How Do Physical Therapists Treat People with Knee Osteoarthritis, and What Drives Their Clinical Decisions? A Population-Based Cross-Sectional Survey

Abstract

Purpose: It is unclear how physical therapists in Florida currently treat people with knee osteoarthritis and whether current best evidence is used in clinical decision making. Methods: We conducted a survey of physical therapists in Florida. We assessed the perceived effectiveness and actual use of physical therapy (PT) interventions and quantified the association between the actual use of interventions and different characteristics of physical therapists. Results: A total of 413 physical therapists completed the survey. Most respondents perceived therapeutic exercise (94%) and education (93%) as being effective or very effective. Interventions least perceived as effective or very effective were electrotherapy (28%), wedged insole (20%), and ultrasound (19%). Physical therapists who followed the principles of evidence-based practice were more likely to use therapeutic exercise (OR 3.89; 95% CI: 1.21, 12.54) and education (OR 3.63; 95% CI: 1.40, 9.43) and less likely to use ultrasound (OR 0.32; 95% CI: 0.16, 0.63) and electrotherapy (OR 0.32; 95% CI: 0.17, 0.58). Results also indicated that older physical therapists were more likely to use ultrasound (OR 3.57; 95% CI: 1.60, 7.96), electrotherapy (OR 2.53; 95% CI: 1.17, 5.47), kinesiology tape (OR 3.82; 95% CI: 1.59, 9.18), and ice (OR 1.95; 95% CI: 1.02, 3.73). Conclusions: In line with clinical guidelines, most physical therapists use therapeutic exercise and education to treat people with knee osteoarthritis. However, interventions that lack scientific support, such as electrotherapy and ultrasound, are still used. A modifiable therapist characteristic, adherence to evidence-based practice, is positively associated with the use of interventions supported by scientific evidence.

Osteoarthritis is the most common joint disease and the main cause of pain in elderly people.¹ The knee is the most commonly affected joint in the lower extremities, and knee osteoarthritis is strongly associated with physical disability.^1–3 As life expectancy increases, the prevalence of knee osteoarthritis is also expected to increase.⁴ Managing knee osteoarthritis is mainly symptom oriented because there is still no effective disease-modifying treatment.⁵

Physical therapy (PT) plays an important role in the treatment of knee osteoarthritis symptoms.⁶ Current evidence from randomized controlled trials (RCTs) has suggested that the most effective individual PT treatment modalities for knee osteoarthritis are cardiovascular or strengthening exercise and education (e.g., instructions on physical exercises and weight loss).^7–10 However, in clinical practice, physical therapists most often use a combination of two or more interventions when treating people with knee osteoarthritis.¹¹ For example, the use of exercise may be preceded by the use of physical agents such as ultrasound, diathermy, and electro analgesia, for example interferential current or transcutaneous electrical nerve stimulation.

The low representation in RCTs of treatment strategies commonly used in the clinical setting may pose a challenge to evidence-based practice.¹¹ Therefore, researchers need to know how physical therapists actually treat people with knee osteoarthritis so that they can design representative trials and identify gaps between evidence and practice and therefore understand where changes in PT education are needed. Although previous studies have been conducted in the United Kingdom,^12,13 it is unclear how physical therapists in Florida currently treat people with knee osteoarthritis and whether they use current best evidence in their clinical decision making. The purpose of this study was, therefore, to conduct a survey to understand how physical therapists treat people with knee osteoarthritis and what drives their clinical decision making.

Methods

Study design

This population-based, cross-sectional, electronic survey study was approved by the Florida International University Institutional Review Board (IRB-14–0381).

Target population

The survey was conducted in February 2015 with licensed physical therapists in Florida. No further eligibility restrictions were applied. We obtained a contact list of physical therapists from the Florida Department of Health and contacted them by means of an email that included an invitation to participate and a link to the online survey.

Data collection

We developed a survey based on information obtained from a literature review that searched for previously published related surveys, guidelines for conservative non-pharmacological knee osteoarthritis treatment, and discussions with experts in the field, including researchers who had previously conducted surveys of PT interventions among physical therapists. We pilot tested our survey with a convenience sample of 20 Doctor of Physical Therapy students from the Florida International University, who gave feedback on the content, form, and time needed to complete the survey. We included their suggestions in the final version of the survey and made minor modifications to improve flow and clarity. As a result, we considered that further pilot testing was unnecessary and that the survey questions were relevant for the objectives of our study. The survey was designed so that it took a maximum of 10 minutes to complete. A copy of the final survey is provided in the Appendix.

The survey consisted of 12 questions. The first section gathered information on general demographic characteristics and general clinical experience as well as clinical expertise specific to osteoarthritis management: gender, age, years of practice, whether currently practising, number of people with knee osteoarthritis seen per month, any postgraduate training related to osteoarthritis treatment, and familiarity with guidelines of clinical practice (GCP) for knee osteoarthritis treatment.

The second section assessed the perceived effectiveness of the following interventions for the treatment of knee osteoarthritis: manual therapy, therapeutic exercise, aquatic exercise, electrotherapy, ultrasound, ice, heat, wedge insoles, knee braces, kinesiology tape, education, and rest. Perceived effectiveness was assessed on a 4-point Likert-type scale with the following options: ineffective, somewhat effective, effective, and very effective. Because different physical therapists may have different perspectives on what constitutes an effective treatment, we presented in our survey a case vignette of a patient with painful osteoarthritis and consequent functional limitations that are considered major outcomes when assessing the effectiveness of knee osteoarthritis treatments.¹⁴ We asked physical therapists to assess the expected effectiveness of different PT interventions, taking this case vignette into consideration.

The third section of the survey concerned the frequency of actual use by physical therapists of each of the interventions listed earlier to treat people with knee osteoarthritis. Frequency of actual use of these interventions was assessed on a 4-point Likert-type scale with the following options: never, rarely, sometimes, and often. The fourth section assessed whether physical therapists use the principles of evidence-based practice to define treatment strategies. Evidence-based practice was defined as the combined use of clinical experience, patient preference, and evidence from peer-reviewed articles as the main sources of information to define treatment strategies.

We distributed the survey electronically using Qualtrics research software (Qualtrics, LLC, Provo, UT). The survey Web site outlined the research project, identified the research team, discussed the privacy of the data, offered a contact email address and telephone number if potential participants had questions, and included other information recommended by the ethics committee. By submitting a completed survey, a respondent was considered to be giving consent to participate.

Data analysis

We tabulated data on demographics and clinical expertise to describe our study population, using percentages, means, and SDs as appropriate, and plotted the results to compare perceived effectiveness of PT interventions with their actual use. We conducted multivariable logistic regression models to derive ORs with 95% CIs to quantify the association between the actual use of PT interventions, our dependent variable, and different characteristics of physical therapists, our independent variables. ORs greater than 1 imply that physical therapists with the characteristic of interest are more likely to use a specific intervention. We conducted multiple imputation to account for missing answers by using gender, age, years of clinical practice, number of patients treated with knee osteoarthritis per month, postgraduate training related to osteoarthritis treatment, and evidence-based practice as variables in the imputation model; as a result, we created 20 imputed data sets.^3,15 The α level was set at 0.05. We performed all analyses using Stata, version 14 (StataCorp LP, College Station, TX).

Results

General population demographics

Of the 13,296 physical therapists who were sent invitations by email, 413 completed the survey and were included in our analysis (3% response rate). Table 1 displays the characteristics of the physical therapists who answered the survey. Respondents were on average aged 44 years, mainly women (64%), with more than 10 years of clinical practice (66%), and they defined treatment strategies following the principles of evidence-based practice (69%). Fewer than 30% saw more than 10 people with knee osteoarthritis per month or had postgraduate training related to osteoarthritis treatment.

Table 1.

Characteristics of Physical Therapists Included in the Survey (n=413)

Characteristic	Participants, no. (%)*
Age, y, mean (SD)	43.8 (11.8)
Male	150 (36.3)
>10 y of practice	271 (65.6)
>10 patients per month	121 (29.3)
Postgraduate training	118 (28.6)
Evidence-based practice	286 (69.2)
Familiar with GCP	317 (76.8)

^*Unless otherwise indicated.

GCP=guidelines of clinical practice.

Perceived effectiveness, and use, of physical therapy interventions

Figure 1 displays the percentage of respondents who perceived interventions as being effective or very effective and the percentage of respondents who classified the actual use of the interventions when treating people with knee osteoarthritis as often. Most respondents perceived therapeutic exercise (94%) and education (93%) as effective or very effective. Interventions least perceived as effective or very effective were electrotherapy (28%), wedged insole (20%), and ultrasound (19%). Actual use of most interventions agreed with their perceived effectiveness, with most respondents using therapeutic exercise (96%) and education (94%) and the least number of respondents using kinesiology tape (9%), knee braces (8%), and wedged insoles (3%). Although 88% of respondents considered aquatic exercise to be effective or very effective, only 19% reported actually using this intervention with patients.

Figure 1.

Percentage of respondents who perceived knee osteoarthritis treatments as effective or very effective, and percentage of respondents who classified their actual use of treatments as often.

Association between physical therapists' characteristics and use of knee osteoarthritis treatments

Figure 2 shows the association between physical therapists' characteristics and use of knee osteoarthritis treatments. Male therapists were more likely to recommend rest (OR 1.84; 95% CI: 1.08, 3.14) and less likely to use kinesiology tape (OR 0.34; 95% CI: 0.14, 0.86). Older physical therapists (>60 y) were more likely to use ultrasound (OR 3.57; 95% CI 1.60, 7.96), electrotherapy (OR 2.53; 95% CI: 1.17, 5.47), kinesiology tape (OR 3.82; 95% CI: 1.59, 9.18), and ice (OR 1.95, 95% CI: 1.02, 3.73). Physical therapists who treated more than 10 people with knee osteoarthritis per month were more likely to recommend rest (OR 2.27; 95% CI: 1.31, 3.95) and ice (OR 2.10; 95% CI: 1.32, 3.34). Physical therapists with postgraduate training related to osteoarthritis treatment were more likely to use manual therapy (OR 1.99; 95% CI 1.25, 3.18). Physical therapists who follow the principles of evidence-based practice were more likely to use therapeutic exercise (OR 3.89; 95% CI: 1.21, 12.54) and education (OR 3.63; 95% CI: 1.40, 9.43) and less likely to use ultrasound (OR 0.32; 95% CI: 0.16, 0.63) and electrotherapy (OR 0.32; 95% CI: 0.17, 0.58). Physical therapists familiar with GCP for the treatment of people with knee osteoarthritis were more likely to use manual therapy (OR 1.74; 95% CI: 1.07, 2.85).

Figure 2.

Association between physical therapists' characteristics and frequent use of knee osteoarthritis treatments. *Statistically significant at p<0.05. Estimates are derived from multivariable models with frequency of use of knee osteoarthritis treatments as the dependent variable and physical therapists' characteristics as independent variables. OA=osteoarthritis; GCP=guidelines of clinical practice.

Combined use of different physical therapy interventions to treat people with knee osteoarthritis

Table 2 shows the frequency of the combined use of different interventions for the treatment of people with knee osteoarthritis. The most commonly used combination of treatments was therapeutic exercise, education, and manual therapy (10%), followed by the combination of these three therapies plus the use of ice (7%). Given the high heterogeneity of interventions used across different combinations, fewer than 1% of physical therapists used most combinations (61%). Only 3% of physical therapists reported not using combined interventions in their treatment strategies.

Table 2.

Frequency of the Combined Use of Different Interventions for Treating People with Knee Osteoarthritis

Therapeutic exercise	Education	Manual therapy	Ice	Heat	Electrotherapy	Aquatic exercise	Kinesiology tape	Rest	Ultrasound	Knee brace	Wedged insole	Frequency, %
•	•	•										9.75
•	•	•	•									6.82
•	•	•	•	•								2.69
•	•	•		•								2.20
•		•										2.00
•	•											1.96
•	•	•		•	•							1.95
•	•	•				•						1.95
•	•	•		•	•							1.69
•	•	•					•					1.48
•	•	•	•		•							1.47
•	•					•						1.21
•	•	•	•				•					1.21

Note: Of most combinations of interventions, 61.1% were used by <1% of physical therapists and are not displayed here; 2.5% of physical therapists used a single intervention.

Discussion

The results of this survey of 413 physical therapists indicate that the most commonly used PT interventions to treat people with knee osteoarthritis are therapeutic exercise and education, and ultrasound and electrotherapy are among the least used. Our results also indicate that evidence-based practice and therapist age have an important influence on how frequently some interventions are prescribed to people with knee osteoarthritis. Unsurprisingly, how physical therapists perceived the effectiveness of interventions agreed with how frequently they used them. Aquatic exercise was the main exception, however. Although 88% of the physical therapists believed that aquatic exercise was effective or very effective, only 19% reported often using it with patients; this may be explained by limited access to clinics with appropriate facilities and similar effectiveness compared with land-based therapeutic exercise.^16,17 Finally, more than 99% of the physical therapists reported using a combination of two or more interventions; the combination of therapeutic exercise, education, and manual therapy was most frequently used.

Two similar surveys were previously conducted of physical therapists from the United Kingdom. In 2006, Holden and colleagues¹² conducted a cross-sectional survey with 538 physical therapists randomly sampled from the general population. They reported that, similar to our findings, physical therapists in the United Kingdom commonly used therapeutic exercises and education to treat people with knee osteoarthritis, and they less frequently used kinesiology tape or recommended that patients rest, avoiding physical activity. This is in line with current GCP.^9,10 However, physical therapists in the United Kingdom made frequent use of ice, heat, and electrotherapy. Little evidence supports the effectiveness of these interventions (and, in fact, all of them appear to be less frequently used by U.S. physical therapists);^18–20 consequently, this leads to weaker recommendations regarding their use.⁹ The differences in approach may be accounted for by the fact that Holden and colleagues' study was conducted almost 10 years ago; practice is always evolving, and new evidence is incorporated into practice over time. In addition, in 2008, Walsh and Hurley¹³ published a cross-sectional survey of 83 managers of PT clinics in the United Kingdom. They reported that therapeutic exercise was the most commonly used treatment in people with knee osteoarthritis, and, in agreement with Holden and colleagues, they reported that U.K. physical therapists commonly used electrotherapy.

Our survey is the first to investigate the association between the characteristics of physical therapists and their clinical decision making in knee osteoarthritis treatment. Adhering to the principles of evidence-based practice and therapist age seem to be the strongest drivers of clinical decision making. Physical therapists who adhere to the principles of evidence-based practice are more likely to use therapeutic exercise and education and less likely to use ultrasound and electrotherapy. Indeed, evidence has indicated that the former interventions are effective when treating these patients and that not enough evidence supports the use of the latter interventions,^7,8,18,19 findings reflected in recommendations presented in recently published GCP to treat knee osteoarthritis.^9,10,21,22

Our results also indicate that older physical therapists are more likely to use ultrasound, electrotherapy, kinesiology tape, and ice, therapies that are not supported by the evidence, as previously mentioned. This finding may be explained by several factors. Perhaps older therapists have less interest in or lack the skills to incorporate new evidence into their practice or they are less exposed to new and up-to-date knowledge than younger therapists.^23,24

This is the first study to describe how a sample of U.S. physical therapists treats people with knee osteoarthritis. This is also the first study to investigate the influence of physical therapists' characteristics on the clinical decision making in treatment of knee osteoarthritis. The study was a population-based survey with a relatively large sample size—413 participants sampled from the general population of physical therapists in Florida. Because of the economy of the method used to recruit participants, it was possible to contact all potential participants rather than a sample.

However, our study did have a few limitations. One of the main limitations is that it included physical therapists only from Florida. Although we are not aware of evidence that physical therapists' skills vary across the United States, our results may arguably be limited to physical therapists in Florida. Another main limitation is the method used to sample physical therapists. Whether the contact information available through the Florida Department of Health was current was unclear; if it was not, it may explain the extremely low response rate of 3%. It is likely that physical therapists interested in osteoarthritis and evidence-based practice participated disproportionately; however, a comparison between physical therapists who participated in our survey and those who are members of the Florida Physical Therapy Association (FPTA) indicates that our sample may, in fact, be representative of physical therapists in Florida. FPTA physical therapists are, on average, aged 45 years, 36% are male, and 59% have more than 10 years of clinical practice (personal communication, FPTA, February 16, 2016). Physical therapists included in our survey were on average aged 44 years; 36% were male, and 66% had more than 10 years of clinical practice.

Another limitation was that we asked physical therapists about electrotherapy use, which might have lacked specificity because several interventions could be categorized as electrotherapy. Finally, using parameters to determine frequency of use may have led to a random error in distinguishing between sometimes and rarely. However, we believe that if a random error took place, it would have little influence on the results presented in Figure 1.

Our results suggest that incorporating the principles of evidence-based practice may lead physical therapists to use treatments that are supported by the best evidence available when treating people with knee osteoarthritis. What evidence-based practice training method would be most effective in encouraging physical therapists to use the best evidence available is currently unclear. Some preliminary work has found that small-group work, along with interactive and personal education, tends to be most effective.^25–27 Future studies, ideally using a randomized controlled design, should be conducted to identify which training methods are likely to result in improved clinical decisions based on the best available evidence by physical therapists of different ages and at different professional stages and eventually to determine the influence of these training methods on patients' outcomes. Our results also stress that current research on PT treatment of knee osteoarthritis does not reflect clinical practice. Although most trials investigate the effectiveness of single interventions, our results indicated that fewer than 1% of therapists use single interventions when treating people with osteoarthritis. This substantiates the need for RCTs with factorial designs to investigate the combined effect of complex PT interventions, which are postulated to be larger than the effect of single interventions.²⁸

Physical therapists in Florida mainly used therapeutic exercise and education to treat people with knee osteoarthritis, which is in line with the best evidence available and current GCP. Although less frequently used, interventions that lack scientific support, such as electrotherapy and ultrasound, are still used; however, they may lead to unnecessarily long treatment sessions, thereby wasting financial resources and possibly compromising patient compliance.²⁹ A modifiable therapist characteristic, adherence to evidence-based practice, is positively associated with the use of interventions supported by scientific evidence. Finally, because more than 99% of physical therapists reported using complex treatment interventions, future trials should address the heterogeneous nature of PT treatment for people with knee osteoarthritis.

Key Messages

What is already known on this topic

It is known that in the United Kingdom, physical therapists commonly use therapeutic exercise, education, ice, heat, and electrotherapy to treat people with knee osteoarthritis, and they less frequently use kinesiology tape or recommend that patients rest and avoid physical activities. However, how physical therapists in North America treat people with knee osteoarthritis, and what drives their clinical decision making, is unclear.

What this study adds

We learned from a sample of Florida physical therapists that people with knee osteoarthritis are mainly treated with therapeutic exercise and education. These physical therapists also still use interventions that lack scientific support, such as electrotherapy and ultrasound, although less frequently. A modifiable therapist characteristic, adherence to evidence-based practice, is positively associated with the use of interventions supported by scientific evidence.

Appendix

Survey Conducted in Qualtrics (https://www.qualtrics.com)

Intro: Thank you for taking this short survey. Please try to answer all questions, regardless of how frequently you treat patients with knee osteoarthritis. The survey has a total of 12 questions and should take approximately 5 minutes to complete.

Q1 What is your gender?

□ Female

□ Male

Q2 How old are you?

Q3 What year did you become licensed to practice physical therapy?

Q4 How many years have you practiced as a physical therapist?

□ 0

□ 1–5

□ 6–10

□ 11–15

□ 16–20

□ >20

Q5 Do you currently practice as a physical therapist?

□ Yes

□ No

Q6 Approximately how many patients with knee osteoarthritis do you treat per month?

□ 0

□ 1–5

□ 6–10

□ >10

Q7 Do you have any specific postgraduate training related to osteoarthritis treatment?

□ Yes

□ No

Q8 Please use the following case scenario to answer question 8. A 65-year-old patient with knee osteoarthritis has moderate pain (4/10 on visual analogue scale), causing moderate limitations of ADLs. The patient has no contraindications for physical therapy interventions and has good cognitive function. In your opinion, how effective would the following interventions be to treat this patient?

	Ineffective	Somewhat Effective	Effective	Very Effective
Manual Therapy	□	□	□	□
Therapeutic Exercise	□	□	□	□
Aquatic Exercise	□	□	□	□
Electrotherapy/E-Stim	□	□	□	□
Ultrasound	□	□	□	□
Ice	□	□	□	□
Heat	□	□	□	□
Wedge Insoles	□	□	□	□
Knee Brace/Wrap	□	□	□	□
Kinesio Tape	□	□	□	□
Education	□	□	□	□
Rest	□	□	□	□

Q9 How often do you use the interventions below to treat patients with knee osteoarthritis?

	Never	Rarely	Sometimes	Often
Manual Therapy	□	□	□	□
Therapeutic Exercise	□	□	□	□
Aquatic Exercise	□	□	□	□
Electrotherapy/E-Stim	□	□	□	□
Ultrasound	□	□	□	□
Ice	□	□	□	□
Heat	□	□	□	□
Wedge Insoles	□	□	□	□
Knee Brace/Wrap	□	□	□	□
Kinesio Tape	□	□	□	□
Education	□	□	□	□
Rest	□	□	□	□

Q10 What combination of interventions do you commonly use to treat patients with knee osteoarthritis?

□ Manual Therapy

□ Therapeutic Exercise

□ Aquatic Exercise

□ Electrotherapy/E-Stim

□ Ultrasound

□ Ice

□ Heat

□ Wedge Insoles

□ Knee Brace Wrap

□ Kinesio Tape

□ Education

□ Rest

□ I do not combine interventions

Q11 Rank the sources of information used to guide your clinical decision making on knee OA treatment (Top=Main Source / Bottom=Last Source). Click and drag items to reorganize in order of importance.

1. ______ Courses I attended

2. ______ Textbooks

3. ______ Peer-reviewed articles

4. ______ Peer advice

5. ______ Patient preference

6. ______ Clinical experience

Q12 Are you aware of clinical practice guidelines that you could use to help your clinical decision making?

Yes

No