The Effectiveness of Corticosteroid Injection Versus Night Splints for Carpal Tunnel Syndrome

24-Month Follow-Up of a Randomized Trial

Claire Burton; Trishna Rathod-Mistry; Steven Blackburn; Milica Blagojevic-Bucknall; Linda Chesterton; Graham Davenport; Krysia Dziedzic; Adele Higginbottom; Sue Jowett; Helen Myers; Raymond Oppong; Danielle van der Windt; Elaine Hay; Edward Roddy


Rheumatology. 2023;62(2):546-554. 

In This Article

Abstract and Introduction


Objectives: This follow-up study of the INSTinCTS (INjection vs SplinTing in Carpal Tunnel Syndrome) trial compared the effects of corticosteroid injection (CSI) and night splinting (NS) for the initial management of mild-to-moderate CTS on symptoms, resource use and carpal tunnel surgery, over 24 months.

Methods: Adults with mild-to-moderate CTS were randomized 1:1 to a local corticosteroid injection or a night splint worn for 6 weeks. Outcomes at 12 and 24 months included the Boston Carpal Tunnel Questionnaire (BCTQ), hand/wrist pain intensity numeric rating scale (NRS), the number of patients referred for and undergoing CTS surgery, and healthcare utilization. A cost–utility analysis was conducted.

Results: One hundred and sixteen participants received a CSI and 118 a NS. The response rate at 24 months was 73% in the CSI arm and 71% in the NS arm. By 24 months, a greater proportion of the CSI group had been referred for (28% vs 20%) and undergone (22% vs 16%) CTS surgery compared with the NS group. There were no statistically significant between-group differences in BCTQ score or pain NRS at 12 or 24 months. CSI was more costly [mean difference £68.59 (95% CI: −120.84, 291.24)] with fewer quality-adjusted life-years than NS over 24 months [mean difference −0.022 (95% CI: −0.093, 0.045)].

Conclusion: Over 24 months, surgical intervention rates were low in both groups, but less frequent in the NS group. While there were no differences in the clinical effectiveness of CSI and NS, initial treatment with CSI may not be cost-effective in the long-term compared with NS.


CTS is a symptomatic compression neuropathy of the median nerve at the wrist.[1] Clinical symptoms include localized pain and/or discomfort, paraesthesia, and functional loss. CTS is the most common peripheral entrapment neuropathy[2] with 36 patients per 10 000 person-years consulting with CTS in primary care in 2013.[3]

Management options are surgical or conservative (non-surgical). Surgical carpal tunnel decompression is usually offered to those with severe CTS or those not improving with conservative management. Conservative treatment options include local corticosteroid injections (CSI) and night splinting (NS), which are the most used conservative interventions and are recommended in national care pathways and guidelines.[4,5] In one systematic review, 57–66% of affected people initially treated with a conservative approach were reported to eventually have required surgery[6]

Cochrane and other systematic reviews of randomized and quasi-randomized trials have evaluated the effectiveness of NS[7] and CSI for CTS.[8,9] They found limited evidence that NS is more effective than no treatment in the short term (<3 months), but concerns were raised about allocation concealment and blinding.[7] There was evidence of short-term benefit from CSI compared with placebo injection or other conservative treatment options.[8] Evidence of longer-term benefit was insufficient, and reviews concluded that more research was needed.

We investigated the clinical and cost-effectiveness of CSI vs NS for CTS in the INSTinCTS (INjection vs SplinTing in Carpal Tunnel Syndrome) trial.[10,11] We found significantly greater improvements in pain and function at 6 weeks with CSI than NS, although there were no significant between-group differences at 6 months. CSI was also cost effective over 6 months when compared with NS.[10] The aim of long-term trial follow-up was to compare the effect of CSI and NS on hand and wrist pain and function, the number of participants referred for and undergoing CTS surgery, and health care resource use, over 24 months.[11]