OBJECTIVES

To assess the association between intermediate and patient-centered outcomes and harms with physical therapy interventions in community-dwelling adults with chronic knee pain secondary to osteoarthritis and to examine validity and minimum clinically important differences of the tools for outcome measurement.

DATA SOURCES

We searched major electronic bibliographic databases including MEDLINE, the Cochrane Library, the Physiotherapy Evidence Database, and Allied and Complementary Medicine and trial registries up to February 29, 2012.

REVIEW METHODS

We performed a systematic review of randomized and nonrandomized studies published in English to synthesize rates or means of measured pain, function, and quality of life with physical therapy interventions. Observational studies provided evidence of the association between changes in knee joint functional tests and patient-centered outcomes and minimum clinically important differences in validated tools for outcome measures. We performed meta-analyses of standardized mean differences using random effects models to synthesize the evidence.

RESULTS

Of 4,266 retrieved references, 154 eligible references examined the association between patient-centered and intermediate outcomes and 422 eligible references examined physical therapy interventions. Of these, 193 randomized controlled trials (RCTs) reported on knee pain, disability, quality of life, and functional outcomes after physical therapy interventions. Pooling criteria were met by 84 RCTs that provided evidence for 12 physical therapy interventions on pain (n = 58), physical function (n = 36), and disability (n = 29). Most studies reported physical therapy effects at followups of 3 months or less. Evidence on longer-term physical therapy effects was available for seven intervention-outcome pairs. Meta-analyses at the longest time of followup provided low-strength evidence that aerobic (n = 11) and aquatic exercise (n = 3) improved disability; aerobic exercise (n = 19), strengthening exercise (n = 17), and ultrasound (n = 6) reduced pain and improved function. Six of 11 individual RCTs demonstrated clinically important improvements in pain and disability with aerobic exercise. Pain relief was consistent in RCTs that reported physical therapist supervision of aerobic exercise. Diathermy, orthotics, and magnetic stimulation demonstrated no benefit. Limited direct comparative effectiveness evidence demonstrated similar benefits in disability measures with aerobic, aquatic, and strengthening exercise. Evidence from individual RCTs did not permit robust conclusions about which physical therapy interventions are most effective or whether differences in effect could be attributed to patient characteristics. Patients with high compliance to exercise tended to have better treatment responses. We found no association between the duration of examined interventions and better intermediate or patient-centered outcomes. Adverse events were uncommon and not severe enough to deter participants from continuing treatment. Gait, mobility restrictions, muscle strength, and range-of-motion measures were associated with disability measures in individual studies. Minimum clinically important differences in scales were determined for 26 tools but have not been used in RCTs to examine the clinical importance of improvements. The definition of the Patient Acceptable Symptom State that accounts for patient satisfaction was available for the Western Ontario McMaster Universities Osteoarthritis Index, the Visual Analog Scale for Pain, and the Patient Global Assessment Scale.

CONCLUSIONS

Low-strength evidence suggested that core physical therapy interventions, including aerobic, aquatic, strengthening, and proprioception exercise, improved patient outcomes. Risk of bias in studies and heterogeneity in populations and physical therapy interventions downgraded the strength of evidence to low or moderate in most cases. Studies focused on a single modality of physical therapy rather than the combinations typically used in practice. Benefits with physical therapy interventions were not consistently evaluated according to the clinical importance of improvement in scales and tests. Adverse events were uncommon and not severe enough to deter participants from continuing treatment. Evidence about long-term adherence to and benefits of available physical therapy interventions is lacking.