Physical Therapy With Unsupervised Home Exercise Does Not Reduce Pain or Improve Function in Patients With Moderate-to-Severe Hip Osteoarthritis Pain

EBM Focus - Volume 9, Issue 23

Read the Complete EBM Focus and Earn CEU.

Reference: JAMA 2014 May 21;311(19):1987 (level 1 [likely reliable] evidence)

Symptomatic hip osteoarthritis is associated with joint pain and reduced physical function. The Osteoarthritis Research Society International (OARSI) guideline recommends referral to a physical therapist (OARSI Level IV) for patients with symptomatic hip osteoarthritis (Osteoarthritis Cartilage 2008 Feb;16(2):137). These recommendations state that physical therapy can provide initial evaluation and instruction in appropriate exercises to reduce pain and improve functional capacity, and may also help by providing assistive devices (canes or walkers) when appropriate. However, clinical data supporting physical therapy for this patient population have been lacking. A recent randomized trial compared a multimodal physical therapy regimen to a sham therapy in 102 community-dwelling patients (mean age 64 years) with radiograph-confirmed symptomatic hip osteoarthritis.

Patients were randomized to physical therapy vs. sham therapy for 10 sessions over 12 weeks and followed to 36 weeks. The physical therapy group received education and advice, manual therapy, and a walking stick if deemed appropriate, followed by 4-6 unsupervised home exercises performed 4 times weekly. The sham group received an inactive ultrasound plus inert gel followed by self-applied gel 3 times weekly during follow-up. At baseline, all patients had a hip pain score ≥ 40 mm (mean 58 mm) on a visual analog scale (range 0-100 mm, with higher scores indicating more severe pain). The mean hip physical function score at baseline was 32.4 points on Western Ontario and McMaster Universities Osteoarthritis Index hip-specific physical function subscale (range 0-68 points, with higher scores indicating worse function). Minimal clinically important differences were defined as 18 mm for the pain score and 6 points for the physical function score. The primary outcome was based on assessments at 13 weeks and secondary outcomes were based on results at 36 weeks.

The loss to follow-up was 6% at week 13 and 19% at week 36, but all patients were included in the efficacy analyses. Patients in the physical therapy group attended mean 9.6 of 10 treatment sessions and adherence to home exercise was mean 85% based on self-reported questionnaires. The mean overall hip pain score at 13 weeks was 40.1 mm with physical therapy vs. 35.2 mm with sham therapy. The 95% confidence interval for the between-group difference ranged from 3.9 mm in favor of physical therapy to 17.7 mm in favor of sham therapy. Similarly, the mean hip physical function score at 13 weeks was 27.5 points with physical therapy vs. 26.4 points with sham therapy. The 95% confidence interval for the between-group difference ranged from 3.8 points in favor of physical therapy to 6.5 points in favor of sham therapy. The frequency of mild adverse events, primarily increased hip pain or pain in other regions, by 13 weeks was 41% with physical therapy vs. 14% with sham therapy (p = 0.003, NNH 3). The results at 36 weeks also did not show significant differences between the 2 groups for either hip pain or hip physical function scores. The proportion of patients believing they were receiving active treatment was 62% in the physical therapy group vs. 25% in the sham group at week 36. The cost of the multimodal physical therapy intervention evaluated was not reported.

Physical therapy for patients with symptomatic hip osteoarthritis often employs a multimodal approach, including education, manual therapy, and various exercises. Although some previous studies have evaluated some of the specific components involved, this is the first study to rigorously evaluate several aspects of physical therapy using a credible sham control, which potentially reduces the risk of bias in outcomes such as pain scores that are subjective in nature. Because of this, these data provide the best evidence to date for the effect of a broad set of physical therapy components, most importantly a combination of manual therapy and unsupervised home exercise, on hip pain or function. The findings suggest that physical therapy including manual therapy and unsupervised home exercise is of limited value for these patients, and may even harm patients by increasing the frequency of mild adverse events such as increased hip pain or pain in other regions. Although a higher proportion of patients in the physical therapy group were aware that they received active treatment at 36 weeks, this is unlikely to have made a substantial impact on either the pain or physical function outcomes.

A recent update of a Cochrane review identified 10 randomized trials comparing land-based exercise (aquatic exercise is reviewed separately) vs. nonexercise active control or no treatment in 554 adults with hip osteoarthritis (Cochrane Database Syst Rev 2014 Apr 22;(4):CD007912). The inclusion criteria allowed both supervised exercise as well as home exercise with minimal monitoring. In a meta-analysis of 9 trials, exercise was associated with a statistically significant but clinically unimportant improvement in pain. Similarly, a separate meta-analysis of 9 trials showed that exercise was associated with a statistically significant improvement in physical function, with a 95% confidence interval including both clinically important and unimportant differences.

The term physical therapy is an umbrella term that may refer to many different interventions, and this makes it difficult to make general conclusions about its use in patients with symptomatic hip osteoarthritis. In particular, the use of unsupervised home exercise evaluated in this study may be a limitation, since therapists had limited ability to ensure that exercises were being done correctly. Furthermore, although adherence to home exercise was reported to be about 85% in patients receiving physical therapy, this was by necessity a self-reported estimate among unsupervised patients. Nonetheless, the evidence from this trial does not support physical therapy as an effective treatment in this clinical context.

For more information see the Degenerative joint disease of the hip topic in DynaMed.