What is the evidence for symptom reduction and functional improvements after corticosteroid injections in patients with carpal tunnel syndrome?
AMSTAR Rating
Summary
Patient Population:
A total of 14 randomized control trials (RCTs) were included in this meta-analysis, but only nine RCTs with a total of 639 participants had data usable for quantitative analysis. All patients included were over the age of 18 and diagnosed with carpal tunnel syndrome (CTS).
Intervention:
All study active treatment groups received local corticosteroid injections (LCI) into or near the carpal tunnel, with or without local anesthetic (LA).
Comparison:
Study control groups received no treatment, a placebo/sham saline injection, LA injection, or splinting.
Outcome:
- Improvement in symptoms at ≤ 3 months was assessed with the visual analogue scale (VAS) and Boston Carpal Tunnel Questionnaire (BCTQ). With a moderate certainty rating, there was greater improvement in symptoms with LCI than control groups (SMD -0.77; 95% CI: -0.94 to –0.59).
- Subgroup analysis of the dose of LCI used demonstrated no evidence of a difference between the low dosages (~20mg equivalent of methylprednisolone), medium dosages (~40mg equivalent of methylprednisolone), and high dose dosages (~80mg equivalent of methylprednisolone).
- Subgroup analysis of the duration of LCI action demonstrated no evidence of difference between intermediate-acting LCI lasting 12 to 36 hours, and long-acting LCI lasting more than 48 hours.
- Subgroup analysis of the type of corticosteroid used demonstrated no evidence of a difference between mineralocorticoid-acting LCI and non-mineralocorticoid-acting LCI.
- Improvement in symptoms at > 3 months was assessed by the VAS and BCTQ. With a moderate certainty rating, there was greater improvement in symptoms with LCI than control groups (SMD –0.58; 95%B CI: -0.89 to –0.28). There was also a mean improvement in BCTQ symptom severity scores favouring LCI of –0.24 points (95% CI: -0.395 to –0.09) over controls.
- Subgroup analysis of the dose of LCI used demonstrated greater improvement in the LCI group using high dosages (~80mg equivalent of methylprednisolone) compared to low (~20mg equivalent of methylprednisolone) and medium dosages (~40mg equivalent of methylprednisolone) at six months.
- Improvement in function at ≤ 3 months was assessed with the BCTQ and Disabilities of the Arm, Shoulder, and Hand (DASH). With a moderate certainty rating, there was an improvement in function favouring LCI over control (SMD –0.62; 95% CI: -0.87 to –0.38).
- Subgroup analysis of the dose of LCI used demonstrated no evidence of a difference between the low dosages (~20mg equivalent of methylprednisolone), medium dosages (~40mg equivalent of methylprednisolone), and high dose dosages (~80mg equivalent of methylprednisolone).
- Subgroup analysis of the duration of LCI action demonstrated no evidence of difference between intermediate-acting LCI lasting 12 to 36 hours, and long-acting LCI lasting more than 48 hours.
- Subgroup analysis of the type of corticosteroid used demonstrated no evidence of a difference between mineralocorticoid-acting LCI and non-mineralocorticoid-acting LCI.
- Improvement in neurophysiological parameters at ≤ 3 months was assessed with median nerve distal motor latency (DML). With a very low certainty rating, the risk difference with LCI was insignificantly higher than control groups (MD –0.37ms; 95% CI: -0.75 to 0.02).
- Subgroup analysis of the dose of LCI used demonstrated no evidence of a difference between the low dosages (~20mg equivalent of methylprednisolone), medium dosages (~40mg equivalent of methylprednisolone), and high dose dosages (~80mg equivalent of methylprednisolone).
- Subgroup analysis of the duration of LCI action demonstrated no evidence of difference between intermediate-acting LCI lasting 12 to 36 hours, and long-acting LCI lasting more than 48 hours.
- Subgroup analysis demonstrated greater improvement when using mineralocorticoid-acting LCI compared to non-mineralocorticoid-acting LCI (MD for mineralocorticoid-acting −0.95 ms, 95% CI −1.08 to −0.82 compared with MD non-mineralocorticoid-acting −0.24 ms, 95% CI −0.43 to −0.05).
- Requirement for carpal tunnel surgery at one year had a risk ratio (RR) of 0.84 favouring LCI over control, with a moderate certainty rating (RR 0.84; 95% CI: 0.72 to 0.98).
- Improvement in quality of life was assessed with the Short-Form Six-Dimension Instrument (SF6D). With a moderate certainty rating, the mean quality of life score improved in LCI groups more than in controls (MD 0.07; 95% CI: 0.02 to 0.12).
- Adverse events were uncommon but included pain and localized swelling for two weeks in both LCI (65%) and placebo (16%) groups. One study reported 2/364 injections caused severe pain, and 1/364 injections caused a sympathetic reaction that resolved in 20 minutes.
Guideline Recommendations
| Source | Recommendation |
|---|---|
| AAOS Clinical Practice Guideline (2024) | STRONG evidence that injection DOES NOT provide long-term improvement |
Outcomes Assessed
- Benefit
- Harm
- Inconclusive
BENEFITS
CTS Symptoms (up to 3 months)
Function (up to 3 months)
Neurophysiology parameters (up to 3 months)
Q of L (up to 3 months)
Reduced need for surgery at 1 year
HARMS
Pain (up to 2 weeks)
Localized swelling (up to 2 weeks)
Rare onset of severe pain
INCONCLUSIVE
Steroid type and dosage
Relevant Clinical Info
Atroshi I, Flondell M, Hofer M, Ranstam J. Methylprednisolone injections for the carpal tunnel syndrome: a randomized, placebo-controlled trial. Ann Intern Med. 2013 Sep 3;159(5):309-17. doi: 10.7326/0003-4819-159-5-201309030-00004. PMID: 24026316.
Atroshi et al., (2013) was the largest RCT included in the meta-analysis with a total of 111 participants (Male=30, Female=81, average age of 46.7 years). Evaluated to have a low risk of bias, this study split participants into three groups. Group 1 received 2mL of 80 mg of methylprednisolone plus 1mL of lidocaine, group 2 received 1mL of 40 mg of methylprednisolone plus 1mL of lidocaine, and group 3 received a placebo injection of 2mL saline plus 1mL lidocaine. Changes in CTS symptom severity scores were based on the short Disabilities of the Arm, Shoulder, and Hand (QuickDASH) score, short form-36 (SF-36) health survey bodily pain scores, SF-6D scores, and treatment satisfaction at 10 weeks and one year. CTS symptom severity scores at 10 weeks improved more in patients who received methylprednisolone than those who received placebo (p=0.003 for 80mg and P<0.001 for 40mg of methylprednisolone). Specifically, group 1 changed by a mean difference (MD) of -0.64 (95% CI: -1.06 to –0.21; P=0.003) compared to placebo, and group 2 changed by –0.88 (95% CI: -1.30 to –0.46; P<0.001) compared to placebo. There was no significant difference in scores between group 1 and 2 (MD=2.98; 95% CI: -0.20 to 0.69). There were no significant differences at one year. The one year rates for surgery were 73% in group 1, 81% in group 2, and 92% in group 3. Compared to those in the placebo group, those in group 1 were less likely to have surgery (odds ratio 0.24 [CI: 0.06 to 0.57]; P=0.042), but no significant difference was found between group 2 and placebo or group 1. Time from injection to surgery was also longer for group 1 (P=0.003) and group 2 (P=0.022) compared to the placebo group. At 10 weeks, both treatment groups had greater improvement than the placebo group in other outcome measures assessing symptom severity and treatment satisfaction (all P<0.025), while at 24 weeks and one year, there were no differences between treatment groups and placebo (all P>0.100). Pain after injection was reported in 24 treatment participants compared to 6 placebo participants (P<0.001). Additionally, local swelling was reported by 8 patients in treatment groups and 2 in the placebo group.
Participant Information
The sample size was 639
There were 14 studies used.
