best Pract Res Clin Rheumatol. 2004;18:813-826.

(7) Nolte PA, Klein-Nulend J, Albers GH, et al. Low-intensity ultrasound stimulates endochondral ossification in vitro. J Orthop Res. 2001;19:301-307.

(8) Takikawa SC, Matsui N, Kokubu T, et al. Low-intensity pulsed ultrasound initiates bone healing in rat nonunion fracture model. J Ultrasound Med. 2001;20:197-205.

(9) Busse JW, Bhandari M, Kulkarni AV, Tunks E. The effect of low-intensity pulsed ultrasound therapy on time to fracture healing: a meta-analysis. Can Med Assoc J. 2002; 166:437-441.

(10) Warden S. A new direction for ultrasound therapy in sports medicine. Sports Med. 2003;33:95-107.

(11) Kopakkala-Tani M, Leskinen JJ, Karjalainen HM, et al. Ultrasound stimulates proteoglycan synthesis in bovine primary chondrocytes. Biorheology. 2006;43:271-282.

(12) Cook SD, Salkeld SL, Popich-Patron LS, et al. Improved cartilage repair after treatment with low-intensity pulsed ultrasound. Clin Orthop Rel Res. 2001;(391 suppl):S231-S243.

(13) Lu H, Qin L, Fok P, et al. Low-intensity pulsed ultrasound accelerates bonetendon junction healing. Am J Sports Med. 2006;34:1287-1296.

(14) Warden S, Afin K, Beck E, DeWolf M, et al. Low-intensity pulsed ultrasound accelerates and a nonsteroidal anti-inflammatory drug delays knee ligament healing. Am J Sports Med. 2006;34:1094-1102.

(15) Warden S, Fuchs R, Kessler C, et al. Ultrasound produced by a conventional therapeutic ultrasound unit accelerates fracture repair. Plays Ther. 2006;86:1118-1127.

(16) Robertson VJ, Ward A, Jung P. The contribution of heating to tissue extensibility: a comparison of deep and superficial heating. Arch Phys Med Rehabil. 2005; 86:819-825.

(17) Lloyd JJ, Evans JA. A calibration survey of physiotherapy ultrasound equipment in North Wales. Physiotherapy. 1988;74: 56-61.

(18) Artho PA, Thyne JG, Warring BP, et al. A calibration study of therapeutic ultrasound units. Phys Ther. 2002;82:257-263.

(19) Chapman R. A Survey of Output Calibrations of Ultrasound Therapy Equipment Used in Physiotherapy Practice. Christchurch, New Zealand: Department of Health; 1985.

(20) Pye SD, Milford C. The performance of ultrasound therapy machines in the Lothian region. Ultrasound Med Biol. 1984;20:347-359.

(21) Robertson VJ. Therapeutic ultrasound: re-evaluating the evidence. Physiotherapy Singapore. 2003;6(2):28-35.

(22) Johns L, Straub S, Howard S. Analysis of effective radiating area, power, intensity, and field characteristics of ultrasound transducers. Arch Phys Med Rehabil. 2007;88:124-129.

(23) Belanger A. Therapeutic Physical Agents. Baltimore, Md: Lippincott Williams & Wilkins; 2002.

Author Response

Rita A Wong, Britta Schumann, Rose Townsend, Crystal A Phelps

We thank Robertson for her insightful commentary on our study (1) and for his documentation of the emerging potential for future uses of ultrasound (US) in physical therapist practice. We agree that recent research holds promise for the benefit of therapeutic US for bone healing (2,3) and for the benefit of diagnostic US using specific imaging technology. (4)

However, the purpose of our study was to examine current and common practices among physical therapists who had achieved the designation of Orthopaedic Certified Specialist (OCS) for using therapeutic US to manage musculoskeletal impairments. It was not the purpose of this study to examine US's untapped potential. We believe our survey design provided ample opportunity for early adopters of emerging therapeutic US applications to identify and describe these uses. Each respondent had the option of self-selecting up to 2 "other" impairments for which they would use US and answering all survey questions based on these self-selected uses. As Robertson noted, only 19 of 205 respondents identified uses of US other than the 6 researcher-selected options. No pattern of "other" uses emerged. One person identified "bone healing" as a condition for which US is used. The low number of "other" uses suggested to us that we had adequately captured the common uses of US by OCSs.

We agree that there is growing evidence supporting the benefit of low-intensity US for bone healing. However, the low-intensity pulsed ultrasound (LIPUS) machine described in the articles by Heckman et al (2) and Warden (3) is typically not used by physical therapists. The LIPUS machine utilizes parameters not available in the typical clinical model US machine, and LIPUS most typically is self-applied at home using a unit purchased by the patient. It is unclear at this time whether the skills of a physical therapist are needed to deliver this intervention specifically for bone healing. A recent animal-model study, (5) published after our study was completed, reports promising findings for the ability of US to promote bone healing using a conventional US machine. However, human subject studies still are needed to fully confirm the benefit in humans with fracture. The skills of a physical therapist would seem to be necessary if a conventional US machine is used, as the risk of injury from excessive heating is a real danger.

Our study did not inquire about OCS use of diagnostic US. In 2003, when preparing our survey instrument for distribution, we found no evidence in the literature that physical therapists were using US technology for diagnostic imaging, nor was this use identified by any of the physical therapists who helped pilot test the survey instrument for content validity related to the uses of US in the United States. Whether diagnostic US imaging will (or should) be applied by physical therapists in the future is an intriguing consideration. However, we see little evidence that this is an emerging area for consideration for physical therapist practice at this time.

Robertson identified several specific concerns for which additional clarification is helpful. We are happy to have the opportunity to clarify.

Concern was expressed that we did not include a category "promoting the repair of ligaments, tendons, cartilaginous tissues, and muscle." However, "tissue healing" was 1 of the 6 researcher-identified categories that subjects responded to throughout the survey. The survey instrument did not divide this category any further into specific tissues or specific stages of healing. We captured only the general category "tissue healing." Indeed, identification of specific tissues or specific stages of healing would have enhanced this category. The decision to use one overall category stemmed from our overriding concern that a lengthy survey instrument would negatively affect response rate. We believe the category "tissue healing" would be chosen by respondents who use US to promote soft tissue repair. Forty-seven percent of the respondents believed US was clinically important for tissue healing.

When reporting the temporal average intensity (TAI) scores for this study, we did not separate TAI scores by preferred US frequency (1 MHz or 3 MHz). However, we did examine the scores for each category in our initial analysis of data. The difference in TAI scores for respondents who chose 1-MHz versus 3-MHz frequency was no more than 0.2 W/[cm.sup.2] for any of the 12 categories (6 impairment categories, each with a superficial tissue and a deep tissue option). Thus, we aggregated scores. This finding supports Robertson's concern that respondents who use US at 3-MHz frequency may be using an excessively high TAI.

Robertson provides greater specificity about the history of US. We agree that reports of the potential for therapeutic benefit from US, based on both animal and human studies, were available well before the 1950s. However, the intent of the statement in our study, consistent with the purpose of our study, was to indicate that, by the early 1950s, US had moved from the early stages of development and testing into the mainstream of use.

Robertson suggests that one explanation for the 56% of OCSs who chose not to respond to the survey could be that clinicians who do not value US did not see the survey as useful and, therefore, did not respond. This would result in a biased sample. Although any interpretation of why individuals chose to respond or not to respond is conjecture, the organization of the survey instrument provided some guarding against such bias. The wording of the survey provided clear and easy opportunities for respondents to choose "would not use" if they did not believe US was warranted. It is our opinion that those with a strong position about the lack of usefulness of US would be just as likely to respond to the survey (expressing their opinion that US was not useful) as those with a strong opinion about a positive benefit of US. For example, respondents were asked directly about the importance they placed on US as an adjunctive modality. In response to this question (summarized in Tab. 2), 27.4% indicated they would not use US for soft tissue swelling, and 20.2% indicated they would not use US for pain management.

We undertook this study very aware of the limited evidence to either clearly support or clearly refute the use of US for many conditions for which US has historically been used. As indicated in the introduction to our study, we examined 15 systematic reviews of US spanning a variety of musculoskeletal conditions. A consistent conclusion, identified in 11 (6-16) of the 15 (6-20) systematic reviews, was that there is insufficient high-quality evidence to make an informed judgment about the clinical benefit of US. This is a very different conclusion from one that there is sufficient evidence to state that an intervention is ineffective. Each systematic review called for additional well-designed clinical studies to help answer the question of effectiveness. The statement in the discussion section of our article that indicates there are very few clinical trials supporting the clinical effectiveness of US is followed by a statement in the next sentence indicating that there also is insufficient evidence to refute the effectiveness of US. Our study did not delve into the rationale for why specific impairments and parameters were preferred. This is an area for future research. Did practitioners base their judgment on a critical assessment of the void in the literature combined with their critical reflection on the effectiveness of US on their patients? Or, rather, was their judgment based on uncritical habits and unexamined expectations of effectiveness?

We do not suggest that therapeutic US should continue to be used simply because it is currently being used by advanced practice clinicians. However, in the absence of convincing scientific evidence to either support or refute any commonly used technique, the opinion of expert practitioners is a level of evidence that should not be discounted without careful examination. This study gathered self-report information about the clinical conditions for which advanced practice clinicians use US. Future researchers should incorporate this information as one of their decision-making factors driving their prioritization of conditions most in need of intervention effectiveness studies. Well-designed clinical trials should provide the evidence to definitively answer these questions.

Evidence-based practice utilizes a hierarchy of evidence to guide clinical decisions. When insufficient scientific evidence exists, evidence from expert clinicians often is used. The judgment of the advanced practice clinicians in our study was that US is a useful adjunctive modality for several conditions. Thus, it seems reasonable that conditions for which these advanced practice clinicians use US merit further investigation. Our findings suggest that soft tissue inflammation, tissue extensibility limitations, and scar tissue remodeling are the most commonly identified conditions for which US is used and identified as clinically important. Well-designed clinical studies will help determine whether the perception of benefit is justified.

DOI: 10.2522/ptj.20050392.ar

References

(1) Wong RA, Schumann B, Townsend R, Phelps CA. A survey of therapeutic ultrasound use by physical therapists who are orthopaedic certified specialists. Phys Ther. 2007;87:986-994.

(2) Heckman JD, Ryaby JP, McCabe J, et al. Acceleration of tibial fracture-healing by non-invasive, low-intensity pulsed ultrasound. J Bone Joint Surg Am. 1994; 76:26-34.

(3) Warden SJ. A new direction for ultrasound therapy in sports medicine. Sports Med. 2003;33:95-107.

(4) Grassi W, Filippucci E, Busilacchi P. Musculoskeletal ultrasound. Best Pract Res Clin Rheumatol. 2004;18:813-826.

(5) Warden SJ, Fuchs RK, Kessler CK, et al. Ultrasound produced by a conventional therapeutic ultrasound unit accelerates fracture repair. Phys Ther. 2006;86: 1118-1127.

(6) Goodyear-Smith F, Arroll B. What can family physicians offer patients with carpal tunnel syndrome other than surgery? A systematic review of nonsurgical management. Ann Fam Med. 2004;2:267-273.

(7) Brosseau L, Casimiro L, Robinson V, et al. Therapeutic ultrasound for treating patellofemoral pain syndrome. Cochrane Database Syst Rev. 2001(4):CD003375.

(8) Casimiro L, Brosseau L, Robinson V, et al. Therapeutic ultrasound for the treatment of rheumatoid arthritis. Cochrane Database Syst Rev. 2002(3):CD003787.

(9) Gam A, Johannsen, F. Ultrasound therapy in musculoskeletal disorders: a meta analysis. Pain. 1995;63:85-91.

(10) O'Connor D, Marshall S, Massy-Westropp N. Non-surgical treatment (other than steroid injection) for carpal tunnel syndrome. Cochrane Database Syst Rev. 2003(1): CD003219.

(11) Ottawa Panel Evidence-Based Clinical Practice Guidelines for Electrotherapy and Thermotherapy Interventions in the Management of Rheumatoid Arthritis in Adults. Phys Ther. 2004;84:1016-1043.

(12) Philadelphia Panel Evidence-Based Clinical Practice Guidelines on Selected Rehabilitation Interventions for Shoulder Pain. Phys Ther. 2001;81:1719-1730.

(13) Philadelphia Panel Evidence-Based Clinical Practice Guidelines on Selected Rehabilitation Interventions for Neck Pain. Phys Ther. 2001;81:1701-1717.

(14) Philadelphia Panel Evidence-Based Clinical Practice Guidelines on Selected Rehabilitation Interventions for Knee Pain. Phys Ther. 2001;81:1675-1700.

(15) Philadelphia Panel Evidence-Based Clinical Practice Guidelines on Selected Rehabilitation Interventions for Low Back Pain. Phys Ther. 2001;81:1641-1674.

(16) van der Heijden GJ, van der Windt DA, de Winter AF. Physical therapy for patients with soft tissue shoulder disorders: a systematic review of randomized clinical trials. Br Med J (Clin Res Ed). 1997;315:25-30.

(17) Robertson VJ, Baker KG. A review of therapeutic ultrasound: effectiveness studies. Phys Ther. 2001;81:1339-1350.

(18) van der Windt DA, van der Heijden GJ, van den Berg SG, et al. Ultrasound therapy for musculoskeletal disorders: a systematic review. Pain. 1999;81:257-271.

(19) van Der Windt DA, van der Heijden GJ, van den Berg SG, et al. Ultrasound therapy for acute ankle sprains. Cochrane Database Syst Rev. 2002(1):CD001250.

(20) Welch V, Brosseau L, Peterson J, et al. Therapeutic ultrasound for osteoarthritis of the knee. Cochrane Database Syst Rev. 2001(3):CD003132.

All authors provided concept/idea/research design, writing, and data collection and analysis. Dr Wong provided project management.

This study was approved by the Marymount University Institutional Review Board for the Use of Human Subjects in Research.

This article was received December 19, 2005, and was accepted March 15, 2007.

DOI: 10.2522/ptj.20050392

References

(1) Warden SJ, McMeeken JM. Ultrasound usage and dosage in sports physiotherapy. Ultrasound Med Biol. 2002;28: 1075-1080.

(2) ter Haar G, Dyson M, Oakley EM. The use of ultrasound by physiotherapists in Britain, 1985. Ultrasound Med Biol. 1987; 13: 659-663.

(3) Robertson VJ, Baker KG. A review of therapeutic ultrasound: effectiveness studies. Phys Ther. 2001;81:1339-1350.

(4) Robinson AJ, Snyder-Mackler, L. Clinical application of electrotherapeutic modalities. Phys Ther. 1988;68:1235-1238.

(5) Lindsay DM, Dearness J, McGinley CC. Electrotherapy usage trends in private physiotherapy practice in Alberta. Physiother Can. Winter 1995;47:30-34.

(6) Roebroeck ME, Dekker J, Oostendorp RA. The use of therapeutic ultrasound by physical therapists in Dutch primary health care. Phys Ther. 1998;78:470-478.

(7) Speed CA. Therapeutic ultrasound in soft tissue lesions. Rheumatology. 2001;40: 1331-1336.

(8) Gersten JW. Ultrasonics and muscle disease. Am J Phys Med. 1954;33:68-74.

(9) Lehmann JF, Erickson DJ, Martin GM, Krusen FH. Comparison of ultrasonic and microwave diathermy in the physical treatment of periarthritis of the shoulder: study of the effects of ultrasonic and microwave diathermy when employed in conjunction with massage and exercise. Arch Phys Med Rehabil. 1954;35:627-634.

(10) Lehmann JF, De Lateur B. Therapeutic heat. In: Lehmann JF, ed. Therapeutic Heat and Cold. 4th ed. Baltimore, Md: Williams & Wilkins; 1990:504-581.

(11) Dyson M, Pond JB. The effect of pulsed ultrasound on tissue regeneration. Physiotherapy. 1970;56:136-142.

(12) Dyson M, Pond JB, Joseph J, Warwick R. The stimulation of tissue regeneration by means of ultrasound. Clin Sci. 1968;35: 273-285.

(13) Goodyear-Smith F, Arroll B. What can family physicians offer patients with carpal tunnel syndrome other than surgery? A systematic review of nonsurgical management. Ann Fam Med. 2004;2:267-273.

(14) Brosseau L, Casimiro L, Robinson V, et al. Therapeutic ultrasound for treating patello-femoral pain syndrome. Cochrane Database Syst Rev. 2001;(4):CD003375.

(15) Casimiro L, Brosseau L, Robinson V, et al. Therapeutic ultrasound for the treatment of rheumatoid arthritis. Cochrane Database Syst Rev. 2002;(3):CD003787.

(16) Gam AN, Johannsen F. Ultrasound therapy in musculoskeletal disorders: a meta analysis. Pain. 1995;63:85-91.

(17) O'Connor D, Marshall S, Massy-Westropp N. Non-surgical treatment (other than steroid injection) for carpal tunnel syndrome. Cochrane Database Syst Rev. 2003; (1):CD003219.

(18) Ottawa Panel Evidence-Based Clinical Practice Guidelines for Electrotherapy and Thermotherapy Interventions in the Management of Rheumatoid Arthritis in Adults. Phys Ther. 2004;84:1016-1043.

(19) Philadelphia Panel Evidence-Based Clinical Practice Guidelines on Selected Rehabilitation Interventions for Shoulder Pain. Phys Ther. 2001;81:1719-1730.

(20) Philadelphia Panel Evidence-Based Clinical Practice Guidelines on Selected Rehabilitation Interventions for Neck Pain. Phys Ther. 2001;81:1701-1717.

(21) Philadelphia Panel Evidence-Based Clinical Practice Guidelines on Selected Rehabilitation Interventions for Knee Pain. Phys Ther. 2001;81:1675-1700.

(22) Philadelphia Panel Evidence-Based Clinical Practice Guidelines on Selected Rehabilitation Interventions for Low Back Pain. Phys Ther. 2001;81:1641-1674.

(23) van der Heijden GJ, van der Windt DA, de Winter AF. Physiotherapy for patients with soft tissue shoulder disorders: a systematic review of randomised clinical trials. BMJ. 1997;315:25-30.

(24) van der Windt DA, van der Heijden GJ van den Berg SG, et al. Ultrasound therapy for musculoskeletal disorders: a systematic review. Pain. 1999;81:257-271.

(25) van der Windt DA, van der Heijden GJ, van den Berg SG, et al. Ultrasound therapy for acute ankle sprains. Cochrane Database Syst Rev. 2002;(2):CD001250.

(26) Welch V, Brosseau L, Peterson J, et al. Therapeutic ultrasound for osteoarthritis of the knee. Cochrane Database Syst Rev. 2001;(3):CD003132.

(27) Sparrow K. Therapeutic ultrasound. In: Michlovitz SL, Nolan TP, eds. Modalities for Therapeutic Intervention. 4th ed. Philadelphia, Pa: FA Davis Co; 2005:79-96.

(28) American Board of Physical Therapy Specialties. Orthopaedic Minimum Eligibility Criteria for Specialist Certification. Available at: http://www.apta.org/AM/Template. cfm?Section= Exams&CONTENTID= 22488& TEMPLATE =/CM/ContentDisplay.cfm. Accessed August 30, 2005.

(29) Portney LG, Watkins MP. Foundations of Clinical Research: Applications to Practice. 2nd ed. Upper Saddle River, NJ: Prentice Hall Health; 2000.

(30) Baker KG, Robertson VJ, Duck FA. A review of therapeutic ultrasound: biophysical effects. Phys Ther. 2001;81: 1351-1358.

(31) Draper DO. Don't disregard ultrasound yet--the jury is still out [letter]. Phys Ther. 2002;82:190-191.

* SPSS Inc, 233 S Wacker Dr, Chicago, IL 60606.

RA Wong, PT, EdD, is Professor and Chair, Department of Physical Therapy, Marymount University, 2807 N Glebe Rd, Arlington, VA 22207 (USA). Address all correspondence to Dr Wong at: [e-mail].

B Schumann, PT, DPT, is Clinic Director, Results Rehab & Fitness, Centreville, Va.

R Townsend, PT, MSPT, is Physical Therapist, National Children's Center-Northwest Campus, Washington, DC.

CA Phelps, PT, MSPT, is a physical therapist with Inova Health System, Ashburn, Va.

[Wong RA, Schumann B, Townsend R, Phelps CA. A survey of therapeutic ultrasound use by physical therapists who are orthopaedic certified specialists. Phys Ther. 2007;87:986-994.]

Invited Commentary

Val J Robertson

Ultrasound offers many clinical users considerable potential

Ultrasound offers many clinical users considerable potential. Many current uses of ultrasound (US) made by physical therapists, however, probably contribute little to patients' outcomes. In years to come, the findings of the survey by Wong et al (1) will provide an interesting reality check for physical therapists: When did we accept how little evidence there is that supports how physical therapists use US? The findings suggest that, instead of adjusting practice according to research evidence, many physical therapists persist with long-discredited ideas and practices that are not evidence based. It also is not apparent that they are adopting practices justified by evidence.

Wong et al (1) chronicled the self-reported uses of US by physical therapy experts for 6 impairments. The 205 contributing experts (44.9% effective return rate, with 457 physical therapists surveyed) were all orthopaedic certified specialists (OCS), and the specified impairments were "pain, soft tissue inflammation, decreased tissue extensibility, delayed tissue healing, soft tissue swelling, ... and scar tissue remodeling." The authors indicate that they chose common impairments and provided opportunities for OCS respondents to add "other" options. The most frequently reported "other" impairments that the respondents identified managing with ultrasound were muscle spasm, calcium deposits, and hematoma. Other uses of US that received only one mention each included to increase blood flow and enhance bone healing.

The article also provides a summary of the most frequently reported parameters used (Tab. 4). To summarize, a frequency of 3 MHz was used for superficial lesions and a frequency of 1 MHz was used for deep lesions (not defined). A typical intensity, apparently irrespective of frequency, was 1 to 1.5 W/[cm.sup.2]. The space-averaged, time-averaged intensity ([I.sub.SATA]) (which Wong et al (1) called the "temporal average intensity") varied from 0.02 to 3.3 W/[cm.sup.2] for superficial tissues and from 0.10 to 2.5 W/[cm.sup.2] for deep tissues. These findings suggest little understanding by the OCS respondents of the physical properties of US, including the effects of frequency on the absorption of US energy. (2(p262))

More than 50% of the OCS respondents reported US as clinically important (somewhat important, essential, or very important) for treating 3 of the 6 impairments: soft tissue inflammation, decreased tissue extensibility, and when scar tissue remodeling was needed. The absence of supporting peer-reviewed evidence for this finding suggests that practice possibly is unduly altered by reimbursement pressures or the need to meet patients' expectations of what a physical therapy treatment should include.

There are 2 ways of interpreting the findings of the survey by Wong et al. (1) One way is with some optimism that maybe the nonresponding experts (55.1% of those surveyed) recognized the contradiction in investigating uses of a modality to manage conditions for which the researchers recognized there is no evidence to support its effectiveness. The alternative is further confirmation that introducing evidence-based change to practice is very difficult. The present commentary will assume this latter interpretation: nearly 50% of experts in a relevant area reported uses that are difficult to justify on the basis of peer reviewed research evidence. With this in mind, I will focus on 3 issues: (1) how the findings of the survey by Wong et al contribute to practice development, (2) the effectiveness of therapeutic US, and (3) dosage issues.

Practice Development

There is an underlying contradiction in the study by Wong et al (1): as the authors report in the "Discussion" section, "there are very few controlled trials that have found US to be clinically effective." Despite this, the researchers investigated the use of US for 6 conditions for which the authors indicate there is little or no high-quality evidence that it improves the outcomes. (3-5) Why, when, and how is investigating a seemingly ineffective practice of using a modality such as US justified?

Investigating practices known as unlikely to be effective can be very informative. From a "big picture" perspective, results such as those reported by Wong et al (1) can provide readers with insights into whether practice is adopting uses supported by research.

Wong and colleagues' findings, (1) therefore, are quite disappointing. Their findings suggest that research published over the past 10 years has had little effect on clinical practices in physical therapy that involve using therapeutic US (as distinct from diagnostic US). The same practices appear to be continuing, apparently with few new options being adopted.

The questionnaire that Wong et al (1) used in their survey was not provided, but the data and description of the study suggest that the researchers missed an ideal opportunity to monitor the adoption of new uses of US by physical therapists, including for diagnostic imaging and for biofeedback during muscle retraining. It is difficult to accept that the study by Wong et al will contribute, as they suggested in the last sentence of their abstract, to helping "researchers prioritize needs for future research on the clinical effectiveness of US." Rather, the study suggests a need to investigate how to provide clinicians with more ways of accessing and integrating knowledge into practice. This need likely extends beyond US, electrical stimulation, and other modalities.

Effectiveness of Ultrasound

The second issue related to the study by Wong et al (1) is: When, and for what, is there evidence that applying US is effective? It is used frequently for imaging a growing list of tissues as the technology improves, for guiding needles during procedures, (6) for biofeedback, and for mixing compounds. Ignoring those and many other related uses, there also is clear evidence that US can promote healing in some types of tissues under some circumstances.

A number of articles published around the turn of the century attest to the value of using US to manage fractures and delayed union. (7-9) The outcomes and the possible mechanisms were well described in an article written for sports medicine clinicians. (10) That same article also canvassed other possible uses of US for managing soft tissue problems. They included promoting the repair of ligaments, tendons, cartilaginous tissues, and muscle. None of these conditions were included or mentioned by Wong et al. (1)

Laboratory researchers have investigated additional possibilities. A cursory search retrieved many current and relevant studies. For example, at the cellular level, pulsed US can promote proteoglycan synthesis in primary bovine chondrocytes. (11) This effect is more than just that induced by the heat produced by US. Such findings may result in future clinical uses in humans, or they may not be relevant in an in vivo context. In rabbits, osteochondral defects treated with US for 40 minutes a day improved faster than those treated for any lesser time. (12)

Surgically induced damage at the bone-to-tendon junction of the patella-patellar tendon complex of rabbits repaired better if treated with low-intensity pulsed US. (13) By week 4, those rabbits treated with US showed evidence of more advanced repairs, and by week 16, this difference was clear. At that stage, biomechanical testing demonstrated a higher failure load and greater strength in the bone-to-tendon junction in those rabbits treated with US than in the control group. It is unclear whether the control group was handled identically and sham ultrasound also was applied. This possible confound was managed in a study using rats and examining, in part, the effects of US on knee ligament healing. (14) Again, the findings showed that US applications produced, even after 2 weeks, an increased ultimate strength and stiffness in the US-treated ligaments.

These results from animal studies suggest that US holds considerable potential for clinical uses in physical therapy. Some caution is warranted despite the findings being consistent with evidence that low-intensity pulsed US promotes fracture union in humans: controlled in vivo studies in humans have not yet been reported, and the parameters and dosages that Wong et all reported physical therapists now use are very different.

Fracture repair is promoted by low-intensity pulsed US. The parameters usually used are a frequency of 1.5 MHz, a beam nonuniformity ratio of 2.16, an effective radiating area of 4 [cm.sup.2], and an [I.sub.SATA] of 0.03 W/[cm.sup.2].(2) In addition, the dosage is usually daily, 20-minute applications, using a stationary applicator. The parameters are produced by a system specifically designed and used to treat fractures (Exogen). The extent to which these parameters may be varied is not yet clear, but this dosage and system of usage raises more questions.

By definition, low-intensity pulsed US has an [I.sub.SATA] of less than 0.1 W/[cm.sup.2]. (15) Warden et al (15) investigated whether using a standard, conventional therapeutic US unit could produce the same outcome as that produced by the purpose-designed US equipment known to promote bone union. They used rats with surgically created femoral fractures. After 40 days, more healing had occurred in those rats treated with the active US than in those rats treated with the sham low-intensity pulsed US. The parameters used included a frequency of 1.0 MHz, pulsed so that the [I.sub.SATA] was 0.1 W/[cm.sup.2]; a beam nonuniformity ratio of less than 6; an effective radiating area effective radiating area of 5 [cm.sup.2]; and daily 20-minute applications 5 times a week with a stationary applicator. The machine output was tested and confirmed weekly during the study.

Questions remain: How effective is conventional equipment for managing human fractures and bone damage? To what extent can US improve the rate and extent of recovery of a range of types of tissue damage in humans? This includes ligament and cartilaginous damage, injuries for which there currently is little convincing evidence.

Another issue is, how realistic is it to expect US to improve the rate of healing of normal tissues? The conditions for which there is supporting evidence typically concern tissues that are slow to heal, where healing has been delayed for various reasons, or detailed biomechanical studies are done after sacrificing an animal. The initial stages of acute soft tissue healing normally take less than 2 weeks, and patients may not want to make return daily visits for additional weeks to get an unseen benefit. This also raises an interesting problem for clinical users of US: If a condition with a usual rate of initial repair of 14 days is speeded up by, say, 10% by using US, would the outcome measures used clinically detect this change? Similarly, with changes in the ability of ligaments to respond to lengthening, will it ever be possible clinically to demonstrate a faster rate of strength development in vivo?

Fractures are typically managed with US for 20-minute sessions once per day. Few physical therapy clinicians apply US this often and for this long to a dosage area. Given the parameters that Wong et al (1) reported as favored, this is a blessing. Skin burns of possibly considerable depth would be inevitable with either frequency of US, especially if applied with a stationary applicator.

Existing clinical and laboratory findings suggest that considerable change is needed to existing uses of US made by physical therapists. Disappointingly, only one respondent in the study by Wong et al seemed aware of the use of US for promoting bone healing (ie, from among 205 clinical experts).

Another issue with US is, when should this modality be preferred? For example, as a method of increasing tissue extensibility, when should it be preferred over an application of heat, of heat and stretching, or of stretching alone? The relevant component of US for increasing tissue extensibility is accepted as heat, but this can be provided by other means.

Shortwave diathermy (SWD) and microwave, for example, are known to be effective methods of deep heating. Even in young, uninjured people, applying SWD can increase the range of movement available in an adjacent joint. (16) The mechanism is not understood, but the effect is readily demonstrated, and SWD enables heating of a large tissue volume. By contrast, the temperature of only small volumes of tissue can be effectively heated using therapeutic US.

There is no evidence that US is effective for managing pain, and little is known about the comparative effectiveness of modalities including heat and cold for relieving different types of pain or how well they compare with different types of medication. Similarly, the relative benefits of different modalities and electrical stimulation (eg, pulsed current from a transcutaneous electrical nerve stimulation machine) are not known. Physical therapists need evidence-based guidelines for making choices from a range of specific modalities. Clinicians are best placed to investigate these types of questions, particularly experts who would usually treat a higher number of patients with relevant presenting problems.

In summary, in vitro and some in vivo research shows that US can improve some outcomes for patients. However, these uses were not those investigated by Wong et al, (1) nor did more than 0.4% (1/205) of the experts who responded indicate knowledge of the types of findings for which research-based evidence exists. This is despite a recent article (15) in this same journal reporting the possible relevance of standard therapeutic machines for managing bony damage in vitro. Neither do the self-reported types of uses made of US suggest that many of the therapists surveyed recognize that alternatives to US might be more effective. Pain and reduced tissue extensibility, for example, are possibly better managed by alternatives to US.

Dosages

The third issue to address in this commentary is dosage. Dosage problems have dogged studies of therapeutic US. One problem concerns the lack of stability of output produced by US equipment. This problem has been well documented in locations ranging from North Wales (17) to Texas (18) to New Zealand (19) and has been known for a long time (19,20) and their implications discussed often and in many places. (21) More recent evidence shows that, although the tested equipment met Food and Drug Administration regulations and was consistent with the manufacturer's advice, considerable differences could exist in the space-averaged intensity produced by transducers with the same frequency. (22) A serious omission in the survey by Wong et al (1) was a question asking clinicians how often they have their equipment tested for electrical safety and the output calibrated. The results of that question might have been very enlightening.

Ultrasound has been used clinically for nearly 70 years, (23) well prior to the 1950s as Wong et al (1) suggested. Despite that, no dose-response relationship has been identified. This makes it difficult to understand why therapists choose the dosages they use. The patterns of absorption of US at frequencies of 1 and 3 MHz are different. Dosages should reflect this difference, but they do not, suggesting that some users lack an understanding of the basic properties of US. To apply US at a frequency of 3 MHz and an intensity of approximately 1 W/[cm.sup.2] ([I.sub.SATA] range=0.023.3 W/[cm.sup.2]) over a superficial lesion seems excessive unless there is no underlying bone and the applicator is moved rapidly. Otherwise, the dosage does not appear to account for the high rate of absorption of 3-MHz US in a small volume of tissue if over a bone. (2) Even without bone directly underneath a superficial lesion, this is still a high dosage and not seemingly consistent with using US with a frequency of 1 MHz at only slightly different intensities, from 0.1 to 2.5 W/[cm.sup.2], over deep tissue. These assumptions may be incorrect, because Wong et al collapsed the 1- and 3-MHz US dosages but indicated that respondents preferred them for deep and superficial lesions, respectively. This leaves readers in a conundrum: Do respondents not know and understand the differences and modify dosages accordingly, or is this interpretation a product of how the results are presented? Hopefully, the higher intensities reported were for US applied at a frequency of 1 MHz, not 3 MHz, and a few survey responses skewed the findings unduly.

Variations in dosages reported in studies of US as used by physical therapists are always intriguing. Robertson and Baker, in an earlier systematic review of clinical uses of US, commented on this, saying, "the dosages of ultrasound used in the studies we reviewed varied considerably and for reasons that were not always clear." (5(p1348)) The findings of the survey by Wong et al imply that nothing has changed and that even experts use a very wide range of dosages. Admittedly, in this instance, users were asked to self-report, no checking was possible, and respondents had to imagine an ideal patient for each of the 6 conditions included.

Taken as a whole, the dosages reported by Wong et al (1) should concern the profession, given they suggest a limited understanding of the relevant biophysical properties of US. This is compounded by long-identified problems with US equipment. The question was not asked, so we can only assume that much of the equipment used is not necessarily in optimal condition and recently tested for electrical safety and output.

Summary

The study by Wong et al (1) raises concerns as to how well physical therapists integrate current research findings and use them to change practice. The results suggest that a considerable number of experts use practices that the researchers themselves reported as not supported by evidence. Whether this was intentional or a by-product of pressures such as for reimbursement and patient expectations is not known.

The near 50% nonresponse rate might suggest that many experts recognized that the questions asked by the researchers were only exploring uses known not to be effective rather than the effect of new findings on the profession. Perhaps, but the findings also suggest unjustifiable differences in dosages, another fallout from the lack of supporting peer-reviewed evidence and the continuing absence of a dose-response relationship.

I am still left with 2 major puzzles. First, why did Wong et al (1) investigate patterns of usage of a modality for which there is little evidence of effectiveness? If this was a test of integration of research into practice, then discussion of this issue and the findings could have been very interesting. Second, why claim to investigate the "conditions [for which US is] ... most often used" and then constrain respondents to the 6 categories with only an option of "other"? This was an ideal opportunity to identify what OCS therapists use US for and why.

The main contribution of the article by Wong et al (1) will remain a reality check for those concerned with the integration of peer-reviewed evidence into practice, and the findings provide cause for concern as this study was of clinical experts in physical therapy.

DOI: 10.2522/ptj.20050392.ic

References

(1) Wong RA, Schumann B, Townsend R, Phelps CA. A survey of therapeutic ultrasound use by physical therapists who are orthopedic certified specialists. Phys Ther. 2007;87:986-994.

(2) Robertson VJ, Ward AR, Low J, Reed A. Electrotherapy Explained. 4th ed. London, United Kingdom: Elsevier Science Ltd; 2006.

(3) Speed C. Therapeutic ultrasound in soft tissue lesions. Rheumatology. 2001;40:1331-1336.

(4) Gursel K, Ulus Y, Bilgic A, et al. Adding ultrasound in the management of soft tissue disorders of the shoulder: a randomized placebo-controlled trial. Phys Ther. 2004;84:336-343.

(5) Robertson VJ, Baker KG. A review of therapeutic ultrasound: effectiveness studies. Phys Ther. 2001;81:1339-1350.

(6) Grassi W, Filippucci E, Busilacchi P. Musculoskeletal ultrasound. Best Pract Res Clin Rheumatol. 2004;18:813-826.

(7) Nolte PA, Klein-Nulend J, Albers GH, et al. Low-intensity ultrasound stimulates endochondral ossification in vitro. J Orthop Res. 2001;19:301-307.

(8) Takikawa SC, Matsui N, Kokubu T, et al. Low-intensity pulsed ultrasound initiates bone healing in rat nonunion fracture model. J Ultrasound Med. 2001;20:197-205.

(9) Busse JW, Bhandari M, Kulkarni AV, Tunks E. The effect of low-intensity pulsed ultrasound therapy on time to fracture healing: a meta-analysis. Can Med Assoc J. 2002; 166:437-441.

(10) Warden S. A new direction for ultrasound therapy in sports medicine. Sports Med. 2003;33:95-107.

(11) Kopakkala-Tani M, Leskinen JJ, Karjalainen HM, et al. Ultrasound stimulates proteoglycan synthesis in bovine primary chondrocytes. Biorheology. 2006;43:271-282.

(12) Cook SD, Salkeld SL, Popich-Patron LS, et al. Improved cartilage repair after treatment with low-intensity pulsed ultrasound. Clin Orthop Rel Res. 2001;(391 suppl):S231-S243.

(13) Lu H, Qin L, Fok P, et al. Low-intensity pulsed ultrasound accelerates bonetendon junction healing. Am J Sports Med. 2006;34:1287-1296.

(14) Warden S, Afin K, Beck E, DeWolf M, et al. Low-intensity pulsed ultrasound accelerates and a nonsteroidal anti-inflammatory drug delays knee ligament healing. Am J Sports Med. 2006;34:1094-1102.

(15) Warden S, Fuchs R, Kessler C, et al. Ultrasound produced by a conventional therapeutic ultrasound unit accelerates fracture repair. Plays Ther. 2006;86:1118-1127.

(16) Robertson VJ, Ward A, Jung P. The contribution of heating to tissue extensibility: a comparison of deep and superficial heating. Arch Phys Med Rehabil. 2005; 86:819-825.

(17) Lloyd JJ, Evans JA. A calibration survey of physiotherapy ultrasound equipment in North Wales. Physiotherapy. 1988;74: 56-61.

(18) Artho PA, Thyne JG, Warring BP, et al. A calibration study of therapeutic ultrasound units. Phys Ther. 2002;82:257-263.

(19) Chapman R. A Survey of Output Calibrations of Ultrasound Therapy Equipment Used in Physiotherapy Practice. Christchurch, New Zealand: Department of Health; 1985.

(20) Pye SD, Milford C. The performance of ultrasound therapy machines in the Lothian region. Ultrasound Med Biol. 1984;20:347-359.

(21) Robertson VJ. Therapeutic ultrasound: re-evaluating the evidence. Physiotherapy Singapore. 2003;6(2):28-35.

(22) Johns L, Straub S, Howard S. Analysis of effective radiating area, power, intensity, and field characteristics of ultrasound transducers. Arch Phys Med Rehabil. 2007;88:124-129.

(23) Belanger A. Therapeutic Physical Agents. Baltimore, Md: Lippincott Williams & Wilkins; 2002.

VJ Robertson, PT, PhD, is Professor, Teaching & Research Unit, Gosford Hospital, Bldg 14, Corner of Holden Rd and Racecourse Rd, Gosford, New South Wales 2250, Australia. Address all correspondence to Dr Robertson at: [e-mail]

Author Response

Rita A Wong, Britta Schumann, Rose Townsend, Crystal A Phelps

We thank Robertson for her insightful commentary on our study (1) and for his documentation of the emerging potential for future uses of ultrasound (US) in physical therapist practice. We agree that recent research holds promise for the benefit of therapeutic US for bone healing (2,3) and for the benefit of diagnostic US using specific imaging technology. (4)

However, the purpose of our study was to examine current and common practices among physical therapists who had achieved the designation of Orthopaedic Certified Specialist (OCS) for using therapeutic US to manage musculoskeletal impairments. It was not the purpose of this study to examine US's untapped potential. We believe our survey design provided ample opportunity for early adopters of emerging therapeutic US applications to identify and describe these uses. Each respondent had the option of self-selecting up to 2 "other" impairments for which they would use US and answering all survey questions based on these self-selected uses. As Robertson noted, only 19 of 205 respondents identified uses of US other than the 6 researcher-selected options. No pattern of "other" uses emerged. One person identified "bone healing" as a condition for which US is used. The low number of "other" uses suggested to us that we had adequately captured the common uses of US by OCSs.

We agree that there is growing evidence supporting the benefit of low-intensity US for bone healing. However, the low-intensity pulsed ultrasound (LIPUS) machine described in the articles by Heckman et al (2) and Warden (3) is typically not used by physical therapists. The LIPUS machine utilizes parameters not available in the typical clinical model US machine, and LIPUS most typically is self-applied at home using a unit purchased by the patient. It is unclear at this time whether the skills of a physical therapist are needed to deliver this intervention specifically for bone healing. A recent animal-model study, (5) published after our study was completed, reports promising findings for the ability of US to promote bone healing using a conventional US machine. However, human subject studies still are needed to fully confirm the benefit in humans with fracture. The skills of a physical therapist would seem to be necessary if a conventional US machine is used, as the risk of injury from excessive heating is a real danger.

Our study did not inquire about OCS use of diagnostic US. In 2003, when preparing our survey instrument for distribution, we found no evidence in the literature that physical therapists were using US technology for diagnostic imaging, nor was this use identified by any of the physical therapists who helped pilot test the survey instrument for content validity related to the uses of US in the United States. Whether diagnostic US imaging will (or should) be applied by physical therapists in the future is an intriguing consideration. However, we see little evidence that this is an emerging area for consideration for physical therapist practice at this time.

Robertson identified several specific concerns for which additional clarification is helpful. We are happy to have the opportunity to clarify.

Concern was expressed that we did not include a category "promoting the repair of ligaments, tendons, cartilaginous tissues, and muscle." However, "tissue healing" was 1 of the 6 researcher-identified categories that subjects responded to throughout the survey. The survey instrument did not divide this category any further into specific tissues or specific stages of healing. We captured only the general category "tissue healing." Indeed, identification of specific tissues or specific stages of healing would have enhanced this category. The decision to use one overall category stemmed from our overriding concern that a lengthy survey instrument would negatively affect response rate. We believe the category "tissue healing" would be chosen by respondents who use US to promote soft tissue repair. Forty-seven percent of the respondents believed US was clinically important for tissue healing.

When reporting the temporal average intensity (TAI) scores for this study, we did not separate TAI scores by preferred US frequency (1 MHz or 3 MHz). However, we did examine the scores for each category in our initial analysis of data. The difference in TAI scores for respondents who chose 1-MHz versus 3-MHz frequency was no more than 0.2 W/[cm.sup.2] for any of the 12 categories (6 impairment categories, each with a superficial tissue and a deep tissue option). Thus, we aggregated scores. This finding supports Robertson's concern that respondents who use US at 3-MHz frequency may be using an excessively high TAI.

Robertson provides greater specificity about the history of US. We agree that reports of the potential for therapeutic benefit from US, based on both animal and human studies, were available well before the 1950s. However, the intent of the statement in our study, consistent with the purpose of our study, was to indicate that, by the early 1950s, US had moved from the early stages of development and testing into the mainstream of use.

Robertson suggests that one explanation for the 56% of OCSs who chose not to respond to the survey could be that clinicians who do not value US did not see the survey as useful and, therefore, did not respond. This would result in a biased sample. Although any interpretation of why individuals chose to respond or not to respond is conjecture, the organization of the survey instrument provided some guarding against such bias. The wording of the survey provided clear and easy opportunities for respondents to choose "would not use" if they did not believe US was warranted. It is our opinion that those with a strong position about the lack of usefulness of US would be just as likely to respond to the survey (expressing their opinion that US was not useful) as those with a strong opinion about a positive benefit of US. For example, respondents were asked directly about the importance they placed on US as an adjunctive modality. In response to this question (summarized in Tab. 2), 27.4% indicated they would not use US for soft tissue swelling, and 20.2% indicated they would not use US for pain management.

We undertook this study very aware of the limited evidence to either clearly support or clearly refute the use of US for many conditions for which US has historically been used. As indicated in the introduction to our study, we examined 15 systematic reviews of US spanning a variety of musculoskeletal conditions. A consistent conclusion, identified in 11 (6-16) of the 15 (6-20) systematic reviews, was that there is insufficient high-quality evidence to make an informed judgment about the clinical benefit of US. This is a very different conclusion from one that there is sufficient evidence to state that an intervention is ineffective. Each systematic review called for additional well-designed clinical studies to help answer the question of effectiveness. The statement in the discussion section of our article that indicates there are very few clinical trials supporting the clinical effectiveness of US is followed by a statement in the next sentence indicating that there also is insufficient evidence to refute the effectiveness of US. Our study did not delve into the rationale for why specific impairments and parameters were preferred. This is an area for future research. Did practitioners base their judgment on a critical assessment of the void in the literature combined with their critical reflection on the effectiveness of US on their patients? Or, rather, was their judgment based on uncritical habits and unexamined expectations of effectiveness?

We do not suggest that therapeutic US should continue to be used simply because it is currently being used by advanced practice clinicians. However, in the absence of convincing scientific evidence to either support or refute any commonly used technique, the opinion of expert practitioners is a level of evidence that should not be discounted without careful examination. This study gathered self-report information about the clinical conditions for which advanced practice clinicians use US. Future researchers should incorporate this information as one of their decision-making factors driving their prioritization of conditions most in need of intervention effectiveness studies. Well-designed clinical trials should provide the evidence to definitively answer these questions.

Evidence-based practice utilizes a hierarchy of evidence to guide clinical decisions. When insufficient scientific evidence exists, evidence from expert clinicians often is used. The judgment of the advanced practice clinicians in our study was that US is a useful adjunctive modality for several conditions. Thus, it seems reasonable that conditions for which these advanced practice clinicians use US merit further investigation. Our findings suggest that soft tissue inflammation, tissue extensibility limitations, and scar tissue remodeling are the most commonly identified conditions for which US is used and identified as clinically important. Well-designed clinical studies will help determine whether the perception of benefit is justified.

DOI: 10.2522/ptj.20050392.ar

References

(1) Wong RA, Schumann B, Townsend R, Phelps CA. A survey of therapeutic ultrasound use by physical therapists who are orthopaedic certified specialists. Phys Ther. 2007;87:986-994.

(2) Heckman JD, Ryaby JP, McCabe J, et al. Acceleration of tibial fracture-healing by non-invasive, low-intensity pulsed ultrasound. J Bone Joint Surg Am. 1994; 76:26-34.

(3) Warden SJ. A new direction for ultrasound therapy in sports medicine. Sports Med. 2003;33:95-107.

(4) Grassi W, Filippucci E, Busilacchi P. Musculoskeletal ultrasound. Best Pract Res Clin Rheumatol. 2004;18:813-826.

(5) Warden SJ, Fuchs RK, Kessler CK, et al. Ultrasound produced by a conventional therapeutic ultrasound unit accelerates fracture repair. Phys Ther. 2006;86: 1118-1127.

(6) Goodyear-Smith F, Arroll B. What can family physicians offer patients with carpal tunnel syndrome other than surgery? A systematic review of nonsurgical management. Ann Fam Med. 2004;2:267-273.

(7) Brosseau L, Casimiro L, Robinson V, et al. Therapeutic ultrasound for treating patellofemoral pain syndrome. Cochrane Database Syst Rev. 2001(4):CD003375.

(8) Casimiro L, Brosseau L, Robinson V, et al. Therapeutic ultrasound for the treatment of rheumatoid arthritis. Cochrane Database Syst Rev. 2002(3):CD003787.

(9) Gam A, Johannsen, F. Ultrasound therapy in musculoskeletal disorders: a meta analysis. Pain. 1995;63:85-91.

(10) O'Connor D, Marshall S, Massy-Westropp N. Non-surgical treatment (other than steroid injection) for carpal tunnel syndrome. Cochrane Database Syst Rev. 2003(1): CD003219.

(11) Ottawa Panel Evidence-Based Clinical Practice Guidelines for Electrotherapy and Thermotherapy Interventions in the Management of Rheumatoid Arthritis in Adults. Phys Ther. 2004;84:1016-1043.

(12) Philadelphia Panel Evidence-Based Clinical Practice Guidelines on Selected Rehabilitation Interventions for Shoulder Pain. Phys Ther. 2001;81:1719-1730.

(13) Philadelphia Panel Evidence-Based Clinical Practice Guidelines on Selected Rehabilitation Interventions for Neck Pain. Phys Ther. 2001;81:1701-1717.

(14) Philadelphia Panel Evidence-Based Clinical Practice Guidelines on Selected Rehabilitation Interventions for Knee Pain. Phys Ther. 2001;81:1675-1700.

(15) Philadelphia Panel Evidence-Based Clinical Practice Guidelines on Selected Rehabilitation Interventions for Low Back Pain. Phys Ther. 2001;81:1641-1674.

(16) van der Heijden GJ, van der Windt DA, de Winter AF. Physical therapy for patients with soft tissue shoulder disorders: a systematic review of randomized clinical trials. Br Med J (Clin Res Ed). 1997;315:25-30.

(17) Robertson VJ, Baker KG. A review of therapeutic ultrasound: effectiveness studies. Phys Ther. 2001;81:1339-1350.

(18) van der Windt DA, van der Heijden GJ, van den Berg SG, et al. Ultrasound therapy for musculoskeletal disorders: a systematic review. Pain. 1999;81:257-271.

(19) van Der Windt DA, van der Heijden GJ, van den Berg SG, et al. Ultrasound therapy for acute ankle sprains. Cochrane Database Syst Rev. 2002(1):CD001250.

(20) Welch V, Brosseau L, Peterson J, et al. Therapeutic ultrasound for osteoarthritis of the knee. Cochrane Database Syst Rev. 2001(3):CD003132.

Table 1.
Response Frequency to the Survey Question "Indicate the Percentage
of Patients With the Following Primary Impairments for Whom You
Would Include US, With or Without Phonophoresis, as an Adjunctive
Therapy" (a)

Impairment                     No. of      Likely to     Unlikely to
                               Responses   Use US (b)    Use US (c)

Pain                           203         49.3% (100)   50.7% (103)
Soft tissue inflammation       207         83.6% (173)   16.4% (34)
  (ie, tendinitis/bursitis)
Tissue extensibility           206         70.9% (146)   29.1% (60)
Tissue healing                 202         52.5% (106)   47.5% (96)
Soft tissue swelling           205         35.1% (72)    64.9% (133)
  (ie, edema/joint effusion)
Scar tissue remodeling         205         68.8% (141)   31.2% (64)

(a) Data are displayed as percentage of subjects (n).

(b) Respondents chose 1 of 4 survey responses "for about 25% of my
patients," "for about 50% of my patients," "for about 75% of my
patients," or "for more than 90% of my patients."

(c) Respondents chose the survey response: "with less than 10% of
my patients."

Table 2.
Response Frequency to the Survey Question "For Those Patients Whom You
Do Treat With Ultrasound (US), How Clinically Important Do You Believe
US Is in Achieving the Each of the Following Therapeutic Goals?"

Impairment                     Total No. of   US Is Clinically
                               Responses      Important (b)

Pain                           203            39.4% (80)
Soft tissue inflammation       203            62.6% (127)
  (ie, tendinitis/bursitis)
Tissue extensibility           204            71.1% (145)
Tissue healing                 200            47.0% (94)
Soft tissue swelling           201            27.9% (56)
  (ie, edema/joint effusion)
Scar tissue remodeling         202            53.0% (107)

Impairment                     US Is Not Clinically   I Would Not
                               Important (c)          Use US

Pain                           40.4% (82)             20.2% (41)
Soft tissue inflammation       28.6% (58)              8.9% (18)
  (ie, tendinitis/bursitis)
Tissue extensibility           22.1% (45)              6.9% (14)
Tissue healing                 38.0% (76)             15.0% (30)
Soft tissue swelling           44.8% (90)             27.4% (55)
  (ie, edema/joint effusion)
Scar tissue remodeling         36.6% (74)             10.4% (21)

(a) Data are displayed as percentage of subjects (n).

(b) Clinically important responses: "US is somewhat important,"
"US is very important," and "US is essential."

(c) Not clinically important response: "US is not important" or
"US is minimally important."

Table 3.
Correlation Between the Frequency of Use of Ultrasound (US) and the
Clinical Importance of US for Each Impairment (a)

Impairment                                           Correlation
                                                     ([r.sub.s])

Pain                                                 .68 (b)
Soft tissue inflammation (ie, tendinitis/bursitis)   .71 (b)
Tissue extensibility                                 .75 (b)
Tissue healing                                       .67 (b)
Soft tissue swelling (ie, edema/joint effusion)      .72 (b)
Scar tissue remodeling                               .69 (b)

(a) Spearman correlation for use of US was based on the scores on the
5-point ordinal scale: 1=less than 10% of patients, 2=about 25% of
patients, 3=about 50% of patients, 4=about 75% of patients, and
5=greater than 90% of patients. Spearman correlation for clinical
importance was based on the scores on the 6-point ordinal scale:
0=would not use US, 1=not important, 2=minimally important,
3=somewhat important, 4=very important, 5=essential.

(b) Significant at P < .001.

Table 4.
Frequency Responses for Preferred Ultrasound (US) Parameters

  US Parameter                Impairment

                              Pain          Soft Tissue
                                            Inflammation
Duty cycle (a)

  [less than or equal          7.9% (11)     26.5% (44)
    to] 20% pulse
  50% pulse                   17.1% (24)     44.0% (73)
  100% pulse                  75.0% (105)    29.5% (49)
Frequency: superficial
    tissues (a)
  1 MHz                       14.0% (17)     10.1% (15)
  3 MHz                       86.0% (104)    89.9% (133)
Frequency: deep tissues (a)
  1 MHz                       95.0% (113)    94.0% (140)
  3 MHz                        5.0% (6)       6.0% (9)
Average pulse intensity:
  superficial tissues (a)
  <1 W/[cm.sup.2]             22.3% (23)     26.9% (32)
  1 [less than or equal       71.8% (74)     66.4% (79)
    to] 2 W/[cm.sup.2]
  >2 W/[cm.sup.2]              5.8% (6)       6.7% (8)
Average pulse intensity:
  deep tissues (a)
  < 1 W/[cm.sup.2]             4.0% (4)       9.8% (12)
  1 [less than or equal       88.0% (88)     82.8% (101)
    to] 2 W/[cm.sup.2]
  >2 W/[cm.sup.2]              8.0% (8)       7.4% (9)
Temporal average intensity:
  superficial tissue (b)
  Mean                          1.04           0.68
  SD                            0.59           0.50
  Median                        1.0            0.50
  Range                         0.09-3-30      0.02-2.50
  Total respondents           101           115
Temporal average intensity:
  deep tissue (b)
  Mean                          1.58           0.82
  SD                            0.51           0.55
  Median                        1.50           0.75
  Range                         0.10-2.50      0.10-2.50
  Total respondents           149           116

  US Parameter                Impairment

                              Tissue          Tissue
                              Extensibility   Extensibility
Duty cycle (a)

  [less than or equal           1.2% (2)       16.8% (21)
    to] 20% pulse
  50% pulse                     5.3% (9)       30.4% (38)
  100% pulse                   93.6% (160)     52.8% (66)
Frequency: superficial
    tissues (a)
  1 MHz                        15.8% (23)      13.5% (14)
  3 MHz                        84.2% (123)     86.5% (90)
Frequency: deep tissues (a)
  1 MHz                        93.9% (139)     95.2% (99)
  3 MHz                         6.1% (9)        4.8% (5)
Average pulse intensity:
  superficial tissues (a)
  <1 W/[cm.sup.2]               9.8% (12)      18.8% (18)
  1 [less than or equal        83.7% (103)     75.0% (72)
    to] 2 W/[cm.sup.2]
  >2 W/[cm.sup.2]               6.5% (8)        6.3% (6)
Average pulse intensity:
  deep tissues (a)
  < 1 W/[cm.sup.2]              7.2% (12)       6.2% (6)
  1 [less than or equal        82.4% (103)     87.6% (85)
    to] 2 W/[cm.sup.2]
  >2 W/[cm.sup.2]              10.4% (13)       6.2% (6)
Temporal average intensity:
  superficial tissue (b)
  Mean                           1.22            0.89
  SD                             0.49            0.49
  Median                         1.20            0.80
  Range                          0.10-3.30       0.10-2.50
  Total respondents            118              90
Temporal average intensity:
  deep tissue (b)
  Mean                           1.37            1.05
  SD                             0.44            0.55
  Median                         1.50            1.00
  Range                          0.10-2.50       0.10-2.50
  Total respondents            113              93

  US Parameter                Impairment

                              Soft           Scar Tissue
                              Tissue         Remodeling
                              Swelling
Duty cycle (a)

  [less than or equal          37.5% (36)     4.4% (7)
    to] 20% pulse
  50% pulse                    44.8% (43)    13.8% (22)
  100% pulse                   17.7% (17)    81.8% (130)
Frequency: superficial
    tissues (a)
  1 MHz                        11.1% (9)     10.4% (14)
  3 MHz                        88.9% (72)    89.6% (120)
Frequency: deep tissues (a)
  1 MHz                        93.9% (77)    94.0% (126)
  3 MHz                         6.1% (5)      6.0% (8)
Average pulse intensity:
  superficial tissues (a)
  <1 W/[cm.sup.2]              35.8% (24)    16.7% (20)
  1 [less than or equal        59.7% (40)    75.0% (90)
    to] 2 W/[cm.sup.2]
  >2 W/[cm.sup.2]               4.5% (3)      8.3% (10)
Average pulse intensity:
  deep tissues (a)
  < 1 W/[cm.sup.2]             16.9% (11)     5.8% (7)
  1 [less than or equal        80.0% (52)    80.8% (97)
    to] 2 W/[cm.sup.2]
  >2 W/[cm.sup.2]               3.1% (2)     13.3% (16)
Temporal average intensity:
  superficial tissue (b)
  Mean                           0.52          1.14
  SD                             0.40          0.57
  Median                         0.40          1.00
  Range                        0.10-1.50      0.10-3.00
  Total respondents           67            116
Temporal average intensity:
  deep tissue (b)
  Mean                           0.58          1.29
  SD                             0.43          0.50
  Median                         0.50          1.50
  Range                        0.10-2.00      0.10-2.50
  Total respondents           65            116

(a)  Data are displayed as percentage of subjects (n).

(b) Calculated as: duty cycle(%) x average pulse intensity
(W/[cm.sup.2]).

Table 5.
Percentage of Respondents (N=207) Who Indicated They Would Use
Ultrasound to Deliver Phonophoresis

Impairment                                            Use
                                                      Phonophoresis (a)

Pain                                                  22.2% (46)
Soft tissue inflammation (ie, tendinitis/bursitis)    54.1% (112)
Tissue extensibility                                   5.3% (11)
Tissue healing                                         8.2% (17)
Soft tissue swelling (ie, edema/joint effusion)       19.8% (41)
Scar tissue remodeling                                 3.9% (8)

(a) Data are displayed as percentage of subjects (n).