Background Clinical and radiographic diagnoses of scaphoid fractures are often challenging at the time of injury. Patients are therefore usually reassessed which has cost implications. Various investigations exist but MRI has been suggested as effective in diagnosing these injuries early.

Aim To determine whether early MRI in suspected occult scaphoid fractures is more clinically and cost effective than conventional management with immobilisation and reassessment.

Methods All patients presenting to the Emergency Department at a district general hospital with a suspected occult scaphoid fracture were randomised into two groups, MRI (early scan of the wrist, discharged if no injury) and control (reassessment in clinic).

Results 84 patients were randomised into MRI (45) and control (39) groups. There were no baseline differences apart from greater dominant hand injuries in the MRI group (62% (26) vs 36% (14), p=0.02). There were three (6.7%) scaphoid fractures in the MRI group and four (10.3%) in the control group (p=0.7). More fractures (15.6% (7) vs 5.1% (2), p=0.9) and other injuries were detected in the MRI group who had fewer mean clinic appointments (1.1±0.5 vs 2.3±0.8, p=0.001) and radiographs (1.2±0.8 vs 1.7±1.1, p=0.03). Mean management costs were £504.13 (MRI) and £532.87 (control) (p=0.9). The MRI group had better pain and satisfaction scores (not significant) with comparable time off work and sporting activities.

Conclusion Early MRI in occult scaphoid fractures is marginally cost saving compared with conventional management and may reduce potentially large societal costs of unnecessary immobilisation. It enables early detection and appropriate treatment of scaphoid and other injuries.