Clinical Question

In nonoperative management of scaphoid fractures, does using a thumb spica cast lead to superior outcomes compared with a cast that does not include the thumb?

Evidence-Based Answer

The type of cast used does not appear to impact outcomes. A short arm cast allows the patient to have better function and mobility during immobilization without compromising the healing process. (Strength of Recommendation [SOR]: B, randomized controlled trials and cross-sectional studies.) Thumb spica casts are no better than casts that do not include the thumb in fracture healing rates or long-term improvement in range of motion, strength, or pain. (SOR: B, multiple randomized controlled trials.) Testing on uninjured individuals shows no difference in the range of motion or functional movement of the wrist when immobilized in a thumb spica cast compared with a cast that does not include the thumb. (SOR: C, cross-sectional anatomic studies.) Patients wearing thumb spica casts also have more functional limitations than patients able to use their thumb while casted. (SOR: C, cross-sectional studies.)

Evidence Summary

FRACTURE UNION

A 2014 multicenter randomized controlled trial from the Netherlands compared below-elbow casting with and without inclusion of the thumb for the treatment of nondisplaced or minimally displaced fractures of the scaphoid waist (n = 55) and distal scaphoid (n = 7).¹ Included patients (n = 62) were 18 years or older and presented less than four weeks after injury with a scaphoid fracture confirmed by magnetic resonance imaging or computed tomography (CT) with less than 1 mm of gapping or translation between fracture fragments and no other associated fracture or perilunate injury. Patients were treated with cast immobilization for 10 weeks, with or without the thumb included, at which time a CT scan was performed, and fracture union was measured as a percentage of the fracture line crossed by bridging bone. The primary outcome was extent of union at 10 weeks, with secondary outcomes measured at six months, including radiographic union, flexion/extension, grip strength and wrist motion, and arm-specific disability. At 10 weeks, there was a statistically significant difference in extent of union favoring the no-thumb cast (n = 31) compared with the thumb cast (n = 31; 85% vs. 70%; P = .048), although the authors note that this likely does not have any clinical significance. None of the secondary outcomes measured at six months showed any statistically significant differences between the groups. Study limitations included unintentional inclusion of fractures of the distal third of the scaphoid, unreliable measurement of union on CT, and patients lost to follow-up.

A 1991 prospective randomized clinical trial from the United Kingdom compared scaphoid casts, which include the thumb, with forearm gauntlet casts, which do not include the thumb (Colles cast), for the treatment of scaphoid fractures.² Patients (n = 292; mean age = 29.7 years; 76% male) with a radiographically confirmed scaphoid fracture who presented within two weeks of injury were randomized to a scaphoid cast or a Colles cast for treatment. Cast integrity and presence of functional problems were reviewed at two and four weeks after receiving the cast. At eight weeks, the cast was removed, and the fracture was clinically and radiographically assessed for union. Cast treatment was discontinued after confirmation of acceptable healing, defined as appearance of trabeculae crossing the fracture line or signs of increased bone density. Lack of acceptable healing resulted in four more weeks of wrist immobilization in the same type of cast. Final follow-up was six months after injury with radiographic examination of fracture union. At six months, there was no significant difference in fracture nonunion with the Colles cast (n = 148 with 15 nonunions) compared with the scaphoid cast (n = 143 with 14 nonunions; relative risk = 1.04; 95% CI, 0.52 to 2.07).

SCAPHOID MOVEMENT

A 2019 cross-sectional study measured anatomic and functional range of motion for Colles-type casts and splints compared with scaphoid-type casts and splints.³ Ten healthy volunteers (mean age = 28 years) with no previous wrist injury or disability had flexion, extension, radioulnar deviation, and circumduction measured to determine their baseline achievable range of motion as well as baseline functional testing, including opening a jar lid, pouring from a jar, drinking the last drops from a cup, and simulated dart throwing. For each patient, these measurements were taken with their wrist free and with it casted. There were no statistically significant differences between Colles-type and scaphoid-type casts for restriction of wrist movement (P > .406; no comparison data provided). Study limitations were the small study size and use of healthy participants and, thus, a lack of generalization to the population with scaphoid fractures.