ACR Appropriateness Criteria®
Clinical Condition: Acute Hand or Wrist Trauma
Variant 1: Wrist trauma, first exam.
Radiologic Procedure |
Rating |
Comments |
RRL* |
X-ray, wrist |
9 |
Including at least a PA, lateral, and semipronated oblique. Semisupinated oblique view may increase yield for distal radius fractures, especially when added for ulnar-sided pain. |
Min |
CT wrist without contrast |
2 |
|
Min |
MRI wrist without contrast |
2 |
|
None |
NUC Tc-99m bone scan wrist |
2 |
|
Med |
US wrist |
1 |
|
None |
Rating Scale: 1=Least appropriate, 9=Most appropriate |
*Relative Radiation Level |
Note: Abbreviations used in the tables are listed at the end of the "Major Recommendations" field.
Variant 2: Suspect acute distal radius fracture. Radiographs normal. Next procedure.
Radiologic Procedure |
Rating |
Comments |
RRL* |
Cast and repeat x-ray wrist in 10-14 days |
8 |
|
Min |
MRI wrist without contrast |
8 |
If immediate confirmation or exclusion of fracture is required. |
None |
CT wrist without contrast |
5 |
Only if casted and repeat radiographs are negative. |
Min |
NUC Tc-99m bone scan wrist |
2 |
|
Med |
US wrist |
2 |
|
None |
Rating Scale: 1=Least appropriate, 9=Most appropriate |
*Relative Radiation Level |
Note: Abbreviations used in the tables are listed at the end of the "Major Recommendations" field.
Variant 3: Comminuted, intra-articular distal radius fracture on radiographs. Surgical planning
Radiologic Procedure |
Rating |
Comments |
RRL* |
CT wrist without contrast |
9 |
|
Min |
MRI wrist without contrast |
2 |
|
None |
NUC Tc-99m bone scan wrist |
1 |
|
Med |
US wrist |
1 |
|
None |
Rating Scale: 1=Least appropriate, 9=Most appropriate |
*Relative Radiation Level |
Note: Abbreviations used in the tables are listed at the end of the "Major Recommendations" field.
Variant 4: Suspect acute scaphoid fracture, first exam.
Radiologic Procedure |
Rating |
Comments |
RRL* |
X-ray wrist |
9 |
Including at least a PA, lateral, and semipronated oblique. The panel recommends the inclusion of a fourth view: a PA projection with ulnar deviation and/or cephalad tube angulation (scaphoid view) if a scaphoid fracture is suspected. |
Min |
CT wrist without contrast |
1 |
|
Min |
MRI wrist without contrast |
1 |
|
None |
NUC Tc-99m bone scan wrist |
1 |
|
Med |
US wrist |
1 |
|
None |
Rating Scale: 1=Least appropriate, 9=Most appropriate |
*Relative Radiation Level |
Note: Abbreviations used in the tables are listed at the end of the "Major Recommendations" field.
Variant 5: Suspect acute scaphoid fracture. Radiographs normal. Next procedure.
Radiologic Procedure |
Rating |
Comments |
RRL* |
Cast and repeat x-ray wrist in 10-14 days |
8 |
Choice of casting or MRI should be tailored to clinical circumstances in the individual case. |
Min |
MRI wrist without contrast |
8 |
Choice of casting or MRI should be tailored to clinical circumstances in the individual case. |
None |
CT wrist without contrast |
4 |
A reasonable third option if the above two choices are contraindicated. |
Min |
NUC Tc-99m bone scan wrist |
1 |
|
Med |
US wrist |
1 |
|
None |
Rating Scale: 1=Least appropriate, 9=Most appropriate |
*Relative Radiation Level |
Note: Abbreviations used in the tables are listed at the end of the "Major Recommendations" field.
Variant 6: Suspected occult scaphoid fracture. Initial radiographs and repeat radiographs after 10-14 days of casting normal. Continued clinical suspicion of scaphoid fracture. Next procedure.
Radiologic Procedure |
Rating |
Comments |
RRL* |
MRI wrist without contrast |
9 |
|
None |
CT wrist without contrast |
7 |
If MRI cannot be performed. |
Min |
NUC Tc-99m bone scan wrist |
5 |
If MRI or CT cannot be performed |
Med |
US wrist |
1 |
|
None |
Rating Scale: 1=Least appropriate, 9=Most appropriate |
*Relative Radiation Level |
Note: Abbreviations used in the tables are listed at the end of the "Major Recommendations" field.
Variant 7: Suspect distal radioulnar joint subluxation.
Radiologic Procedure |
Rating |
Comments |
RRL* |
X-ray wrist |
9 |
Bilateral wrist CT is indicated in addition to radiographs of the affected side. |
Min |
CT wrist without contrast bilateral |
9 |
Bilateral wrist CT (pronated and supinated) is indicated in addition to radiographs of the affected side |
Min |
MRI wrist without contrast |
2 |
Include contralateral wrist for comparison and scan in both supinated and prone positioning. |
None |
NUC Tc-99m bone scan wrist |
1 |
|
Med |
US wrist |
1 |
|
None |
Rating Scale: 1=Least appropriate, 9=Most appropriate |
*Relative Radiation Level |
Note: Abbreviations used in the tables are listed at the end of the "Major Recommendations" field.
Variant 8: Suspect hook of the hamate fracture. Initial radiographs normal or equivocal.
Radiologic Procedure |
Rating |
Comments |
RRL* |
X-ray wrist |
9 |
Including semisupinated and carpal tunnel views. |
Min |
CT wrist without contrast |
9 |
If additional projections are negative or equivocal. |
Min |
NUC Tc-99m bone scan wrist |
2 |
|
Med |
MRI wrist without contrast |
2 |
|
None |
US wrist |
1 |
|
None |
Rating Scale: 1=Least appropriate, 9=Most appropriate |
*Relative Radiation Level |
Note: Abbreviations used in the tables are listed at the end of the "Major Recommendations" field.
Variant 9: Suspect metacarpal fracture or dislocation.
Radiologic Procedure |
Rating |
Comments |
RRL* |
X-ray hand |
9 |
Including at least a PA, lateral, and semipronated oblique (off-lateral view). |
Min |
CT hand without contrast |
7 |
If strong clinical concern exists following negative or equivocal radiograph. |
Min |
MRI hand without contrast |
3 |
Limited added value. |
None |
NUC Tc-99m bone scan wrist |
1 |
|
Med |
US hand |
1 |
|
None |
Rating Scale: 1=Least appropriate, 9=Most appropriate |
*Relative Radiation Level |
Note: Abbreviations used in the tables are listed at the end of the "Major Recommendations" field.
Variant 10: Suspect phalangeal fracture or dislocation.
Radiologic Procedure |
Rating |
Comments |
RRL* |
X-ray hand or finger |
9 |
Including at least a PA, lateral, and externally rotated oblique view. Internally rotated oblique is appropriate but not always routine. Either a hand or finger radiograph is indicated. |
Min |
CT hand or finger without contrast |
2 |
|
Min |
NUC Tc-99m bone scan hand |
1 |
|
Med |
MRI hand or finger without contrast |
1 |
|
None |
US hand or finger |
1 |
|
None |
Rating Scale: 1=Least appropriate, 9=Most appropriate |
*Relative Radiation Level |
Note: Abbreviations used in the tables are listed at the end of the "Major Recommendations" field.
Variant 11: Suspect thumb fracture or dislocation.
Radiologic Procedure |
Rating |
Comments |
RRL* |
X-ray thumb |
9 |
Including at least AP or PA, lateral and rotated oblique. |
Min |
CT thumb without contrast |
5 |
May be useful for surgical planning for complex, intra-articular fractures of the first metacarpal base. |
Min |
MRI thumb without contrast |
4 |
|
None |
NUC Tc-99m bone scan hand |
1 |
|
Med |
US thumb |
1 |
|
None |
Rating Scale: 1=Least appropriate, 9=Most appropriate |
*Relative Radiation Level |
Note: Abbreviations used in the tables are listed at the end of the "Major Recommendations" field.
Variant 12: Suspect gamekeeper injury (thumb MCP ulnar collateral ligament injury).
Radiologic Procedure |
Rating |
Comments |
RRL* |
X-ray thumb |
9 |
Including at least a PA and lateral. |
Min |
MRI thumb without contrast |
8 |
If no fracture visible on radiographs. |
None |
X-ray thumb with valgus stress and contralateral comparison |
6 |
Controversy concerning accuracy and creation of Stener lesion. |
Min |
US thumb |
6 |
If expertise exists, reliable alternative to MRI. |
None |
MR arthrography thumb |
3 |
|
None |
X-ray arthrography thumb |
2 |
|
Min |
CT thumb without contrast |
2 |
|
Min |
NUC Tc-99m bone scan hand |
1 |
|
Med |
Rating Scale: 1=Least appropriate, 9=Most appropriate |
*Relative Radiation Level |
Note: Abbreviations used in the tables are listed at the end of the "Major Recommendations" field.
Summary of Literature Review
For most patients with trauma of the hand, wrist, or both, radiographs provide adequate diagnostic information and guidance for the treating physician. However, in one large study, wrist fractures, especially those of the distal radius and scaphoid, accounted for more delayed diagnoses than any other traumatized region in patients with initially normal emergency room radiographs. Thus, when initial radiographs are equivocal, or in the presence of certain clinical or radiographic findings, further imaging is appropriate. This may be as simple as additional radiographic projections, or it may include sonography, bone scintigraphy, computed tomography (CT), or magnetic resonance imaging (MRI).
As is true for many extremities, a two-view radiographic examination is not adequate for detecting fracture in the wrist, hand, or fingers. In most patients with suspected distal radius fractures, a three-view radiographic examination (posteroanterior [PA], lateral, and 45-degree semipronated oblique) suffices, while a recent study suggests that the routine addition of a fourth projection--a semisupinated oblique projection--would increase the yield for distal radius fractures, which may be visible only on this fourth view. Nevertheless, when high-field or low-field MRI is performed in addition to radiographs, radiographically occult fractures of the distal radius as well as unsuspected fractures of the carpal bones are frequently demonstrated. In injured wrists with normal or suspicious radiographs that do not account for the clinical symptoms, performing MR imaging results in a change in diagnosis in 55% of cases, and a change in management in 66% of cases. However, a recent randomized controlled trial showed that routine performance of an immediate, abbreviated, low-field MRI study in acutely injured wrists did not predict the need for further treatment any better than the combination of physical examination and radiography. Furthermore, there was no statistically significant difference in outcomes measures--including quality of life, time lost from work, and total costs–with this strategy compared to performing radiographs alone. In addition to MRI, multidetector CT can both show radiographically occult carpal fractures and exclude or confirm suspected fractures, when initial radiographs are equivocal, but the panel rated MRI as a more appropriate modality to use before CT, if there are no contraindications to MRI. CT does have an important role in treatment planning when initial radiographs show a complex fracture-dislocation of the carpus.
Successful treatment of distal radius fractures is predicated on re-establishment of radial length, inclination, and tilt, as well as restoration of the articular surfaces. Specifically, the presence of a coronally oriented fracture line, die-punch depression, and more than three articular surface fragments may indicate the need for operative reduction. Less than 2 mm residual step-off of the distal radial articular surface is considered a congruent reduction necessary for good long-term outcome. Most patients with intra-articular fractures of the distal radius develop radiographic radiocarpal osteoarthritis that progresses over time, even when the original fracture was treated with open reduction and internal fixation; however, the development of radiographic osteoarthritis does not correlate with function, even 15 years after initial injury. CT examination reveals involvement of the radiocarpal and distal radioulnar articular surfaces, intra-articular displacements and depressions, and comminuting more accurately than radiographs.
Measurements of articular surface gap and step-off are more reproducible when performed using CT compared with radiographs, and for displacements >2 mm, there is poor correlation between radiographic and CT findings. Thus, in distal radius fractures where there is a high likelihood of intra-articular incongruence (e.g., fractures in young adults, which frequently result from high-energy impact loading), selective or even routine use of CT to supplement the standard radiographic examination is warranted. The distal radial articular surface is best evaluated by multidetector CT with multiplanar reformatted images; if multidetector CT is not available, direct sagittal images can be obtained, but the imaging process may be difficult if the patient has a cast or external fixator. The addition of 3D surface-rendered reconstructions to the standard 2D CT images may increase interobserver agreement and will change planned management of intra-articular distal radius fractures in up to 48% of cases. MRI also shows intra-articular extension of distal radius fractures more frequently than does radiography and demonstrates concomitant intra-articular soft-tissue injuries—predominantly tears of the scapholunate interosseous ligament—that may affect surgical treatment. However, current evidence suggests that MRI performed immediately at the time of injury has no added value for predicting whether additional treatment will be necessary for soft-tissue injuries, and the panel recommends CT over MRI for surgical planning of complex, intra-articular distal radius fractures.
The diagnosis of distal radioulnar joint (DRUJ) subluxation is problematic. The symptoms and physical findings are often nonspecific, and the condition is difficult to confirm radiographically. Traumatic subluxation or dislocation of the DRUJ may occur as an isolated injury or be associated with other conditions. If optimum positioning of the wrist is not possible because of the injury or overlying cast, CT scanning is recommended. Both wrists should be scanned simultaneously in both pronated and supinated positions. While this examination can also be performed with MRI, repositioning the patient and scanning both wrists is logistically more complex, more time-consuming, and less comfortable with MRI compared to CT.
An additional fourth radiographic projection--an elongated PA view with approximately 30 degrees of cephalad beam angulation and the wrist positioned in 10 to 15 degrees of ulnar deviation--is recommended as a routine whenever there is clinical suspicion of a scaphoid fracture. However, scaphoid fractures are notoriously difficult to see on initial radiographs (regardless of the views), being radiographically occult in up to 20% of cases. Standard practice in patients with clinically suspected scaphoid fractures but normal initial radiographs is to apply a cast and to repeat the clinical evaluation and radiographs in 10 to 14 days, when resorption at the fracture line may make previously occult fractures visible. If the repeat radiographs are still normal, or equivocal at that time and there continues to be a strong clinical suspicion of scaphoid fracture, imaging with a second modality -- bone scintigraphy, CT, or MRI -- is indicated. There is little evidence favoring either scintigraphy or CT in this scenario, while a recent meta-analysis found that MRI is superior to scintigraphy for showing occult scaphoid fractures. A survey of worldwide institutions found that MRI is most commonly used in these cases, although many hospitals still do CT or scintigraphy, and the choice of modality often depends on local preferences, expertise, and equipment.
Recent studies have evaluated the role of tomography, ultrasonography, scintigraphy, CT, and MRI (with standard equipment or a dedicated, extremity-only scanner), in uncertain cases of scaphoid fracture at the time of or shortly after the initial injury. If one or more of these studies is sufficiently sensitive and specific, presumptive casting can be eliminated in normal cases, and definitive care can be instituted earlier for fractures.
A tomographic wrist examination using a Panorex machine has been suggested to help clarify cases where the initial four-view radiograph is suspicious, but the panoramic study requires a custom-built adaptor for the wrist, which limits its applicability. Bone scintigraphy, with either delayed images or blood pool images, can be used to identify or exclude radiographically occult scaphoid fractures, but this use of scintigraphy has been largely replaced by MRI, which is both more sensitive and more specific than scintigraphy. Scintigraphic false- positive diagnoses of carpal fractures occur due to bone contusions, osteoarthritis, avascular necrosis, and osteomyelitis, any of which may be radiographically occult. MRI evaluation for radiographically occult scaphoid fractures can be performed with high-field or low-field equipment, using a whole-body imaging system and appropriate local coil, or using a dedicated extremity MR scanner. Not only can MRI accurately show scaphoid fractures, but in cases where no scaphoid fracture is present, the MRI often demonstrate other, unsuspected fractures of the distal radius or carpus, or soft tissue injuries. In this role, MRI may be cost-effective, especially if immediate MR examination is performed in lieu of presumptive casting, if MRI is done with a limited protocol and at a reduced charge, and if the total cost of presumptive care, including productivity lost from work, is included in the analysis.
Ultrasonography with high-frequency transducers can identify some cases of radiographically occult scaphoid fractures; however, the current evidence does not support the routine use of sonography in these cases. Ultrasound (US) examination is not sensitive enough to preclude presumptive casting when no fracture is seen. Furthermore, US only interrogates the dorsal scaphoid waist, while a large proportion of wrists with clinically suspected occult scaphoid fractures in reality have a fracture of the distal radius or other carpal bone (or another portion of the scaphoid); all these cases would be missed if a negative ultrasound examination were used to avoid casting. CT examination is more sensitive and specific than scintigraphy for diagnosing radiographically occult scaphoid fractures, though it is less sensitive (and shows fewer additional fractures) than MRI in this situation. Nevertheless, the panel felt that CT is a reasonable alternative to immediate MRI with a claustrophobic patient or when there is a contraindication to MRI (in cases where a decision has been made not to apply a cast and repeat radiographs).
In summary, radiographically occult scaphoid fractures are relatively common and cause future morbidity when missed. In patients with a strong clinical suspicion of a scaphoid fracture but normal radiographs, the panel feels that either presumptive casting with repeated radiographs in 10 to 14 days or immediate MRI is equally acceptable strategies. The choice will depend on the age, hand dominance, and activity level of the patient, the availability of MRI, and local preferences. If repeat radiographs are normal and the patient remains symptomatic, further imaging is required, and the panel favors MRI as the study of choice. For patients with contraindications to MRI, CT is preferred to scintigraphy.
For the scaphoid bone, not only is identification of the fracture important, but many surgeons recommend immediate operative intervention for displaced scaphoid fractures. As little as 1 mm of displacement is important, resulting in a higher rate of nonunion and avascular necrosis. While CT scanning confined to the direct sagittal plane will underestimate radial or ulnar displacement of scaphoid fractures, evaluations with MRI or multiplanar and/or 3D reconstructions from multidetector CT are more sensitive than standard radiographs for showing small amounts of displacement. In cases where the position of the scaphoid fracture fragments is suspect despite normal radiographs, the panel recommends CT. Similarly, the panel recommends CT examination when there is a question about the age of a scaphoid fracture or its healing.
Compared with the scaphoid, the diagnosis of other carpal bone injuries is less problematic. In specific circumstances, however, supplemental studies in addition to the standard wrist examination are useful. Pisiform fractures are best seen on semisupinated anterior-posterior (AP) or carpal tunnel projections, which project the pisiform volar to the rest of the carpus. The same projections may also demonstrate fractures involving the hook of the hamate that are not visible on the standard radiographs. However, if radiographs fail to show a hamate fracture that is strongly suspected clinically, axial CT examination is indicated.
A standard three-view radiographic examination will reveal most fractures and dislocations of the metacarpals and phalanges. CT may be useful for surgical planning in fracture-dislocations of the carpometacarpal joints. For phalangeal injuries, some practices include a PA examination of the entire hand, while others limit the entire examination to the injured finger. An internally rotated oblique projection in addition to the standard externally rotated oblique may increase diagnostic confidence for phalangeal fractures. Unlike the case for the wrist, low-field MRI is less sensitive than radiographs for hand and finger fractures, and its role is limited to cases where specific abnormalities of the soft tissues – including the collateral ligaments, volar plates, tendons, and pulleys – would affect treatment.
Most fractures of the thumb will be visible on a two-view radiographic examination, although there is a slight increase in diagnostic yield with the addition of an oblique projection, which can be obtained together with a PA examination of the whole hand. Tears of the ulnar collateral ligament of the thumb metacarpophalangeal joint (gamekeeper injury) represent a special problem. Unless there is an associated bony avulsion of the distal metacarpal or proximal phalangeal base, the injury will be radiographically occult. In these cases, a stress examination of the joint with manually applied abduction stress (which can be applied by the patient or the examiner) may show subluxation compared to the contralateral, uninjured side, although there is a theoretical risk of converting a nondisplaced ulnar collateral ligament tear into a displaced one by a stress examination. More important for treatment planning is whether the adductor aponeurosis has become interposed between the torn, displaced ligament and its osseous attachment site--a so-called Stener lesion. Torn ligaments with a Stener lesion require operative repair, while most nondisplaced tears without an interposed aponeurosis will heal with conservative treatment. Conventional arthrography, US, MRI, and MR arthrography have each been advocated to distinguish ulnar collateral ligament tears with and without Stener lesions. The choice of which modality to use will depend on local availability and expertise.
Summary and Recommendations
Radiographs should be the first imaging study in patients with acute wrist, hand, or finger injuries, but the examination must include the correct radiographic projections, which in turn depends on an accurate, site-specific clinical history. Nondisplaced wrist fractures–especially those of the distal radius and scaphoid–may be radiographically occult initially; in cases where there is a strong clinical suspicion of a fracture despite normal radiographs, further evaluation with immobilization and repeat radiographs, CT, or MR imaging is indicated, depending on the clinical circumstances. CT has additional roles for evaluating the articular surfaces in intra-articular fractures and for detecting specific injuries, including fractures of the hook of the hamate, subluxations of the distal radioulnar joint, and fractures and dislocations of the metacarpal bases. For many indications, including scaphoid bone injuries and ligament injuries at the base of the thumb, MR imaging is the most sensitive examination.
Abbreviations
- AP, anterior-posterior
- CT, computed tomography
- MCP, metacarpophalangeal
- Med, medium
- Min, minimal
- MRI, magnetic resonance imaging
- NUC, nuclear medicine
- PA, posteroanterior
- Tc, technetium
- US, ultrasound
Relative Radiation Level |
Effective Dose Estimated Range |
None |
0 |
Minimal |
<0.1 mSv |
Low |
0.1-1 mSv |
Medium |
1-10 mSv |
High |
10-100 mSv |