Cost-effectiveness of Physical Therapy vs Intra-articular Glucocorticoid Injection for Knee Osteoarthritis: A Secondary Analysis From a Randomized Clinical Trial

Abstract

Importance: Physical therapy and glucocorticoid injections are initial treatment options for knee osteoarthritis, but available data indicate that most patients receive one or the other, suggesting they may be competing interventions. The initial cost difference for treatment can be substantial, with physical therapy often being more expensive at the outset, and cost-effectiveness analysis can aid patients and clinicians in making decisions.

Objective: To investigate the incremental cost-effectiveness between physical therapy and intra-articular glucocorticoid injection as initial treatment strategies for knee osteoarthritis.

Design, setting, and participants: This economic evaluation is a secondary analysis of a randomized clinical trial performed from October 1, 2012, to May 4, 2017. Health economists were blinded to study outcomes and treatment allocation. A randomized sample of patients seen in primary care and physical therapy clinics with a radiographically confirmed diagnosis of knee osteoarthritis were evaluated from the clinical trial with 96.2% follow-up at 1 year.

Interventions: Physical therapy or glucocorticoid injection.

Main outcomes and measures: The main outcome was incremental cost-effectiveness between 2 alternative treatments. Acceptability curves of bootstrapped incremental cost-effectiveness ratios (ICERs) were used to identify the proportion of ICERs under the specific willingness-to-pay level ($50 000-$100 000). Health care system costs (total and knee related) and health-related quality-of-life based on quality-adjusted life-years (QALYs) were obtained.

Results: A total of 156 participants (mean [SD] age, 56.1 [8.7] years; 81 [51.9%] male) were randomized 1:1 and followed up for 1 year. Mean (SD) 1-year knee-related medical costs were $2113 ($4224) in the glucocorticoid injection group and $2131 ($1015) in the physical therapy group. The mean difference in QALY significantly favored physical therapy at 1 year (0.076; 95% CI, 0.02-0.126; P = .003). Physical therapy was the more cost-effective intervention, with an ICER of $8103 for knee-related medical costs, with a 99.2% probability that results fall below the willingness-to-pay threshold of $100 000.

Conclusions and relevance: A course of physical therapy was cost-effective compared with a course of glucocorticoid injections for patients with knee osteoarthritis. These results suggest that, although the initial cost of delivering physical therapy may be higher than an initial course of glucocorticoid injections, 1-year total knee-related costs are equivalent, and greater improvement in QALYs may justify the initial higher costs.

Trial registration: ClinicalTrials.gov Identifier: NCT01427153.

Conflict of interest statement

Conflict of Interest Disclosures: None reported.

Figures

Figure 1.. Cost-effectiveness Planes and Acceptability Curves for Knee-Related Costs

Filled circles indicate the incremental cost-effectiveness ratios, with the solid blue line showing the mean incremental cost-effectiveness ratio. Dotted lines indicate 95% CIs. WTP indicates willingness to pay; QALY, quality-adjusted life-year.

Figure 2.. Cost-effectiveness Planes and Acceptability Curves for Total Medical Costs (Medical Care for Any Reason)

Filled circles indicate the incremental cost-effectiveness ratios, with the solid blue line showing the mean incremental cost-effectiveness ratio. Dotted lines indicate 95% CIs. WTP indicates willingness to pay; QALY, quality-adjusted life-year.

Figure 3.. Monetary Net Benefit (MNB) Threshold

Dotted lines indicate 95% CIs. WTP indicates willingness to pay.