A systematic search of PUBMED and the Cochrane databases was conducted to identify English language publications evaluating the efficacy and/or safety of non-pharmacologic therapies in the management of OA. Each non-pharmacologic treatment modality was searched separately using medical subject heading (MESH) terms and key words. Non-pharmacologic therapies were those included in the 2008 OARSI guidelines for the management of hip and knee OA¹²⁰: acupuncture; manual therapy; physical therapy; devices (including orthoses, orthotics, footwear, walking aids); education and self-management; weight loss; transcutaneous electrostimulation or Transcutaneous electrical nerve stimulation (TENS); thermal modalities; nutraceuticals; and surgery and rehabilitation. In addition, the Cochrane database was searched for all reviews on OA over the same time period. Articles were included for consideration if they addressed management of OA using any of the non-pharmacologic therapies. Eligible papers were reviewed for quality (e.g., randomization, blinding and treatment concealment, intent-to-treat analysis approach). Studies selected for inclusion in the review presentation were those deemed by the author to be of high quality with potential to affect the management of OA in clinical practice.

There has been conflicting evidence regarding the efficacy of traditional Chinese acupuncture (TCA) in the management of painful knee OA. Variability in results has been attributed in particular to inadequate blinding to treatment, largely due to the use of non-penetrating needling procedures vs superficial but penetrating needling at non-acupuncture (non-meridian) sites²². Further, there have been concerns raised that the effects of placebo may be enhanced by patients’ expectations of treatment benefits, which may be influenced by the treatment provider.

Suarez-Almazor and colleagues conducted a nested, two-stage randomized controlled trial comparing the efficacy of TCA to that of a sham procedure for knee OA; nested within this trial, they evaluated for a moderating effect of acupuncturists’ style of communication (high or neutral expectations). Study subjects were randomized to one of three groups: (1) waiting list controls; (2) high expectations; and (3) neutral expectations. Individuals randomized to groups (2) or (3) were then re-randomized to receive either TCA or a sham procedure (sham). The sham acupuncture was comprised of shallow needling with minimal stimulation at non-meridian points; both sham and TCA were delivered using TENS equipment. Each of six participating acupuncturists was trained to deliver either a ‘high expectations’ or ‘neutral expectations’ style of communication; for the former, acupuncturists were trained to use language such as “I think this will work for you”, while the latter used language such as “It really depends on the patient”. Half way through the trial, the acupuncturists were re-trained to deliver the other style of communication. Acupuncturists received rigorous training both in their communication style as well as their performance of the intervention treatment. Multiple strategies were used to ensure blinding to allocation. Adequacy of blinding was assessed prior to recruitment in volunteers and then at study end. Using an intent-to-treat analysis, the investigators evaluated the treatment arms difference in change OA pain from baseline to 12 weeks based on the following measures: the joint-specific multidimensional assessment of pain (J-MAP), which evaluates pain intensity, frequency and quality, and the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) pain scale. Patient satisfaction with their treatment outcome was also assessed, using a five point Likert scale.

Of 560 subject initially randomized, 527 received a sufficient number of treatments and were included in the analyses (455 received acupuncture – TCA or sham; 72 wait list controls). The groups were comparable with respect to baseline characteristics and dropout rates. At 12 weeks post intervention, no differences were found between those who received TCA vs sham. However, compared with the wait list controls, both treatment groups experienced a significant improvement in pain scores (J-MAP −1.1/7, 95% confidence interval −1.0 to −0.1, P<0.0001; WOMAC pain −13.7/100, 95% confidence interval −14 to −1.7, P<0.001). Compared with the neutral expectations group, the high expectations group had significantly greater improvements in pain and greater treatment satisfaction, but the effect sizes were small (0.25 and 0.22, respectively, at 12 weeks).

Transcutaneous electrical nerve stimulation (TENS), interferential current stimulation, and pulsed electrostimulation are widely used to control both acute and chronic pain arising from many conditions, including OA. However, there remains conflicting results regarding the effectiveness of these treatments.

Rutjes and colleagues¹⁰¹ conducted a systematic Cochrane review of transcutaneous electrostimulation vs sham or no specific intervention on pain, and withdrawal due to adverse events, in individuals with knee OA.

The review identified 18 small trials including a total of 813 subjects with painful knee OA (11 TENS, four interferential current stimulation, one TENS and interferential current stimulation, and two pulsed electrostimulation). Overall, the quality of these trials was poor; there was a high degree of heterogeneity across the studies, limiting comparison across trials. The calculated predicted standardized mean difference for pain intensity was −0.07 (95% confidence interval −0.46 to 0.32), which corresponds to a difference in pain scores between electrostimulation and control of 0.2 cm on a 10 cm visual analog scale for pain.

Tai Chi is a traditional mind–body exercise that enhances balance, strength, flexibility, and self-efficacy, while reducing pain, depression and anxiety in diverse groups with chronic conditions. Tai Chi may be useful in the management of painful OA; the physical component provides exercise consistent with recommendations for OA, while the mental component may address chronic OA pain through its effects of psychological well-being.

Wang and colleagues¹³ conducted a single-blinded randomized, controlled trial of Tai Chi in subjects with symptomatic tibiofemoral knee OA (confirmed radiographically). Subjects were randomized to receive 60 min of Tai Chi vs nutrition education and stretching exercises (the attention control group) twice per week for 12 weeks. Usual OA medications were unchanged. The primary outcome of interest was the difference in change in pain (WOMAC pain score) from baseline to 12 weeks, using an intent-to-treat approach. Secondary outcomes included physical functioning (WOMAC function score; timed chair stand), patient global assessment, depression, self-efficacy and quality of life, using valid and reliable measures for each. WOMAC subscales were scored such that each item within the scale had a maximum score of 100; thus the pain subscale scores ranged from 0 to 500, and the function subscale score from 0 to 1700. To reduce the potential for bias due to lack of blinding to treatment allocation, participants were told that the study objective was to compare two types of exercise, one of which was combined with education.

Forty patients were randomized; their mean age was 65 years, with a mean BMI of 30. Compared with the attention control group, those who received Tai Chi experienced greater improvements in WOMAC pain scores (between group mean difference −118.80, 95% confidence interval −183.66 to −53.94, P=0.0005)(Fig. 1) and all secondary outcome measures. No serious adverse events were reported and there were no study withdrawals. Correction for baseline differences in severity of OA disability (which was higher in the controls), did not change the results.

ACL tears represent a serious knee injury that results in joint instability which leads in turn to reduced physical activity and lower knee-related quality of life. ACL tear is often a precursor for knee OA. Surgical ACL reconstruction is commonly performed, with or without rehabilitation by a physical therapist, to manage this injury. However, high quality evidence from randomized, controlled trials supporting this approach is lacking.

Frobell and colleagues⁹⁴ conducted a randomized, controlled clinical trial in young (18–35 years old), active adults with acute ACL tear (within 4 weeks of injury, confirmed using magnetic resonance imaging). Subjects were randomized to receive structured rehabilitation plus early ACL reconstruction (within 10 weeks) vs structured rehabilitation with optional delayed reconstruction in the presence of symptomatic knee instability (defined as self-reported symptoms of instability with a positive pivot shift test). The primary outcome of interest was the absolute change in the Knee Injury and Osteoarthritis Outcome Score₄, KOOS₄, at 24 months. The KOOS₄ incorporates the average scores for each of four KOOS subscales: pain, symptoms, function in sports and recreation, and knee-related quality of life. Supervised rehabilitations, delivered by nine therapists at nine out-patient clinics, targeted goals for range of motion, muscle function, and functional performance. ACL reconstruction was performed by four senior, high volume knee surgeons who used one of two procedures previously shown to have similar results. Additional meniscal surgery was performed at the surgeons’ discretion as needed.

Of 121 subjects randomized, 62 received rehabilitation plus early ACL reconstruction and 59 received rehabilitation with optional delayed ACL reconstruction. Among those who received optional delayed ACL reconstruction, 23 underwent subsequent ACL reconstruction and 36 did not. Subjects in the early surgery group received a greater number of rehabilitation visits (63 vs 53, P=0.05); no other group differences were found. One subject in the early reconstruction group was excluded from the primary analysis due to missing data at interim visits. At 24 months follow-up, both groups had improved significantly; the absolute between group difference in KOOS₄ scores was 0.18 (95% confidence interval −6.5 to 6.8, P=0.96). In a post-hoc as-treated analysis that compared 24-month outcomes in those assigned to early reconstruction, those who received later reconstruction (5.5–19 months post-randomization) and those who received only structured rehabilitation, no significant by-group differences were observed (Fig. 2)

            .