Knee Osteoarthritis: The Emerging Role of Physical Therapy

Abstract & Commentary

Synopsis: Using a randomized trial, an eight-week course of manual therapy and exercise improved walking distances and WOMAC scores at four and eight weeks. These differences continued at one year, by which time 20% of the placebo group (subtherapeutic ultrasound) compared to only 5% of patients treated with rehabilitation had undergone total knee replacement.

Source: Deyle GD, et al. Effectiveness of manual physical therapy and exercise in osteoarthritis of the knee. A randomized, controlled trial. Ann Intern Med 2000;132:173-181.

Knee osteoarthritis (koa) commonly affects more than 10% of adults, including 33% of patients between the ages of 63 and 94. In a recent review, more than 90% of KOA was treated successfully with nonoperative means, avoiding surgery altogether.1 Many medical options exist in the treatment of KOA, including acetaminophen, nonsteroidal anti-inflammatory agents (NSAIDs), cortisone and hyaluronan injections, and alternatives such as glucosamine and chondroitin sulfate. Use of NSAIDs is commonplace, as are the risks of gastrointestinal (GI) complications in the population older than 60 years of age. Mechanical changes to treat KOA include shoe wear modifications, unloader bracing, and weight loss. Physical therapy has also been considered to be an effective means to treat KOA, including active and passive range of motion exercises, fitness walking, aerobic exercise, and strength training. Deyle and colleagues examined the effectiveness of manual physical therapy and exercise against placebo in the conservative management of KOA.

In an institutional review board (IRB) approved study, Deyle et al created two randomization arms. Forty-two patients (average age, 60 ± 10 years) were assigned to the treatment group and 41 patients (average age, 62 ± 10 years) received a placebo. Demographics including gender were similar for both groups. Exclusion criteria involved having received a cortisone injection within 30 days and patients were instructed to continue taking routine prescription or over-the-counter NSAIDs. Treatment patients received manual physical therapy and supervised exercise. Placebo patients received subtherapeutic ultrasound. Radiographs were obtained on all patients and were graded by a radiologist. Dependent variables included the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) score and a timed six-minute walk test. Manual therapy treatment techniques included passive physiologic and accessory joint movements, muscle stretching, and soft tissue mobilization. Other involved joints, including the lumbar spine, hip, and ankle, were also treated if symptomatic. Progressive knee exercises including strength training were performed over eight weeks. Patients exercised in a painless or minimally painful manner. Exercises were decreased with joint irritation. The placebo group received 10 minutes of ultrasound (subtherapeutic) with similar subjective and objective hands-on evaluation during treatment. The treatment group required an extra 30-45 minutes over placebo to complete each treatment. A home program with an instruction sheet was used for the treatment group during off days. Eight clinic visits (weekly) were performed for both treatment and control groups.

Of the 83 patients enrolled, 69 completed all clinic visits and testing at four weeks, eight weeks, and one year. The average timed walk was better for the treatment group (4 weeks: 82 m better, 8 weeks: 78 m better, P < 0.05) as compared to control. Similarly, WOMAC scores were 416 mm better at four weeks and 472 mm better at eight weeks (P < 0.05) as compared to placebo. At one year, 20% of the placebo group had received a total knee replacement (TKR) vs. 5% of the treatment group (P < 0.039). The placebo group also received more cortisone injections. Walking and WOMAC scores persisted at one year in those patients not undergoing a TKR or injection.

Comment by Robert C. Schenck, Jr., MD

The benefits of physical therapy have been shown in the elderly population in previous studies. In this randomized trial Deyle et al have now shown effectiveness in a general population of patients with KOA. Most impressive is the significant decrease in total joint surgery required in the treatment group. As in any treatment plan, reproducibility is key. Deyle et al include exercise descriptions and tables for the use of their program outside San Antonio, Texas. When applying such a treatment plan, the clinician should educate the treating therapist (and in this reviewer’s case vice versa) on the program and study. (Single reprints can be obtained via the Internet at The combination of clinic- and home-based physical therapy should not be underestimated. The use of patient education, home use materials, and exercise sheets allow for progression of rehabilitation through the week and not just during the single clinic visit. This home-based component may have contributed to the 20-30% improvement rate seen after only two to three visits. Home-based therapy programs have great benefit to patients, including convenience and safety.2

Unfortunately, Deyle et al did not include a cost analysis of their treatment. Depending upon the clinic, physical therapy visits can cost from $50-$150 per visit, not including the initial evaluation. The use of NSAIDs was not controlled and use of a standard medication could be a future consideration for such studies. Furthermore, the cost of both prescription NSAIDs and the cost of treating NSAID complications are also significant. The comparison of two treatment arms, NSAID vs. PT vs. placebo, is an excellent future research topic. The importance of time spent between therapist and patient was acknowledged as different. Could time alone be partially responsible for the improved success? I admit it is unlikely, but time spent with patients is a confusing variable. Lastly, the inability to double blind the study raises questions that in reality can’t be answered due to the nature of the process.

The conservative management of KOA has many angles and avenues for success. Frequently, patients come to the clinician having already tried alternatives (glucosamine/chondroitin sulfate) and some form of over-the-counter NSAIDs. Prescription NSAIDs (especially new COX-2 inhibitors) also have a place in the treatment of OA. Nonetheless, the functional focus (increasing strength, increasing range of motion) of a well-defined "clinic-based/home-extended" physical therapy program is a significant finding by Deyle et al. The actual decrease in total joint arthroplasty in the treatment group as compared to placebo gives concrete reasons for including physical therapy programs in the conservative management of KOA.


1. Buckwalter JA, et al. New approaches to the treatment of osteoarthritis. Instructional Course Lecture, American Academy of Orthopaedic Surgeons. 2000;49:491-494.

2. Schenck RC Jr, et al. A prospective outcome study of rehabilitation programs and anterior cruciate ligament reconstruction. Arthroscopy 1997;13:285-290.

In treating knee osteoarthritis, a physical therapy program was found to:

a. not affect patient outcome.

b. be more expensive than the group treated with NSAIDs.

c. be less effective than treatment with ultrasound.

d. reduce the need for total knee replacement at one year.

e. show only short-term benefit.