Patients with shoulder pain receiving corticosteroid injections (CSI) versus manual physical therapy (MPT) had the same amount of symptom improvement, but the corticosteroid group used more healthcare resources, researchers reported in a study published in Annals of Internal Medicine on Aug. 4.
Patients assigned to receive injections had more physician visits and more additional injections compared with patients receiving manual therapy over the course of 1 year, wrote Daniel Rhon, DPT, and colleagues from Madigan Army Medical and the University of Puget Sound in Tacoma, Wash., and Franklin Pierce University in Manchester, N.H.
“Considering that aversion to injection was the main reason that patients chose not to participate in the study, manual physical therapy may serve as an effective low-risk alternative,” wrote the researchers.
Current accepted approaches to treat shoulder pain include corticosteroid injections, physical therapy, and exercise.
These new findings challenge those in a previous trial looking at patients with shoulder pain that found a corticosteroid injection more cost-effective than physiotherapy, wrote Brooke Coombes, PhD, and Bill Vicenzino, PhD, of the University of Queensland in Brisbane, Australia, in an accompanying editorial.
Although healthcare costs were not measured directly, the study provides insight into the cost relative to the benefit of the two approaches, Coombes and Vicenzino said.
Several study limitations may prevent the findings from changing management just yet.
One limitation was the lack of diagnosis in the patients who presented. Shoulder impingement syndrome (SIS) includes both mechanical and synovial pathology and each responds differently to treatment.
For mechanical problems, MPT is usually a better treatment choice, while synovial problems are more likely to respond to anti-inflammatory injections, author Robert Boyles, DPT, of the University of Puget Sound in Tacoma, Wash., told MedPage Today.
“SIS is typically not a cookie-cutter treatment plan. If there is shoulder stiffness or a weak shoulder, there is more of a role for physical therapy. If it’s just pain, then it could be bursitis and we would treat with cortisone,” said David Geier, MD, an orthopedic surgeon in Charleston, S.C.
But too many cortisone injections can weaken the rotator cuff tendon, leading to tears or recurrent pain, warned Geier.
Geier also wondered about the recurrence rate in future years depending on treatment. “Physical therapy can fix the underlying problem that leads to SIS in the first place,” he said.
Study Design
Researchers followed 104 patients with unilateral shoulder impingement syndrome over 1 year in the study’s single-blind, randomized, controlled trial.
Shoulder impingement syndrome includes many pathologies, such as subacromial bursitis, rotator cuff tendinopathy, and partial rotator cuff tear.
The study recruited patients ages 18 to 65 years with a primary symptom of unilateral shoulder pain who were referred from family practice or orthopedic clinics for physical therapy. The patients included a mix of active-duty and retired military service members and their families.
Exclusion criteria included history of shoulder dislocation, fracture, or adhesive capsulitis, history of CSI or MPT for the shoulder pain in the past 3 months, history of systemic or neurologic disease affecting the shoulder, history of full-thickness rotator cuff tear, a Shoulder Pain and Disability Index (SPADI) score of less than 20%, and inability to attend physical therapy for 3 consecutive weeks.
Patients were allowed to continue any current medications prescribed by their primary care providers.
Two physical therapists provided the MPT, and one physician administered all the injections.
Patients did not receive identical MPT treatments, but the techniques were matched to individual impairments identified on exam. They consisted of joint and soft-tissue mobilizations, manual stretches, contract-relax techniques, and reinforcing exercises directed to the shoulder girdle or thoracic or cervical spine.
Patients in the MPT group were treated twice weekly over a 3-week period, by the same physical therapy in most cases. Home exercises were prescribed as well.
Patients in the CSI group received an injection of 40 mg of triamcinolone acetonide to the subacromial space of the symptomatic shoulder. As many as three total injections, greater than one month apart, could be given by the study physician over the course of 1 year.
The CSI group patients were discouraged from seeking additional care for the first month.
Patients in either group were able to see their primary care provider for additional care as necessary. The primary care providers could manage the patient as he or she thought best, which included a CSI or referral to physical therapy.
Outcomes were measured at baseline, 1 month, 3 months, 6 months, and one year. They were measured on three self-assessed scales:

  • SPADI (Shoulder Pain and Disability Index): pain and disability
  • GRC (Global Rating of Change): perceived changes in the participant’s quality of life
  • NPRS (Numeric Pain Rating Scale): pain intensity

Researchers also tracked shoulder-related visits to physical therapists, primary care providers, rheumatologists, and orthopedists, as well as frequency and types of procedures that included additional steroid injections, magnetic resonance imaging, and radiography.
All patients in the CSI group received at least one injection. Twenty patients (38%) had more than one injection.
All but six patients in the MPT group (88%) received the six physical therapy treatments.
At one year, improvement from baseline in all three outcomes was measured in both groups.
Greater than 50% improvement was seen in the 100-point SPADI from baseline in both groups. A clinically important difference for the SPADI was a 6% to 10% change.
Self-perceived improvement on the GRC was three points (CI 2-4) on a 15-point scale for both groups. Three points was considered clinically meaningful.
Self-reported pain on the 11-point NPRS improved by 1.7 points for the MPT group and 0.8 groups for the CSI group. Two points was considered the minimal clinically important difference.
At 1 year, there was not a significant between-group difference (SPADI: 1.5%, 95% CI minus 6.3%-9.4%; GRC: 0, 95% CI minus 2-1; NPRS: 0.4, 95% CI minus 0.5-1.2) in any of the outcomes.
Moreover, there were no significant between-group differences in any of the outcomes at any time point.
Regardless of treatment, patients had a significant improvement from baseline to 1 month (>50% on the SPADI scale).
Patients in the MPT group made fewer visits to their primary care provider for pain: 37% of patients in the MPT group versus 60% in the CSI had at least one additional visit (risk ratio 0.64, 95% CI 0.43-0.95).
The MPT group also had fewer additional CSIs than the CSI group: 20% of MPT patients compared with 38% of CSI patients had at least one additional injection (risk ratio 0.77, 95% CI 0.59-0.99).
Ten CSI patients (19%) sought physical therapy, and four MPT patients (9%) sought additional physical therapy.