Long-term outcome of local steroid injections versus surgery in carpal tunnel syndrome

Key points

1. 01This was a long-term observational extension of an open randomized trial of injection versus surgery for new-onset carpal tunnel syndrome, with a mean follow-up of 6.3 years (median 5.9).
2. 02Therapeutic failure (needing any new treatment on that wrist) reached 41.8% in the injection group versus 11.6% in the surgery group.
3. 03Injection carried about a 4.5 times higher risk of treatment failure than surgery (95% CI 2.1-9.8, P < .0001).
4. 04Surgery failures occurred mostly in the first year, while injection failures kept accumulating year after year.
5. 05Baseline nerve-test severity (mild, moderate, or severe) did not reliably predict who would fail either treatment.

How it was conducted

Design

Observational long-term extension of a prospective, open, randomized clinical trial (injection versus surgery), original trial run 1998-2001

Participants

Patients with newly diagnosed primary carpal tunnel syndrome, randomized by wrist; 163 randomized wrists, 148 available at final follow-up

Groups

Decompressive surgery (limited palmar incision, one surgeon) versus local corticosteroid injection (paramethasone acetonide 20 mg in 1 mL, one investigator)

Primary outcome

Therapeutic failure, defined as need for any new intervention (surgery, injection, physiotherapy, acupuncture, etc.) on the involved wrist

Analysis

Cox multiple regression; cumulative incidence estimated with withdrawal as a competing risk (Gooley's test)

Follow-up

Mean 6.3 years, median 5.9 years (SD 2.3); range 6.1 to 9.8 years among those contacted

What they found

• Cumulative incidence of therapeutic failure at long-term follow-up was 11.6% in the surgery group versus 41.8% in the injection group.
• Cox multiple regression showed a risk of failure for the injection group of 4.5 (95% CI 2.1-9.8; P < .0001).
• Surgery group cumulative failure rose to 11.6% by 2 years (95% CI 6.0-22.2) and stayed flat through 7 years; injection group rose from 15.2% at 1 year to 41.8% by 6 years (95% CI 32.2-54.2).
• In severe baseline nerve testing, injection failure reached 64.7% (95% CI 45.5-91.9) versus 15.8% (95% CI 5.6-44.6) for surgery at 6 years.
• About 58% of patients in the injection group did not need any further therapeutic intervention during follow-up.

Limitations

• Long-term outcome data were collected observationally after the trial ended, and the final follow-up was done by telephone rather than clinical or electrophysiological exam.
• After the trial, patients received whatever treatment their general practitioner or specialist prescribed, so follow-up care was not standardized.
• Failure was defined broadly as any new intervention, which may not equal symptom recurrence or objective worsening.
• Subgroups by baseline severity were small, giving wide confidence intervals and limited power to detect prognostic differences.

Why it matters

For patients

If you have carpal tunnel syndrome, surgery is more likely to be a one-time fix, but a steroid injection still works long-term for most people and is a sensible first try.

For clinicians

Counsel patients that injection is as effective early on but carries roughly a 4.5 times higher long-term failure rate than surgery, and that baseline nerve-test severity should not be used to choose between them.

For readers

A rare multi-year extension of a randomized carpal tunnel trial showing surgery is more durable while injection still spares most patients further treatment.