Surgery for thumb (trapeziometacarpal joint) osteoarthritis

Surgery is used to treat persistent pain and dysfunction due to osteoarthritis at the base of the thumb, when conservative management is no longer effective. This is an update of a Cochrane review first published in 2005.

To evaluate the benefits and harms of different surgeries for people with thumb (trapeziometacarpal) osteoarthritis.

We searched CENTRAL, MEDLINE and Embase up to 19 May 2025.

We included randomised controlled trials (RCTs) or quasi-RCTs that compared at least two surgical techniques for trapeziometacarpal osteoarthritis. We excluded studies that used non-surgical interventions as the sole comparator (e.g. splinting).

Pain, physical function, quality of life, patient global assessment, adverse events, treatment failure and trapeziometacarpal joint imaging.

We assessed risk of bias using the Cochrane Risk of bias 1.0 tool (RoB 1).

Two authors screened studies, extracted data, assessed risk of bias, and judged the certainty of evidence using the GRADE criteria. We synthesised effect estimates for both dichotomous (risk ratio (RR) using Mantel-Haenszel method) and continuous outcomes (standardised mean difference (SMD) or MD using the inverse-variance method) using a random-effects meta-analysis model.

We included 25 studies with 1591 participants. We identified 11 surgical procedures: trapeziectomy with ligament reconstruction and tendon interposition (LRTI), trapeziectomy, trapeziectomy with ligament reconstruction, trapeziectomy with interpositional arthroplasty, Artelon joint implant, arthrodesis, trapeziometacarpal joint replacement, trapeziectomy and haematoma distraction arthroplasty, trapeziectomy with interpositional arthroplasty with allograft, arthroscopic hemitrapeziectomy, and suture tape suspensionplasty. Other than arthrodesis, no studies compared surgeries that did not involve at least partial trapeziectomy (e.g. first carpometacarpal joint denervation). We found no studies with placebo or non-surgical comparators.

More than 90% of the included studies were at high or unclear risk of performance, detection and selective reporting biases. Below, we summarise the evidence for trapeziectomy with LRTI, the most widely utilised technique. We did not summarise single-study comparisons for trapeziectomy with LRTI. Quality of life, patient global assessment and treatment failure were not reported. Pain was measured on a 100-point visual analogue scale (mm); physical function on a relevant scale (typically the Disabilities of the Arm, Shoulder and Hand questionnaire) presented as a mean difference; adverse events as the proportion of participants who experienced any adverse events; trapeziometacarpal joint imaging as the scapho-metacarpal distance at rest reported in mm. Trapeziectomy with LRTI versus trapeziectomy: compared to trapeziectomy, trapeziectomy with LRTI may make little to no difference to pain (MD -2.80 mm, 95% CI -9.82 to 4.21; I² = 0%; 3 studies, 162 participants), physical function (SMD 0.01, 95% CI -0.70 to 0.72 translated to an MD of 0.03 points (95% CI -1.95 to 1.98); I² = 19%; 3 studies, 211 participants), adverse events (RR 1.89, 95% CI 0.96 to 3.73; I² = 0%; 4 studies, 328 participants) or scapho-metacarpal distance (MD -0.10 mm, 95% CI -0.81 to 0.61; 1 study, 42 participants); all low-certainty evidence. Trapeziectomy with LRTI versus trapeziometacarpal joint replacement: we are uncertain of the effects of trapeziectomy with LRTI compared to trapeziometacarpal joint replacement on pain (MD 1.38 cm, 95% CI 0.7 to 2.07; 2 studies, 206 participants) and physical function (MD 12.53, 95% CI 3.71 to 21.35; I² = 78%; 2 studies, 207 participants); both very low-certainty evidence. There may be little to no difference in adverse events (RR 0.36, 95% CI 0.12 to 1.06; I² = 0%; 3 studies, 261 participants; low-certainty evidence). Trapeziectomy with LRTI versus haematoma distraction arthroplasty: compared to haematoma distraction arthroplasty, we are uncertain of the effects of trapeziectomy with LRTI on pain (MD 0.34 cm, 95% CI -1.19 to 1.87; I² = 97%; 2 studies, 172 participants) and physical function (MD 1.96 points, 95% CI -0.40 to 4.32; I² = 55%; 2 studies, 172 participants); both very low-certainty evidence). LRTI may make little to no difference to adverse events (RR 16.66, 95% CI 0.99 to 279.94; 1 study, 120 participants) or scapho-metacarpal distance (MD 0.0 mm, 95% CI -2.12 to 2.12; 1 study, 120 participants); both low-certainty evidence. Trapeziectomy with LRTI versus trapeziectomy with interpositional arthroplasty: compared to trapeziectomy with interpositional arthroplasty, trapeziectomy with LRTI may slightly reduce pain (MD -0.57cm, 95% CI -1.10 to -0.04; I² = 0%; 2 studies, 98 participants), and may make little to no difference to adverse events (RR 1.29, 95% CI 0.49 to 3.44; 2 studies, 146 participants) and scapho-metacarpal distance (MD -0.70 mm, 95% CI -1.44 to 0.04; 1 study, 46 participants); all low-certainty evidence. We are uncertain of the effects of trapeziectomy with LRTI on physical function (MD -6.75 points, 95% CI -15.88 to 2.39; I² = 70%; 2 studies, 98 participants; very low-certainty evidence).

There are no placebo-controlled trials or studies comparing surgery with non-operative treatment. Low-certainty evidence showed that compared to trapeziectomy, trapeziectomy with LRTI for people with thumb osteoarthritis, may result in little to no difference in pain, function or scapho-metacarpal distance. Compared to trapeziectomy with interpositional arthroplasty, trapeziectomy with LRTI may slightly reduce pain. Trapeziectomy with LRTI may make little to no difference to scapho-metacarpal distance compared to haematoma distraction arthroplasty or trapeziectomy with interpositional arthroplasty. Very low-certainty evidence showed that we are uncertain of the effects of trapeziectomy with LRTI on pain and function compared to trapeziometacarpal joint replacement or haematoma distraction arthroplasty. We are uncertain of the effects of trapeziectomy with LRTI on function compared to trapeziectomy with interpositional arthroplasty. Overall, low-certainty evidence showed that trapeziectomy with LRTI may make little to no difference to adverse events. Low- to very low-certainty evidence from small studies limits our ability to draw conclusions on the benefits and harms of surgical procedures for thumb osteoarthritis.

None.

Protocol (2004) DOI: 10.1002/14651858.CD004631 Original review (2005) DOI: 10.1002/14651858.CD004631.pub2. Review update (2009) DOI: 10.1002/14651858.CD004631.pub3 Review update (2015) DOI: 10.1002/14651858.CD004631.pub4 Review update (2017) DOI: 10.1002/14651858.CD004631.pub5.