Practice Rational Imaging

Investigating suspected scaphoid fracture

BMJ 2013; 346 doi: (Published 27 March 2013) Cite this as: BMJ 2013;346:f1370
  1. Randall L Baldassarre, medical student,
  2. Tudor H Hughes, professor of radiology; vice chair of education; radiology residency program director
  1. 1Department of Radiology, University of California, San Diego Medical Center, 200 West Arbor Drive, San Diego, CA 92103-8756, USA
  1. Correspondence to: T H Hughes thughes{at}

The choice of imaging modality for suspected scaphoid fracture depends on factors such as age, patient activity, cost, and availability of services

Learning points

  • Initial evaluation for suspected acute scaphoid fracture requires radiography, but this modality misses 15-20% of fractures

  • Occult scaphoid fractures may be examined by repeat radiography after casting, magnetic resonance imaging, computed tomography, bone scintigraphy, or ultrasonography

  • The American College of Radiology recommends magnetic resonance imaging or repeat radiography after 10-14 days of casting as the second line investigations of choice

  • The choice of imaging modality for occult scaphoid fracture depends on factors such as age, hand dominance, patient activity, cost, and availability of services

A 26 year old previously healthy woman presented to the emergency department immediately after being involved in a road traffic incident. After clinical examination, scaphoid injury was suspected. A radiograph of the left wrist was obtained and found to be equivocal (fig 1).


Fig 1 Posteroanterior radiograph of the wrist showing an incomplete lucency at the waist of scaphoid suspicious for a fracture but not definitive (arrow)

What is the next investigation?

Scaphoid fractures should be suspected with trauma involving the hand and/or wrist, particularly falls onto an outstretched hand and road traffic incidents. Such fractures most commonly occur among men aged 20-30 years, with about 10% presenting with an associated fracture.1 Less than 20% of patients with a clinically suspected scaphoid fracture have a true fracture.2 In most of the patients with true scaphoid fracture, the fracture is diagnosed with wrist radiography, the initial examination recommended by the Royal College of Radiologists3 and the American College of Radiology’s “appropriateness criteria” (tables 1 and 2), which are based on a literature review incorporating robust meta-analyses, prospective studies, and retrospective case series).4 If initial imaging fails to show a fracture, the hand or wrist is often put in a cast before any further imaging, with the presumption that there is a fracture (presumptive casting). Reflecting a paucity of consensus in the literature, the British and US colleges differ slightly in their recommendations for second line imaging. The British college recommends magnetic resonance imaging (MRI) as the second line approach, with bone scintigraphy and computed tomography as alternatives.3 The US college favors MRI or presumptive casting with repeat radiography to detect subsequently formed fracture lines; it also suggests computed tomography as a third line option.4

Table 1

Suspected acute scaphoid fracture: first examination4

View this table:
Table 2

Suspected acute scaphoid fracture: radiographs normal, next procedure4

View this table:


Obtain four total views—including posteroanterior, lateral, oblique, and ulnar-deviated posteroanterior with cephalad angulation—to evaluate patients with suspected scaphoid fractures.4 5 Compared with other carpal bone injuries, scaphoid fractures are particularly difficult to identify on initial radiographs and may not be seen in 15-20% of cases.6 7 Consider repeating radiographs 10-14 days after initial injury and presumptive casting to allow time for resorption to produce a visible fracture line.4 Meta-analysis has shown that such repeat radiologic assessments have a relatively low sensitivity (91.1%).8 The total cost of radiography is estimated at £24 (€28; $36) per patient.2

Magnetic resonance imaging

MRI is an ideal diagnostic modality when initial plain films are negative. A recent meta-analysis of 30 studies has reported that MRI has excellent sensitivity and specificity (97.7% and 99.8% respectively), especially compared with bone scintigraphy (97.8% and 93.5% respectively) or computed tomography (85.2% and 99.5% respectively).8 MRI is especially sensitive for detecting minor displacements, which is important for determining the need for surgery. Additional carpal, radial, or soft tissue injuries, which may otherwise be missed, can be simultaneously identified, though these injuries are not always clinically relevant. MRI is the most expensive modality, at £200 per patient.2

Computed tomography

Computed tomography is the investigation of choice according to the American College of Radiology’s recommendations when (a) initial radiographs are normal and magnetic resonance imaging is not available, (b) casting is not wanted owing to the patient’s wish for a quick return to activity, and/or (c) surgical planning of complex fractures is needed. Multiplanar and three dimensional reconstructions offer greater capabilities of showing minor displacements and depicting fracture morphology. Computed tomography is more sensitive than MRI for cortical fractures and may be used to evaluate specific injuries such as fracture dislocations of the carpometacarpal joints.4 9 However, computed tomography is otherwise not as sensitive as other modalities in the setting of negative initial radiographs.8 Cost is estimated at £100 per patient.2

Bone scintigraphy

Bone scintigraphy shows fractures via increased tracer uptake at sites of metabolic changes, which are best detected after a period of at least 72 hours after injury. Some authors consider bone scintigraphy to be the optimal procedure for detecting occult fractures owing to its exquisite sensitivity.10 However, a survey of worldwide protocols on scaphoid imaging found that bone scintigraphy was less commonly used than MRI or computed tomography as a second line investigation for suspected scaphoid fractures,11 partly owing to its invasiveness, higher exposure of the whole body to radiation, and expense (£140 per patient).2 Bone scintigraphy is also limited by false-positive results,4 though a few studies have found equal diagnostic ability when compared with MRI.12


Ultrasonography is typically readily available, quick to perform, and does not expose patients to radiation. In addition, ultrasonography is the cheapest imaging modality discussed here, at £10 per patient.2 However, the fracture line, tubercle area, and the proximal and distal poles of the scaphoid are difficult to image, which contributes to the relatively low sensitivity.13 Also, results are assessor dependent. Neither the Royal College of Radiologists nor the American College of Radiology recommends ultrasonography alone to detect scaphoid fractures. A recent study found that general radiologists were able to detect 92% of scaphoid fractures already proved by computed tomography, with a negative predictive value of 97%.14 The authors therefore recommended the use of ultrasonography in emergency room settings as a triage method of determining which patients should receive further imaging for suspected scaphoid fracture.


The patient’s wrist was immobilized in a cast for two days until magnetic resonance imaging could be performed, which showed a scaphoid waist fracture (figs 2 and 3). After six more weeks of wrist immobilization, the cast was removed, and radiography of the wrist showed normal alignment of the fracture site. The patient retained normal function and sensation of the wrist and hand.


Fig 2 Magnetic resonance imaging of the wrist using coronal T1 with fat saturation showing a clear, non-displaced scaphoid waist fracture (arrow)


Fig 3 Magnetic resonance imaging of the wrist using coronal T2 with fat saturation showing a clear, non-displaced scaphoid waist fracture (arrow)


Cite this as: BMJ 2013;346:f1370


  • This series provides an update on the best use of different imaging methods for common or important clinical presentations. The series advisers are Fergus Gleeson, consultant radiologist, Churchill Hospital, Oxford, and Kamini Patel, consultant radiologist, Homerton University Hospital, London. To suggest a topic for this series, please email us at practice{at}

  • Contributors: THH had the idea for the article and selected the patient. RLB designed, drafted, and revised the manuscript. THH reviewed and revised the manuscript. THH is the guarantor.

  • Competing interests: All authors have completed the Unified Competing Interest form at (available on request from the corresponding author) and declare: no support from any organisation for the submitted work; no financial relationships with any organisations that might have an interest in the submitted work in the previous three years; no other relationships or activities that could appear to have influenced the submitted work.

  • Provenance and peer review: Commissioned; externally peer reviewed.

  • Patient consent not required (patient anonymised, dead, or hypothetical).


View Abstract

Sign in

Log in through your institution

Free trial

Register for a free trial to to receive unlimited access to all content on for 14 days.
Sign up for a free trial