Physical Therapy versus Glucocorticoid Injection for Osteoarthritis of the Knee
Deyle GD, Allen C, Allison S, et al. Physical Therapy versus Glucocorticoid Injection for Osteoarthritis of the Knee. The New England Journal of Medicine. 2020;382(15):1420-1429. doi:10.1056/nejmoa1905877Link to the original article: https://www.nejm.org/doi/full/10.1056/NEJMoa1905877
Key Points
- This study compared the effectiveness of physical therapy versus glucocorticoid injections for the treatment of osteoarthritis of the knee.
- The results showed that patients who underwent physical therapy had less pain and functional disability at 1 year compared to those who received glucocorticoid injections.
- The study concluded that physical therapy was more effective than glucocorticoid injections in improving outcomes for patients with osteoarthritis of the knee, suggesting that physical therapy should be considered as a primary treatment option.
Introduction
The management of osteoarthritis of the knee is limited to treatment of symptoms until a knee replacement is needed. A controversy surrounds the injection of glucocorticoid for relief of symptoms despite being used commonly as a primary treatment for osteoarthritis of the knee. Some of the complications include joint infection, accelerated degradation of articular cartilage, and subchondral insufficiency. Clinical Practice Guidelines (CPG) vary in recommendations regarding the use of glucocorticoid injections for osteoarthritis of the knee, but the latest (CPG) strongly recommends glucocorticoid injection. Physical therapy is also suggested for the treatment of osteoarthritis of the knee, but the utilization of physical therapy has decreased over years. Currently, CPGs do not recommend using physical therapy and glucocorticoid injections together. A previous study shows that glucocorticoid injection and physical therapy do not provide further benefits. In this research, a comparison was made for the effectiveness of glucocorticoid injection with that of physical therapy for patients with osteoarthritis of the knee.
Methods
Patients
The study focused on patients within the Military Health System who were active-duty or retired service members or their family members. The eligible patients were 38 years of age or older and were seen at two military hospitals between October 2012 and May 2017. The patients received treatment at various clinics, including physical therapy, rheumatology, and orthopedics. In order to be included in the study, patients had to meet the American College of Rheumatology clinical classification for knee osteoarthritis and show radiographic evidence of osteoarthritis with a Kellgren-Lawrence grade ranging from 1 to 4. Patients who had received glucocorticoid injections, undergone physical therapy for knee pain in the previous 12 months, or had no radiographic evidence of osteoarthritis were excluded from the study. More detailed inclusion and exclusion criteria can be found in the study protocol.
Trial Oversight and Procedure
The study was approved by the institutional review board and the authors confirmed the reliability of the data and adherence to the trial protocol, as well as the comprehensive reporting of adverse events. Patients were informed about the trial during their primary care or physical therapy visits, and research coordinators obtained written informed consent and facilitated their participation. Before randomization, demographic information and baseline measures were collected, and patients were provided with education on the relationship between knee osteoarthritis and physical activity, nutrition, and obesity. The patients were assigned to either physical therapy or glucocorticoid injection in a 1:1 ratio using a random number generator. Outcome assessments were performed by research assistants who were unaware of the treatment group assignments, and steps were taken to ensure blinding was maintained throughout the data collection process. Treatment was administered to both knees if necessary, but outcomes were assessed only in the knee with worse symptoms at baseline.
Glucocorticoid Injections
The research paper describes a study in which orthopedists or rheumatologists administered glucocorticoid injections to patients with knee pain. The injections consisted of 1 ml of triamcinolone acetonide and 7 ml of lidocaine, following sterile technique. Patients were examined by the same providers at 4 months and 9 months to assess their condition and determine if additional injections were needed. The study allowed for a maximum of three injections over the course of one year, as determined by the clinician. This information highlights the methodology and protocol used in the study to administer and monitor the effects of glucocorticoid injections in patients with knee pain.
Physical Therapy
In this research paper, the authors describe a physical therapy intervention for patients with knee osteoarthritis. The intervention involved a combination of exercises, joint mobilizations, and clinical reasoning to address pain and stiffness in the knee joints. During the sessions, the physical therapist performed hands-on techniques to reduce stiffness and pain before the patient performed reinforcing exercises. For example, if the patient had difficulty extending or flexing the knee, the therapist would use passive mobilization techniques to reduce stiffness and then the patient would perform active knee movements. Manual muscle stretching was also performed if the muscles around the knee were tight. Patients underwent up to 8 treatment sessions over a 4-to-6-week period, with the possibility of additional sessions at later reassessments. The physical therapists involved in the study were board certified and fellowship-trained in orthopedic manual physical therapy. This intervention has been shown to improve knee extension in patients with osteoarthritis.
Assessments and Outcomes
The paper assessed outcomes for pain, physical function, and global assessment in a clinical trial for osteoarthritis. The primary outcome measure was the total WOMAC score at 1 year, which is a scale ranging from 0 to 240. Secondary outcome measures included the Global Rating of Change scale, the 1-year cost of knee-related healthcare utilization, and the results of two functional tasks. The minimal clinically important difference for the total WOMAC score was reported to be a 12% or 16% improvement from baseline. A score of +3 or higher on the Global Rating of Change scale was considered clinically meaningful. No published minimal clinically important difference was available for the Alternate Step Test, while estimates for the Timed Up and Go test ranged from 0.8 to 1.2 seconds. Healthcare utilization data was obtained from a database capturing outpatient and inpatient visits. No formal cost-effectiveness analysis was conducted, but descriptive cost values for each group were provided.
Adverse Events
The research paper investigated adverse events in a clinical trial. Aside from serious events such as death, infection, and fracture, adverse events were defined as persistent worsening of symptoms leading to additional treatment outside the trial. Patients were regularly asked to report any complications or symptoms they perceived as adverse outcomes related to their treatment. The study also included examination of claims data from the Military Health System Data Repository to validate reported additional care, including emergency department visits. Overall, the study aimed to identify and document adverse events and their associated treatment outside the trial setting.
Statistical Analysis
The study aimed to determine the required sample size to detect differences in WOMAC scores between treatment groups over time. A sample size of 138 patients was calculated to provide 80% power to detect a 12 percentage point difference in mean WOMAC scores at the first post-treatment assessment. The analysis of the data was performed using the intention-to-treat approach, with a linear mixed-effects model initially planned. However, due to significant positive skewness in the score distributions, a log-linear mixed-effects model was used instead. Analyses were reported as least-squares means and 95% confidence intervals, with no adjustments for multiple comparisons. Missing data were handled using the Markov chain Monte Carlo method. Categorical outcomes were analyzed using contingency tables. Mean costs were compared between groups using a generalized linear model. The statistical analysis plan and software version used for analyses were also specified. Data were missing for 1.4% of all values and 7% of primary and secondary outcomes.
Results
Patients
The study enrolled a total of 156 patients with osteoarthritis of the knee, with an average age of 56.1 years and 48% of them being women. The primary reasons for excluding patients from the study were unwillingness to receive a glucocorticoid injection and having received one in the previous 12 months. A total of 78 patients were randomly assigned to each group, with patients in the glucocorticoid injection group receiving an average of 2.6 injections and patients in the physical therapy group attending an average of 11.8 treatment visits. Baseline characteristics were similar between the two groups, except for the severity of osteoarthritis measured using the Kellgren-Lawrence scale, where more patients in the physical therapy group had a higher grade. Some patients in each group received additional care outside of the trial, and four patients in the glucocorticoid group underwent knee surgery. The overall cost of knee-related medical care during the trial period was similar between the two groups.
Primary Outcome
The study compared the effectiveness of glucocorticoid injections versus physical therapy in patients with osteoarthritis. The primary outcome measure was the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) score at 1 year. The mean WOMAC score was significantly higher in the glucocorticoid injection group compared to the physical therapy group (mean difference of 18.8 points; 95% CI, 5.0 to 32.6; P = 0.008). In a prespecified analysis, a smaller proportion of patients in the physical therapy group (10.3%) did not achieve at least a 12% improvement from baseline in the WOMAC score at 1 year compared to the glucocorticoid injection group (25.6%). The overall direction of results for the primary outcome remained consistent in various sensitivity analyses. The study suggests that physical therapy may be more effective than glucocorticoid injections in improving WOMAC scores in patients with osteoarthritis.
Secondary Outcomes
The study compared two treatment options, physical therapy and glucocorticoid injection, for patients with patellofemoral pain syndrome. At the one-year follow-up, patients in the physical therapy group reported a median score of +5 on the Global Rating of Change scale, indicating a significant improvement. In contrast, the glucocorticoid injection group had a median score of +4, indicating a moderate improvement. A lower proportion of patients in the physical therapy group (14.1%) did not reach a score of +3 or higher on the scale compared to the glucocorticoid injection group (33.3%). The physical therapy group also demonstrated better performance on two physical tests, the Alternate Step Test and the Timed Up and Go Test, compared to the glucocorticoid injection group. However, it should be noted that these results were not part of the prespecified analysis. Only one adverse event, fainting during an injection, was reported in the glucocorticoid group.
Discussion
This trial compared the effectiveness of physical therapy and glucocorticoid injections in patients with symptomatic osteoarthritis in one or both knees. The results showed that physical therapy was more effective than glucocorticoid injections in improving outcomes at 1 year, as assessed by the total WOMAC score. Secondary outcomes, including functional tasks and patient assessment of improvement, also favored physical therapy. Both groups reported perceived improvement, but a higher proportion of patients in the glucocorticoid group reported no improvement or worsening symptoms. The trial also found that physical therapy had a greater reduction from baseline in the mean WOMAC score at 1 year compared to previous studies with shorter treatment duration. The trial acknowledged limitations such as additional interventions received by some patients, different severity levels in each group, and inability to generalize concurrent use of both treatments. In conclusion, physical therapy resulted in better pain and physical function scores than glucocorticoid injection at 1 year.
Opportunities for Future Research
1. Investigate the long-term effectiveness of physical therapy and glucocorticoid injections for osteoarthritis of the knee, beyond 1 year.
2. Examine the cost-effectiveness of physical therapy compared to glucocorticoid injections for the treatment of osteoarthritis of the knee.
3. Compare the effectiveness of physical therapy and glucocorticoid injections in different subgroups of patients (e.g., different age groups, severity of osteoarthritis).
4. Explore the optimal dosage and frequency of glucocorticoid injections for the treatment of osteoarthritis of the knee.
5. Investigate the mechanisms by which physical therapy and glucocorticoid injections exert their effects on pain and function in patients with osteoarthritis of the knee.
6. Assess the long-term safety and potential adverse effects of glucocorticoid injections for the treatment of osteoarthritis of the knee.
7. Investigate the role of other non-surgical interventions, such as exercise therapy or acupuncture, in the management of osteoarthritis of the knee.
8. Evaluate the impact of combining physical therapy with other interventions, such as medication management or weight loss programs, for the treatment of osteoarthritis of the knee.
9. Examine the patient preferences and factors influencing treatment choice between physical therapy and glucocorticoid injections for osteoarthritis of the knee.
