How does surgery compare to sham surgery or physiotherapy as a treatment for tendinopathy? A systematic review of randomised trials

Key points

1. 0112 eligible RCTs with 1,051 participants covering shoulder, lateral elbow, patellar, and Achilles tendinopathies were included
2. 02Surgery was not superior to sham surgery for pain, function, or range of movement at midterm or long-term follow-up
3. 03Physiotherapy (tendon loading exercises) matched surgery for pain, function, range of movement, and quality of life at both midterm and long-term follow-up
4. 04Surgery was superior to no treatment or placebo, but this benefit is likely partly explained by surgical placebo effect and postoperative rehabilitation
5. 05Most included studies were of moderate or poor quality; only 2 of 12 were rated good quality

How it was conducted

Design

Systematic review of randomised controlled trials (PRISMA-compliant)

Databases searched

EMBASE, Medline, CINAHL, and Scopus (searched March 2018)

Included studies

12 RCTs published 1993-2018

Participants

1,051 participants (1,056 affected tendons); mean age 48.0 years (range 18-72); all had chronic tendinopathy (symptoms >3 months)

Tendinopathies covered

Shoulder (7 studies, n=876 tendons), lateral elbow (3 studies, n=122), patellar (1 study, n=40), Achilles (1 study, n=20)

Comparisons

Surgery vs. no treatment/placebo (2 studies), surgery vs. sham surgery (2 studies), surgery vs. physiotherapy (6 studies), surgery vs. ESWT (2 studies), surgery vs. botox or polidocanol (2 studies)

What they found

• Surgery vs. sham surgery (moderate evidence): no statistically significant differences in pain, function, or range of movement at midterm or long-term follow-up in 2 good-quality RCTs (shoulder tendinopathy and lateral elbow tendinopathy)
• Surgery vs. physiotherapy for pain (moderate evidence, midterm): no significant difference across studies of shoulder and patellar tendinopathy
• Surgery vs. physiotherapy for function (moderate evidence, midterm): no significant difference; Haahr et al. reported no differences in Constant score (primary outcome) or sub-scores over 12 months
• Surgery vs. physiotherapy for treatment success (moderate evidence, long-term): no significant difference across studies
• Surgery vs. physiotherapy for quality of life (moderate evidence, long-term): no significant difference in Ketola et al. (n=140, 5-year follow-up using 15D tool)
• Surgery vs. no treatment/placebo (moderate evidence, midterm): surgery superior for pain, function, range of movement, treatment success, and quality of life in Beard et al. (n=313) and Brox et al. (n=125)
• In Bahr et al. (patellar tendinopathy, n=35): VISA score improved significantly in both surgery and physiotherapy groups over time with no intergroup differences at any follow-up stage (3, 6, or 12 months)
• In Ketola et al. (shoulder, n=140): no differences in VAS pain (primary outcome), disability, night pain, or proportion pain-free at 2- and 5-year follow-up between surgery+physiotherapy and physiotherapy alone
• 6 of 12 studies were poor quality, 4 moderate quality, and only 2 good quality by combined internal validity, external validity, and precision assessment

Limitations

• Most evidence was limited or conflicting (Level 3) and only moderate (Level 2) for the key comparisons; no strong (Level 1) evidence was available
• Different tendinopathy sites were pooled for some comparisons, which may not be appropriate despite shared pathophysiology
• Wide variation in physiotherapy regimes and postoperative rehabilitation protocols across studies makes it impossible to determine whether improvement after surgery reflects the operation itself or the rehabilitation
• Meta-analysis was not possible due to heterogeneity in outcome measures across studies

Why it matters

For patients

Patients with chronic tendinopathy should feel reassured that a structured exercise program for at least 12 months is likely to produce the same long-term relief as surgery, without the risks of an operation.

For clinicians

Clinicians should recommend evidence-based loading exercise for at least 12 months before considering surgical referral, as surgery offers no additional benefit over physiotherapy and no benefit over sham surgery.

For readers

This review challenges the use of surgery as a second-line treatment for tendinopathy and highlights the need for larger, well-designed sham-controlled trials before surgery is routinely offered after short-course conservative management.