Can a day 4 bone scan accurately determine the presence or absence of scaphoid fracture?

To evaluate the accuracy of day 4 bone scans in predicting the presence or absence of fracture in patients with "clinical scaphoid fracture."

Prospective sensitivity study of ED patients with clinical scaphoid fractures. Each patient was immobilized in a thumb spica cast and had day 4 bone scans of both wrists and hands. Blinded day 4 bone scan results were ultimately compared with the diagnosis on day 14 when patients returned for repeat clinical examination and radiographs. In cases of equivocal radiographic or clinical examination results, a day 14 bone scan was performed.

Two tertiary care teaching hospital emergency departments.

All ED patients older than 16 years with the diagnosis of clinical scaphoid fracture were eligible.

Ninety-nine patients were enrolled and successfully completed the study protocol from October 1990 through November 1992. One patient had bilateral injury, for a total of 100 completed studies. Day 4 bone scans were 100% sensitive and 92% specific, for a positive predictive value of 65%, a negative predictive value of 100%, and accuracy of 93% (95% confidence interval, 88%, to 98%). Many other types of fractures were identified on the day 4 scans, including those of the triquetra, distal radius, capitate, hamate, trapezoid, trapezium, and metacarpals.

Day 4 bone scans are an accurate means of ruling out scaphoid fracture. However, because of a significant number of false-positive scans at day 4, they do not reliably confirm the diagnosis of scaphoid fracture. The bone scans also permitted identification of several other wrist fractures that had not been radiographically apparent.